Publications
List of publications authored and co-authored.
2024
- EJCRobust data-driven design of a jamming detection filter for airborne electromechanical actuatorsP. Boni, M. Mazzoleni, and F. PrevidiEuropean Journal of Control, 2024
The electrification of the actuations on aircraft is a technological trend towards a greener aviation. Electromechanical Actuators (EMAs) are a key item in the more electric aircraft concept. However, EMAs are still prone to the jamming of their mechanical transmission, leading to possible incidents when flight surfaces can no longer be controlled. Jamming detection algorithms are thus envisaged. In this paper, we conduct a large experimental campaign on a 1:1 scale airborne EMA and design a robust data-driven jamming detection algorithm. The designed algorithm is compared with a traditional non-robust model-based jamming detection algorithm design. The benefits of robust data-driven approach are experimentally evaluated on real faulty data with several performance indexes.
@article{BONI2024100926, title = {Robust data-driven design of a jamming detection filter for airborne electromechanical actuators}, journal = {European Journal of Control}, volume = {75}, pages = {100926}, year = {2024}, issn = {0947-3580}, doi = {10.1016/j.ejcon.2023.100926}, url = {https://www.sciencedirect.com/science/article/pii/S0947358023001541}, author = {Boni, P. and Mazzoleni, M. and Previdi, F.}, keywords = {Electromechanical actuators, Jamming detection}, } - JDSMCA New Physics-Informed Condition Indicator for Cost-Effective Direct Current Solenoid Valves Using Significant Points of the Excitation CurrentLuca Maurelli, Mirko Mazzoleni, Fabio Previdi, and 1 more authorJournal of Dynamic Systems, Measurement, and Control, 2024
This paper proposes a new condition indicator (CI) for cost-effective industrial direct current (DC)-powered solenoid valves (SVs). First, a literature review of the failure modes for SVs is reported. The failure regarding the impeded movement of the valve plunger is considered. The proposed CI for this failure mode is computed from only some significant points of the SVs excitation current and their time instants. The proposed CI is evaluated by a large endurance campaign on a set of SVs. Then, a Remaining Useful Life (RUL) estimation algorithm is proposed and evaluated on the experimental data. Experimental results show the goodness of the proposed CI for RUL estimation of cost-effective industrial DC SVs.
@article{maurelli2024new, title = {A New Physics-Informed Condition Indicator for Cost-Effective Direct Current Solenoid Valves Using Significant Points of the Excitation Current}, author = {Maurelli, Luca and Mazzoleni, Mirko and Previdi, Fabio and Camisani, Andrea}, journal = {Journal of Dynamic Systems, Measurement, and Control}, volume = {146}, number = {3}, pages = {031007}, year = {2024}, publisher = {American Society of Mechanical Engineers}, doi = {10.1115/1.4064602}, url = {https://asmedigitalcollection.asme.org/dynamicsystems/article-abstract/146/3/031007/1195239/A-New-Physics-Informed-Condition-Indicator-for}, } - Revisión de alcance: evaluación de técnicas de aprendizaje automático en el mantenimiento predictivoDaniel Campos-Olivares, Alejandro Carrasco-Muñoz, Mirko Mazzoleni, and 2 more authorsDYNA, 2024
@article{campos2024revision, title = {Revisi{\'o}n de alcance: evaluaci{\'o}n de t{\'e}cnicas de aprendizaje autom{\'a}tico en el mantenimiento predictivo}, author = {Campos-Olivares, Daniel and Carrasco-Mu{\~n}oz, Alejandro and Mazzoleni, Mirko and Ferramosca, Antonio and Luque-Sendra, Amalia}, journal = {DYNA}, volume = {99}, number = {2}, pages = {159--165}, year = {2024}, url = {https://revista-dyna.com/index.php/DYNA/article/view/1676}, } - JIIIDesign of supervision solutions for industrial equipment: Schemes, tools and guidelines for the userMirko MazzoleniJournal of Industrial Information Integration, 2024
The advent of Industry 5.0 envisages production systems that are more resilient, embrace human–machine collaboration and promote sustainability driven by technological research. The development of supervision solutions for industrial equipment fills in this picture as a basis for more proactive Condition-Based Maintenance strategies. The goal of this paper is to provide a self-contained set of guidelines to design such supervision solutions. With respect to existing literature on the topic, we provide a design process with a strong focus on experimental data collection and failure reproduction activities. Moreover, the connections between the steps of the proposed process are clearly highlighted to guide the user. First, the paper provides a set of tools to select the critical items and the methodological approaches for supervision. Then, these tools are used and referenced in the proposed design process. Finally, the proposed process is exemplified on two industrial case studies to show its effectiveness. Considerations, hints, and a user guidelines are given at the end of most sections.
@article{MAZZOLENI2024100667, title = {Design of supervision solutions for industrial equipment: Schemes, tools and guidelines for the user}, journal = {Journal of Industrial Information Integration}, volume = {41}, pages = {100667}, year = {2024}, issn = {2452-414X}, doi = {10.1016/j.jii.2024.100667}, url = {https://www.sciencedirect.com/science/article/pii/S2452414X24001110}, author = {Mazzoleni, Mirko}, keywords = {Supervision, Condition-based maintenance}, } - IFAC SAFEPROCESSThe scenario approach for data-driven prognosticsD. Cesani, M. Mazzoleni, and F. PrevidiIFAC-PapersOnLine, 202412th IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes SAFEPROCESS 2024
Prognostics is the process of forecasting the time-to-failure or the time-to-alarm of an industrial item using degradation models. Data-driven approaches to prognostics employ regression models ft on condition indicators computed from raw run-to-failure data to extrapolate the degradation behaviour of the item. The development of a reliable data-driven degradation model typically requires many run-to-failure acquisitions to understand the degrading behavior. Such experimental tests are destructive and expensive for items manufacturers. Thus, decreasing the number of run-to-failure experiments is key in reducing predictive maintenance costs. In this work, focusing on time-to-alarm prediction to anticipate items breakdown, we propose a data-driven method based on the scenario approach to characterise the degradation behaviour of an industrial item in certain operative conditions using only one run-to-failure experiment, updating the time-to-alarm prediction only when needed. The scenario approach gives probabilistic guarantees on the time-to-alarm predictions.
@article{CESANI2024461, title = {The scenario approach for data-driven prognostics}, journal = {IFAC-PapersOnLine}, volume = {58}, number = {4}, pages = {461-466}, year = {2024}, note = {12th IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes SAFEPROCESS 2024}, issn = {2405-8963}, doi = {https://doi.org/10.1016/j.ifacol.2024.07.261}, url = {https://www.sciencedirect.com/science/article/pii/S2405896324003458}, author = {Cesani, D. and Mazzoleni, M. and Previdi, F.}, keywords = {Prognostics, Scenario approach}, } - IFAC SAFEPROCESSIdentification of relevant symptoms of performance degradation in industrial machinesP. Boni, R. Sala, M. Mazzoleni, and 2 more authorsIFAC-PapersOnLine, 202412th IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes SAFEPROCESS 2024
In the last decades, manufacturing companies increasingly recognized the role of maintenance in guaranteeing high performances for their machines. At the same time, companies realized that, through the analysis of operational data, knowledge on the health status and performance of the machines could be generated, and maintenance-related optimizations and services could be offered to customers. In this setting, the identification of causes leading to degradation of key performance indicators (KPIs) of a machinery is of paramount importance in deciding what actions to take to improve machines performances. In this paper, we propose the use of symptomatology indicators that allow to automatically estimate symptoms of KPI decay in industrial machines. The effectiveness of the proposed symptomatology analysis is experimentally evaluated on real data coming from a set of four shrink wrappers, showing the benefits of the proposed indicators both on client and producer side.
@article{BONI2024467, title = {Identification of relevant symptoms of performance degradation in industrial machines}, journal = {IFAC-PapersOnLine}, volume = {58}, number = {4}, pages = {467-472}, year = {2024}, note = {12th IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes SAFEPROCESS 2024}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2024.07.262}, url = {https://www.sciencedirect.com/science/article/pii/S240589632400346X}, author = {Boni, P. and Sala, R. and Mazzoleni, M. and Pirola, F. and Previdi, F.}, keywords = {performances degradation, key performance indicator, symptomatology indicators}, } - IFAC SAFEPROCESSLeak detection for household pipelines based on a smart valve with single pressure and flow sensorsD. Cesani, M. Mazzoleni, and F. PrevidiIFAC-PapersOnLine, 202412th IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes SAFEPROCESS 2024
Household pipelines leaks are a major source of water wastage, and several leak detection techniques have been developed. As a result, manufacturers started to produce smart valves, able to autonomously detect water leaks in house plumbing networks. A smart valve is an electronically controllable valve provided with some sensors, that must be closed if a leak is detected. The valve closure is controlled by an embedded computational unit that analyses the valve sensors data with a Leak Detection system composed by a set of algorithms. This work proposes a Leak Detection system to monitor a household plumbing network both when an house utility is opened and when all utilities are closed, considering a single flow sensor and single pressure sensor in the smart valve. The proposed Leak Detection system consists of three algorithms: (1) an if-then-else construct (2) a fault detection algorithm based on a logistic regression model, and (3) an anomaly detection algorithm based on the multivariate Gaussian distribution. The Effectiveness of the Leak Detection system is experimentally evaluated on a laboratory house plumbing network, considering different utilities.
@article{CESANI2024408, title = {Leak detection for household pipelines based on a smart valve with single pressure and flow sensors}, journal = {IFAC-PapersOnLine}, volume = {58}, number = {4}, pages = {408-413}, year = {2024}, note = {12th IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes SAFEPROCESS 2024}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2024.07.252}, url = {https://www.sciencedirect.com/science/article/pii/S2405896324003367}, author = {Cesani, D. and Mazzoleni, M. and Previdi, F.}, keywords = {Leak detection, Fault Detection, Anomaly Detection, Water Networks}, } - IFAC SYSIDA comparison of indirect and direct filter designs from data for LTI systems: the effect of unknown noise covariance matricesM. Mazzoleni, L. Maurelli, S. Formentin, and 1 more authorIFAC-PapersOnLine, 202420th IFAC Symposium on System Identification SYSID 2024
@article{mazzoleni2024datadriven, title = {A comparison of indirect and direct filter designs from data for LTI systems: the effect of unknown noise covariance matrices}, journal = {IFAC-PapersOnLine}, volume = {58}, number = {15}, pages = {133-138}, year = {2024}, note = {20th IFAC Symposium on System Identification SYSID 2024}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2024.08.517}, author = {Mazzoleni, M. and Maurelli, L. and Formentin, S. and Previdi, F.} } - IFAC SYSIDAn information theory approach for recursive LPV-ARX model identification via LS-SVMF. Corrini, M. Mazzoleni, F. Ferracuti, and 2 more authorsIFAC-PapersOnLine, 202420th IFAC Symposium on System Identification SYSID 2024
@article{mazzoleni2024datadriveo, title = {An information theory approach for recursive LPV-ARX model identification via LS-SVM}, journal = {IFAC-PapersOnLine}, volume = {58}, number = {15}, pages = {486-491}, year = {2024}, note = {20th IFAC Symposium on System Identification SYSID 2024}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2024.08.576}, url = {https://www.sciencedirect.com/science/article/pii/S2405896324013569}, author = {Corrini, F. and Mazzoleni, M. and Ferracuti, F. and Cavanini, L. and Previdi, F.} } - IFAC TDSAn explicit expression of the steady-state error in Smith Predictor applied to linear systems with integral actionG. Sonzogni, M. Mazzoleni, M. Polver, and 2 more authors202418th IFAC Workshop on Time Delay Systems (TDS)
@article{mazzoleni2024datadrivep, title = {An explicit expression of the steady-state error in Smith Predictor applied to linear systems with integral action}, author = {Sonzogni, G. and Mazzoleni, M. and Polver, M. and Ferramosca, A. and F.Previdi}, note = {18th IFAC Workshop on Time Delay Systems (TDS)}, year = {2024}, } - IEEE CDCA meal detection approach based on parity space to detect untreated meals in subjects with Type 1 diabetesPaolo Alberto Mongini, Mirko Mazzoleni, Antonio Ferramosca, and 2 more authors202463rd IEEE Conference on Decision and Control (CDC)
@article{mongini2024mealdetection, title = {A meal detection approach based on parity space to detect untreated meals in subjects with Type 1 diabetes}, author = {Mongini, Paolo Alberto and Mazzoleni, Mirko and Ferramosca, Antonio and Magni, Lalo and Toffanin, Chiara}, note = {63rd IEEE Conference on Decision and Control (CDC)}, year = {2024}, } - IFAC SYSDContinuous-time identification of grey-box and black-box models of an industrial ovenDavide Previtali, Matteo Scandella, Leandro Pitturelli, and 3 more authorsIFAC-PapersOnLine, 202420th IFAC Symposium on System Identification SYSID 2024
2023
- COAGLISp-r: a preference-based optimization algorithm with convergence guaranteesDavide Previtali, Mirko Mazzoleni, Antonio Ferramosca, and 1 more authorComputational Optimization and Applications, 2023
Preference-based optimization algorithms are iterative procedures that seek the optimal calibration of a decision vector based only on comparisons between couples of different tunings. At each iteration, a human decision-maker expresses a preference between two calibrations (samples), highlighting which one, if any, is better than the other. The optimization procedure must use the observed preferences to find the tuning of the decision vector that is most preferred by the decision-maker, while also minimizing the number of comparisons. In this work, we formulate the preference-based optimization problem from a utility theory perspective. Then, we propose GLISp-r, an extension of a recent preference-based optimization procedure called GLISp. The latter uses a Radial Basis Function surrogate to describe the tastes of the decision-maker. Iteratively, GLISp proposes new samples to compare with the best calibration available by trading off exploitation of the surrogate model and exploration of the decision space. In GLISp-r, we propose a different criterion to use when looking for new candidate samples that is inspired by MSRS, a popular procedure in the black-box optimization framework. Compared to GLISp, GLISp-r is less likely to get stuck on local optima of the preference-based optimization problem. We motivate this claim theoretically, with a proof of global convergence, and empirically, by comparing the performances of GLISp and GLISp-r on several benchmark optimization problems.
@article{previtali2023glisp, title = {GLISp-r: a preference-based optimization algorithm with convergence guarantees}, author = {Previtali, Davide and Mazzoleni, Mirko and Ferramosca, Antonio and Previdi, Fabio}, journal = {Computational Optimization and Applications}, volume = {86}, number = {1}, pages = {383--420}, year = {2023}, publisher = {Springer US New York}, doi = { 10.1007/s10589-023-00491-2}, url = {https://link.springer.com/article/10.1007/s10589-023-00491-2}, } - IJRNCData-driven mixed-sensitivity control with automated weighting functions selectionNicholas Valceschini, Mirko Mazzoleni, Simone Formentin, and 1 more authorInternational Journal of Robust and Nonlinear Control, 2023
Abstract System identification plays a key role in robust control, as not only it provides a nominal model for model-based design, but also the estimate of the model uncertainty can be employed for guaranteeing robust stability and performance. In this paper, we investigate the use of kernel-based identification methods in mixed-sensitivity control, and we show that, using the uncertainty description returned by such methods, we can also automate the selection of the optimal weights, which represent the most critical knobs in real-world applications. We finally compare our approach with a benchmark prediction-error method on a numerical case study. Simulation results illustrate that kernel-based identification might be more suited for robust control, due to its low-bias modeling capability.
@article{valceschini2022mixed, author = {Valceschini, Nicholas and Mazzoleni, Mirko and Formentin, Simone and Previdi, Fabio}, title = {Data-driven mixed-sensitivity control with automated weighting functions selection}, journal = {International Journal of Robust and Nonlinear Control}, volume = {33}, number = {6}, pages = {3458-3470}, keywords = {data-driven robust control, kernel-based system identification, mixed sensitivity loop-shaping, robust control}, doi = {10.1002/rnc.6579}, url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/rnc.6579}, year = {2023}, } - Determining the Importance of Physicochemical Properties in the Perceived Quality of WinesAmalia Luque, Mirko Mazzoleni, Francisco Zamora-Polo, and 3 more authorsIEEE Access, 2023
Wine is a relevant part of the diet in many countries, showing significant nutritional properties, providing health benefits to consumers, and having a significant weight in economy. Also, wine plays an important role in many cultures as a part of their social relationships, feasts, or religion where some of them may become a sign of luxury and distinction. For those reasons, objective and subjective quality of wines is an important issue in their production and marketing. To improve wine excellence, some production methods try to relate its physicochemical properties to the quality as it is perceived by humans. Then, modern data prescriptive analysis can be applied to measure the importance (the influence) of each wine attribute. This paper examines and compare several metrics of the attribute importance and its application to the quality-aware design and production of wines. Moreover, for the cases where the perceived quality is characterized using a discrete value, a novel importance metric, based on the Jensen-Shannon Divergence (JSD) is introduced and compared to the existing ones. The results show that JSD clearly overperforms other metrics previously proposed in the literature. Also, it can be asserted that JSD properly reflects the importance of discrete multivalued functions. The results, using this metric in an importance performance analysis of a public wine dataset, show that the main physicochemical attributes of a red wine are citric acidity, alcohol, sulphates and fixed acidity. As for the white wine case, the main attributes are alcohol, free sulfure dioxide and pH.
@article{10287348, author = {Luque, Amalia and Mazzoleni, Mirko and Zamora-Polo, Francisco and Ferramosca, Antonio and Lama, Juan Ramón and Previdi, Fabio}, journal = {IEEE Access}, title = {Determining the Importance of Physicochemical Properties in the Perceived Quality of Wines}, year = {2023}, volume = {11}, number = {}, pages = {115430-115449}, url = {https://ieeexplore.ieee.org/abstract/document/10287348}, keywords = {Measurement;Production;Semiconductor device measurement;Probability density function;Performance analysis;Linear regression;Volume measurement;Importance metric;importance performance analysis;Jensen Shannon divergence;wine quality}, doi = {10.1109/ACCESS.2023.3325676}, } - Screening of Machine Learning Techniques on Predictive Maintenance: a Scoping ReviewDaniel Campos Olivares, Alejandro Carrasco Muñoz, Mirko Mazzoleni, and 2 more authorsDyna, 2023
Predictive maintenance (PdM) is a set of actions and techniques to early detect failures and defects on machines before they occur, and the usage of machine learning and deep learning techniques in predictive maintenance has increased during the last years. Even with this increase of the literature, there is still a gap concerning the application of such techniques for PdM in the industry, as there are no clear guidelines about which information to use for a PdM system, how to process the information, and what machine learning techniques should be used in order to obtain acceptable results. This scoping review is performed in order to observe the current status on the use of Machine Learning and Deep Learning in predictive maintenance in academia and provide answer to the questions related to these guidelines. For this purpose, a literature review of the last five years is carried out, using those articles that cover information about sources of information used for PdM, the treatment given to such data and the machine learning (ML) methods or techniques used. The Web of Science: Core Collection database is used as a source of information, specifically the Science Citation Index Expanded (SCIE). The review shows that there are different information sources used for machine learning and deep learning in PdM, depending on the origin of the data and the availability of it, and as well whether the data sets are private or public. Also, we can observe that data used for both training and making predictions does not only use traditional pre-processing techniques, but that about one fifth of the articles even propose new techniques in this field. Additionally, we compare a wide range of techniques and algorithms which are used in Deep Learning - being ANN the most used- and in Machine Learning, being SVM the most used algorithm, closely followed by Random Forest. Based on the results, we provide indications about how to apply ML for PdM in industry.
@article{campos2023screening, title = {Screening of Machine Learning Techniques on Predictive Maintenance: a Scoping Review}, author = {Campos Olivares, Daniel and Carrasco Mu{\~n}oz, Alejandro and Mazzoleni, Mirko and Ferramosca, Antonio and Luque Sendra, Amalia}, journal = {Dyna}, year = {2023}, publisher = {Dyna Sl}, doi = {10.6036/10950}, url = {https://idus.us.es/handle/11441/156632}, } - IFAC SYSIDArtificial Pancreas under a Zone Model Predictive Control based on Gaussian Process models: toward the personalization of the closed loopMarco Polver, Beatrice Sonzogni, Mirko Mazzoleni, and 2 more authorsIFAC-PapersOnLine, 202322nd IFAC World Congress
This work introduces a novel zone model predictive control (MPC) based on Gaussian Process models (GPs) for the artificial pancreas (AP). The main novelty of the proposal is to exploit a GP that is trained on previously collected metabolic data of type 1 diabetes mellitus (T1DM) patients, to regulate the blood glucose levels by means of a personalized MPC strategy that automatically adjusts the basal insulin and the insulin boluses to be injected to the patients. The average closed-loop performance is improved in terms of classical indexes such as time in range, avoidance of critic hypoglycaemia episodes and avoidance of long-term hyperglycaemia events. The controller was evaluated in-silico by means of the FDA-accepted UVA/Padova metabolic simulator on 11 adult T1DM patients, showing promising results.
@article{POLVER20239642, title = {Artificial Pancreas under a Zone Model Predictive Control based on Gaussian Process models: toward the personalization of the closed loop}, journal = {IFAC-PapersOnLine}, volume = {56}, number = {2}, pages = {9642-9647}, year = {2023}, note = {22nd IFAC World Congress}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2023.10.271}, url = {https://www.sciencedirect.com/science/article/pii/S2405896323006225}, author = {Polver, Marco and Sonzogni, Beatrice and Mazzoleni, Mirko and Previdi, Fabio and Ferramosca, Antonio}, keywords = {Artificial Pancreas, Model Predictive Control, Data-driven Control, Gaussian Processes}, } - IFAC SYSIDFRAN-X: An improved diagnostic transfer learning approach with application to ball bearings fault diagnosisL. Pitturelli, M. Mazzoleni, L. Rillosi, and 1 more authorIFAC-PapersOnLine, 202322nd IFAC World Congress
Data-driven diagnostic methods are attractive from an industrial and practical perspective due to their limited amount of required prior knowledge about the process or component under monitoring. However, these methods usually require a large amount of healthy and possibly faulty labeled data. Often, gathering and manually labeling a vast dataset is not feasible in real scenarios. Transfer learning has emerged as an answer to the labeling problem, exploiting the idea that the diagnostic knowledge could be reused across multiple different, but related, machines and operating conditions. In this work, we introduce several improvements to the Feature Representation and Alignment Network (FRAN) architecture described in (Chen et al., 2020) devised with the diagnostic transfer learning purpose. Our approach, named FRAN-X, presents improved transfer and diagnostics performance between identical machines in different operating conditions, and it is computationally lighter than its original counterpart. The FRAN-X approach is evaluated on the CWRU-bearing dataset and on experimental data collected from a Computerized Numerical Control (CNC) workcenter machine.
@article{PITTURELLI20237716, title = {FRAN-X: An improved diagnostic transfer learning approach with application to ball bearings fault diagnosis}, journal = {IFAC-PapersOnLine}, volume = {56}, number = {2}, pages = {7716-7721}, year = {2023}, note = {22nd IFAC World Congress}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2023.10.1175}, url = {https://www.sciencedirect.com/science/article/pii/S2405896323015781}, author = {Pitturelli, L. and Mazzoleni, M. and Rillosi, L. and Previdi, F.}, keywords = {Transfer learning, Fault diagnosis}, } - Chest X-ray at Emergency Admission and Potential Association with Barotrauma in Mechanically Ventilated Patients: Experience from the Italian Core of the First Pandemic PeakPietro Andrea Bonaffini, Francesco Stanco, Ludovico Dulcetta, and 9 more authorsTomography, 2023
Barotrauma occurs in a significant number of patients with COVID-19 interstitial pneumonia undergoing mechanical ventilation. The aim of the current study was to investigate whether the Brixia score (BS) calculated on chest-X-rays acquired at the Emergency Room was associated with barotrauma. We retrospectively evaluated 117 SARS-CoV-2 patients presented to the Emergency Department (ED) and then admitted to the intensive care unit (ICU) for mechanical ventilation between February and April 2020. Subjects were divided into two groups according to the occurrence of barotrauma during their hospitalization. CXRs performed at ED admittance were assessed using the Brixia score. Distribution of barotrauma (pneumomediastinum, pneumothorax, subcutaneous emphysema) was identified in chest CT scans. Thirty-eight subjects (32.5%) developed barotrauma (25 pneumomediastinum, 24 pneumothorax, 24 subcutaneous emphysema). In the barotrauma group we observed higher Brixia score values compared to the non-barotrauma group (mean value 12.18 vs. 9.28), and logistic regression analysis confirmed that Brixia score is associated with the risk of barotrauma. In this work, we also evaluated the relationship between barotrauma and clinical and ventilatory parameters: SOFA score calculated at ICU admittance and number of days of non-invasive ventilation (NIV) prior to intubation emerged as other potential predictors of barotrauma.
@article{tomography9060171, author = {Bonaffini, Pietro Andrea and Stanco, Francesco and Dulcetta, Ludovico and Poli, Giancarla and Brambilla, Paolo and Marra, Paolo and Valle, Clarissa and Lorini, Ferdinando Luca and Mazzoleni, Mirko and Sonzogni, Beatrice and Previdi, Fabio and Sironi, Sandro}, title = {Chest X-ray at Emergency Admission and Potential Association with Barotrauma in Mechanically Ventilated Patients: Experience from the Italian Core of the First Pandemic Peak}, journal = {Tomography}, volume = {9}, year = {2023}, number = {6}, pages = {2211--2221}, url = {https://www.mdpi.com/2379-139X/9/6/171}, pubmedid = {38133075}, issn = {2379-139X}, doi = {10.3390/tomography9060171}, } - IEEE CDCModel Uncertainty-Aware Residual Generators for SISO LTI Systems Based on Kernel Identification and Randomized ApproachesMirko Mazzoleni, Nicholas Valceschini, and Fabio PrevidiIn 2023 62nd IEEE Conference on Decision and Control (CDC), 2023
Robustness of residual signals to model uncertain-ties and noise in the measurements is of paramount importance in model-based fault diagnosis. Model uncertainty has been mainly represented in a structured way by considering known bounds on the model parameters, thus relying on prior knowledge about the plant structure and values of its physical parameters. When the plant is completely unknown, system identification techniques must be used for model-based diagnosis. In this work, we present a data-driven approach to represent the uncertainty in the identified model. This uncertainty is described in the frequency domain using kernel-based identification and robust control tools. The estimated model uncertainty region overlaps with the true uncertainty region with a probability specified by the user. The user choices are thus reduced to the selection of only some interpretable hyperparameters. Then, a residual generator robust to the es-timated model uncertainty and measurements noise is designed by a standard H∞ approach. Simulation results on SISO LTI systems show the effectiveness of the approach in producing a residual signal viable for the detection of additive faults.
@inproceedings{10383264, author = {Mazzoleni, Mirko and Valceschini, Nicholas and Previdi, Fabio}, booktitle = {2023 62nd IEEE Conference on Decision and Control (CDC)}, title = {Model Uncertainty-Aware Residual Generators for SISO LTI Systems Based on Kernel Identification and Randomized Approaches}, year = {2023}, volume = {}, number = {}, pages = {4849-4854}, url = {https://ieeexplore.ieee.org/abstract/document/10383264}, keywords = {Linear systems;Adaptation models;Uncertainty;Simulation;Measurement uncertainty;Generators;System identification}, doi = {10.1109/CDC49753.2023.10383264}, } - IEEE CDCNotch Filter Design with Stability Guarantees for Mechanical Resonance Suppression in SISO LTI Two-Mass Drive SystemsGiulia Sonzogni, Mirko Mazzoleni, Marco Polver, and 2 more authorsIn 2023 62nd IEEE Conference on Decision and Control (CDC), 2023
Although the suppression of mechanical resonances for drive and positioning systems is a well-understood problem in the literature, its importance is still actual as technological developments push towards an increase in performance requirements. In this paper, the design of a notch filter is investigated with the aim of suppressing a single resonant frequency in SISO LTI two-mass drive systems. In the cases where the notch filter is located inside an existing control loop, as assumed in this work, it must not compromise the closed-loop stability of the system, while assuring desired control bandwidth and stability margins. Given a fixed known resonant frequency to suppress, an automatic algorithm is proposed to tune the notch filter parameters to guarantee specified control requirements and stability of the closed-loop system, so as to avoid, whenever possible, the reconfiguration of a preexisting controller.
@inproceedings{10383921, author = {Sonzogni, Giulia and Mazzoleni, Mirko and Polver, Marco and Ferramosca, Antonio and Previdi, Fabio}, booktitle = {2023 62nd IEEE Conference on Decision and Control (CDC)}, title = {Notch Filter Design with Stability Guarantees for Mechanical Resonance Suppression in SISO LTI Two-Mass Drive Systems}, year = {2023}, volume = {}, number = {}, pages = {7745-7750}, url = {https://ieeexplore.ieee.org/abstract/document/10383921}, keywords = {Linear systems;System performance;Computational modeling;Resonant frequency;Bandwidth;Stability analysis;Closed loop systems}, doi = {10.1109/CDC49753.2023.10383921}, } - Use of Artificial Intelligence Techniques in Characterization of Vibration Signals for Application in Agri-Food EngineeringAmalia Luque-Sendra, Daniel Campos Olivares, Mirko Mazzoleni, and 3 more authorsSSRN, 2023
Bottling machinery is an essential element in agri-food industries. The early detection of faulty conditions in this equipment can have a significant impact on its productivity and economic performance. Classifying vibration signals recorded in different health states enables an optimized maintenance procedure. These vibrations are usually characterized using the original signal samples or a very reduced set of features specifically designed for each application. In this paper, instead, it is proposed the use of a set of generic features obtained applying basic signal processing methods to the vibration signal. Artificial intelligence techniques have been used for an application in agri-food engineering, specifically for health monitoring of gripping pliers in bottling plants. As the purpose of this research is to compare different ways of describing vibrations, a particular classifier has been used, the random forest classifier. The results obtained show that the proposed method clearly overperforms the other two alternatives in detecting and classifying faulty conditions while significantly reduce the required computational cost. Additionally, the proposed feature extraction method is more robust against randomly generated perturbations and requires a smaller number of instances to train the corresponding classifier.
@article{luque2023use, title = {Use of Artificial Intelligence Techniques in Characterization of Vibration Signals for Application in Agri-Food Engineering}, author = {Luque-Sendra, Amalia and Campos Olivares, Daniel and Mazzoleni, Mirko and Ferramosca, Antonio and Previdi, Fabio and Carrasco, Alejandro}, journal = {SSRN}, year = {2023}, doi = {10.2139/ssrn.4738239}, url = {https://ssrn.com/abstract=4738239}, }
2022
- IJCKernel-based identification of asymptotically stable continuous-time linear dynamical systemsMatteo Scandella, Mirko Mazzoleni, Simone Formentin, and 1 more authorInternational Journal of Control, 2022
In many engineering applications, continuous-time models are preferred to discrete-time ones, in that they provide good physical insight and can be derived also from non-uniformly sampled data. However, for such models, model selection is a hard task if no prior physical knowledge is given. In this paper, we propose a non-parametric approach to infer a continuous-time linear model from data, by automatically selecting a proper structure of the transfer function and guaranteeing to preserve the system stability properties. By means of benchmark simulation examples, the proposed approach is shown to outperform state-of-the-art continuous-time methods, also in the critical case when short sequences of canonical input signals, like impulses or steps, are used for model learning.
@article{doi:10.1080/00207179.2020.1868580, author = {Scandella, Matteo and Mazzoleni, Mirko and Formentin, Simone and Previdi, Fabio}, title = {Kernel-based identification of asymptotically stable continuous-time linear dynamical systems}, journal = {International Journal of Control}, volume = {95}, number = {6}, pages = {1668--1681}, year = {2022}, publisher = {Taylor \& Francis}, doi = {10.1080/00207179.2020.1868580}, url = {https://www.tandfonline.com/doi/abs/10.1080/00207179.2020.1868580}, } - Visualizing Classification Results: Confusion Star and Confusion GearAmalia Luque, Mirko Mazzoleni, Alejandro Carrasco, and 1 more authorIEEE Access, 2022
Recent developments in machine learning applications are deeply concerned with the poor interpretability of most of these techniques. To gain some insights in the process of designing data-based models it is common to graphically represent the algorithm’s results, either in their final or intermediate stage. Specially challenging is the task of plotting multiclass classification results as they involve categorical variables (classes) rather than numeric results. Using the well-known MNIST dataset and a simple neural network as an example, this paper reviews the existing techniques to visualize classification results, from those centered on a particular instance or set of instances, to those representing an overall performance metric. As classification results are commonly summarized in the form of a confusion matrix, special attention is paid to its graphical representation. From this analysis, a new visualization tool is derived, which is presented in two forms: confusion star and confusion gear. The confusion star is centered on the classification errors, while the confusion gear focuses on the classification hits. The proposed visualization tools are also evaluated when facing: (i) balanced and imbalanced classifiers issues; (ii) the problem of representing errors with different orders of magnitude. By using shapes instead of colors to represent the value of each matrix cell, the new tools significantly improve the readability of the confusion matrices. Furthermore, we show how the area enclosed by the confusion stars and gears are directly related to standard classification metrics. The new graphic tools can be also usefully employed to visualize the performances of a sequence of classifiers.
@article{9658486, author = {Luque, Amalia and Mazzoleni, Mirko and Carrasco, Alejandro and Ferramosca, Antonio}, journal = {IEEE Access}, title = {Visualizing Classification Results: Confusion Star and Confusion Gear}, year = {2022}, url = {https://ieeexplore.ieee.org/abstract/document/9658486}, volume = {10}, number = {}, pages = {1659-1677}, keywords = {Visualization;Gears;Data visualization;Classification algorithms;Measurement;Training;Statistics;Machine learning;classification performance;confusion matrix;data visualization;confusion star;confusion gear}, doi = {10.1109/ACCESS.2021.3137630}, } - EAAIA kernel-based control approach for multi-period assets allocation based on lower partial momentsMirko Mazzoleni, Gabriele Maroni, Simone Formentin, and 1 more authorEngineering Applications of Artificial Intelligence, 2022
In quantitative finance, multi-period portfolio optimization can be reformulated as a stochastic optimal control problem, and standard feedback tools can be employed for its analysis. The performance of the trading solutions strongly depend on the quality of the model of the returns. Therefore, data-driven solutions have been recently proposed to optimize simple-linear allocation policies, based only on a set of possible market scenarios. In this work, kernel-based methods are proposed to design more complex and effective control actions, providing better trade-offs in terms of risk and investment performance with respect to linear ones, by preserving convexity. The proposed approach relies on the minimization of the Lower Partial Moments (LPM) risk measure. The effectiveness of the method is shown on a set of real historical financial data.
@article{MAZZOLENI2022104659, title = {A kernel-based control approach for multi-period assets allocation based on lower partial moments}, journal = {Engineering Applications of Artificial Intelligence}, volume = {110}, pages = {104659}, year = {2022}, issn = {0952-1976}, doi = {10.1016/j.engappai.2021.104659}, url = {https://www.sciencedirect.com/science/article/pii/S0952197621004589}, author = {Mazzoleni, Mirko and Maroni, Gabriele and Formentin, Simone and Previdi, Fabio}, keywords = {Portfolio optimization, Kernel methods}, } - A unified surrogate-based scheme for black-box and preference-based optimizationDavide Previtali, Mirko Mazzoleni, Antonio Ferramosca, and 1 more authorarXiv preprint arXiv:2202.01468, 2022
Black-box and preference-based optimization algorithms are global optimization procedures that aim to find the global solutions of an optimization problem using, respectively, the least amount of function evaluations or sample comparisons as possible. In the black-box case, the analytical expression of the objective function is unknown and it can only be evaluated through a (costly) computer simulation or an experiment. In the preference-based case, the objective function is still unknown but it corresponds to the subjective criterion of an individual. So, it is not possible to quantify such criterion in a reliable and consistent way. Therefore, preference-based optimization algorithms seek global solutions using only comparisons between couples of different samples, for which a human decision-maker indicates which of the two is preferred. Quite often, the black-box and preference-based frameworks are covered separately and are handled using different techniques. In this paper, we show that black-box and preference-based optimization problems are closely related and can be solved using the same family of approaches, namely surrogate-based methods. Moreover, we propose the generalized Metric Response Surface (gMRS) algorithm, an optimization scheme that is a generalization of the popular MSRS framework. Finally, we provide a convergence proof for the proposed optimization method.
@article{previtali2022unified, title = {A unified surrogate-based scheme for black-box and preference-based optimization}, author = {Previtali, Davide and Mazzoleni, Mirko and Ferramosca, Antonio and Previdi, Fabio}, journal = {arXiv preprint arXiv:2202.01468}, year = {2022}, } - AutomaticaKernel-based system identification with manifold regularization: A Bayesian perspectiveMirko Mazzoleni, Alessandro Chiuso, Matteo Scandella, and 2 more authorsAutomatica, 2022
This paper presents a nonparametric Bayesian interpretation of kernel-based function learning with manifold regularization. We show that manifold regularization corresponds to an additional likelihood term derived from noisy observations of the function gradient along the regressors graph. The hyperparameters of the method are estimated by a suitable empirical Bayes approach. The effectiveness of the method in the context of dynamical system identification is evaluated on a simulated linear system and on an experimental switching system setup.
@article{MAZZOLENI2022110419, title = {Kernel-based system identification with manifold regularization: A Bayesian perspective}, journal = {Automatica}, volume = {142}, pages = {110419}, year = {2022}, issn = {0005-1098}, doi = {10.1016/j.automatica.2022.110419}, url = {https://www.sciencedirect.com/science/article/pii/S0005109822002722}, author = {Mazzoleni, Mirko and Chiuso, Alessandro and Scandella, Matteo and Formentin, Simone and Previdi, Fabio}, keywords = {System identification, Kernel methods}, } - CMPBAn agent-based model to assess large-scale COVID-19 vaccination campaigns for the Italian territory: The case study of Lombardy regionAndrea Cattaneo, Andrea Vitali, Mirko Mazzoleni, and 1 more authorComputer Methods and Programs in Biomedicine, 2022
In Italy, the administration of COVID-19 vaccines began in late 2020. In the early stages, the number of available doses was limited. To maximize the effectiveness of the vaccine campaign, the national health agency assigned priority access to at-risk individuals, such as health care workers and the elderly. Current vaccination campaign strategies do not take full advantage of the latest mathematical models, which capture many subtle nuances, allowing different territorial situations to be analyzed aiming to make context-specific decisions. The main objective is the definition of an agent-based model using open data and scientific literature to assess and optimize the impact of vaccine campaigns for an Italian region. Specifically, the aim is twofold: (i) estimate the reduction in the number of infections and deaths attributable to vaccines, and (ii) assess the performances of alternative vaccine allocation strategies. The COVID-19 Agent-based simulator Covasim has been employed to build an agent-based model by considering the Lombardy region as case study. The model has been tailored by leveraging open data and knowledge from the scientific literature. Dynamic mobility restrictions and the presence of Variant of Concern have been explicitly represented. Free parameters have been calibrated using the grid search methodology. The model mimics the COVID-19 wave that hit Lombardy from September 2020 to April 2021. It suggests that 168,492 cumulative infections 2,990 cumulative deaths have been avoided due to the vaccination campaign in Lombardy from January 1 to April 30, 2021. Without vaccines, the number of deaths would have been 66% greater in the 80–89 age group and 114% greater for those over 90. The best vaccine allocation strategy depends on the goal. To minimize infections, the best policy is related to dose availability. If at least 1/3 of the population can be covered in 4 months, targeting at-risk individuals and the elderly first is recommended; otherwise, the youngest people should be vaccinated first. To minimize overall deaths, priority is best given to at-risk groups and the elderly in all scenarios. This work proposes a methodological approach that leverages open data and scientific literature to build a model of COVID-19 capable of assessing and optimizing the impact of vaccine campaigns. This methodology can help national institutions to design regional mathematical models that can support pandemic-related decision-making processes.
@article{CATTANEO2022107029, title = {An agent-based model to assess large-scale COVID-19 vaccination campaigns for the Italian territory: The case study of Lombardy region}, journal = {Computer Methods and Programs in Biomedicine}, volume = {224}, pages = {107029}, year = {2022}, issn = {0169-2607}, doi = {10.1016/j.cmpb.2022.107029}, url = {https://www.sciencedirect.com/science/article/pii/S0169260722004114}, author = {Cattaneo, Andrea and Vitali, Andrea and Mazzoleni, Mirko and Previdi, Fabio}, keywords = {COVID-19, Vaccination, Variants of concern, Agent-based model, Epidemiologic model, Covasim}, } - IFAC SAFEPROCESSModel-based fault diagnosis of sliding gates electro-mechanical actuators transmission components with motor-side measurementsN. Valceschini, M. Mazzoleni, and F. PrevidiIFAC-PapersOnLine, 202211th IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes SAFEPROCESS 2022
This paper presents a model-based fault detection and isolation scheme for the transmission components of Electro-Mechanical Actuators, applied to the actuation of sliding gates. The most important failures are investigated by a Failure Mode, Effects, and Criticality Analysis procedure. Following Failure Mode, Effects, and Criticality Analysis, the components selected for the development of the diagnostic algorithm are the nylon gear and pinion of the Electro-Mechanical Actuator, and the rack of the gate. The proposed diagnostic algorithm is able to isolate two out of the three types of faults. The overall procedure is validated by experimental results.
@article{VALCESCHINI2022784, title = {Model-based fault diagnosis of sliding gates electro-mechanical actuators transmission components with motor-side measurements}, journal = {IFAC-PapersOnLine}, volume = {55}, number = {6}, pages = {784-789}, year = {2022}, note = {11th IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes SAFEPROCESS 2022}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2022.07.222}, url = {https://www.sciencedirect.com/science/article/pii/S2405896322006073}, author = {Valceschini, N. and Mazzoleni, M. and Previdi, F.}, keywords = {FMECA, Fault detection, Fault isolation, Mechatronic systems}, } - IFAC SAFEPROCESSExperimental fault detection of input gripping pliers in bottling plantsN. Valceschini, M. Mazzoleni, L. Pitturelli, and 1 more authorIFAC-PapersOnLine, 202211th IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes SAFEPROCESS 2022
This paper presents a signal-based fault detection scheme for input gripping pliers of the blow molding machine in plastic bottling plants, using accelerometers data. The focus of the diagnosis is on the bearings that support the pliers movements on their mechanical cam. Therationale of the algorithm lies in interpreting the pliers\x92 bearings as the balls in a traditional rolling bearing. Then, strategies inspired by bearing diagnosis are employed and adapted to the specific case of this work. The developed algorithm is validated with experimental tests, following a fault injection step, directly on the real blow molding machine
@article{VALCESCHINI2022778, title = {Experimental fault detection of input gripping pliers in bottling plants}, journal = {IFAC-PapersOnLine}, volume = {55}, number = {6}, pages = {778-783}, year = {2022}, note = {11th IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes SAFEPROCESS 2022}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2022.07.221}, url = {https://www.sciencedirect.com/science/article/pii/S2405896322006061}, author = {Valceschini, N. and Mazzoleni, M. and Pitturelli, L. and Previdi, F.}, keywords = {Fault detection, Mechatronic systems}, } - EAAIA fuzzy logic-based approach for fault diagnosis and condition monitoring of industry 4.0 manufacturing processesMirko Mazzoleni, Kisan Sarda, Antonio Acernese, and 4 more authorsEngineering Applications of Artificial Intelligence, 2022
Since the introduction of the industry 4.0 paradigm, manufacturing companies are investing in the development of algorithmic diagnostic solutions for their industrial equipment, relying on measured data and process models. However, process and fault models are not usually available for complex productions plants and production data are usually unlabeled. Thus, to classify machine status, unsupervised approaches such as anomaly detection and signal processing strategies have to be employed. Due to the unsupervised nature of the problem, it is meaningful to apply several diagnostic algorithms to cover most of the process anomalous behaviors. Additionally, in some contexts, the experience of process operators in grasping the correct functioning of machines as well as their ability in understanding early signs of deterioration is relevant for the diagnosis of incoming failures. However, seldom these information can be included in failure diagnosis algorithms. In this paper, we propose a diagnostic scheme for condition monitoring of mechanical components. The proposed scheme combines anomaly detection algorithms, envelope analysis of vibration data, and eventually additional qualitative information on machine functioning. The combination of all the fault indicators is obtained leveraging on a fuzzy inference system. The proposed scheme is experimentally validated on a steel making plant with real process data, making use of heuristic information such monitoring reports of machine health status.
@article{MAZZOLENI2022105317, title = {A fuzzy logic-based approach for fault diagnosis and condition monitoring of industry 4.0 manufacturing processes}, journal = {Engineering Applications of Artificial Intelligence}, volume = {115}, pages = {105317}, year = {2022}, issn = {0952-1976}, doi = {10.1016/j.engappai.2022.105317}, url = {https://www.sciencedirect.com/science/article/pii/S0952197622003566}, author = {Mazzoleni, Mirko and Sarda, Kisan and Acernese, Antonio and Russo, Luigi and Manfredi, Leonardo and Glielmo, Luigi and {Del Vecchio}, Carmen}, keywords = {Fuzzy logic, Manufacturing process, Fault diagnosis, Condition monitoring, Industry 4.0}, } - IFAC CMWRSEvaluation of robust sensors placement schemes for leaks isolation in water distribution networksM. Mazzoleni, M. Scandella, and F. PrevidiIFAC-PapersOnLine, 20222nd IFAC Workshop on Control Methods for Water Resource Systems CMWRS 2022
This paper proposes and evaluates robust design schemes for pressure sensors placement and leak isolation in Water Distribution Networks (WDNs). The proposed sensors placement strategy consider the impact of measurements noise and varying water demand at terminal nodes of the network. The best configurations are obtained with a Genetic Algorithm (GA) by optimizing different performance costs. The isolation performance are evaluated in simulation on a Discrict Metered Area (DMA) of a small town in northern Italy.
@article{MAZZOLENI202248, title = {Evaluation of robust sensors placement schemes for leaks isolation in water distribution networks}, journal = {IFAC-PapersOnLine}, volume = {55}, number = {33}, pages = {48-53}, year = {2022}, note = {2nd IFAC Workshop on Control Methods for Water Resource Systems CMWRS 2022}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2022.11.008}, url = {https://www.sciencedirect.com/science/article/pii/S2405896322026313}, author = {Mazzoleni, M. and Scandella, M. and Previdi, F.}, keywords = {Fault diagnosis, Water networks}, }
2021
- TACA Note on the Numerical Solutions of Kernel-Based Learning ProblemsMatteo Scandella, Mirko Mazzoleni, Simone Formentin, and 1 more authorIEEE Transactions on Automatic Control, 2021
In the last decade, kernel-based learning approaches typically employed for classification and regression have shown outstanding performance also in dynamic system identification. The typical way to compute the solution of this learning problem subsumes the inversion of the kernel matrix. However, due to limited machine precision, this might not be possible in many practical applications. In this article, we analyze the aforementioned problem and show that the typical estimate is just one of the possible infinite solutions that can be leveraged, considering both the supervised and the semisupervised settings. We show under which conditions the infinite solutions are equivalent, and if it is not the case, we provide a bound on the mismatch between two generic solutions. Then, we propose two specific solutions that are particularly suited to boost sparsity or performance.
@article{9076872, author = {Scandella, Matteo and Mazzoleni, Mirko and Formentin, Simone and Previdi, Fabio}, journal = {IEEE Transactions on Automatic Control}, title = {A Note on the Numerical Solutions of Kernel-Based Learning Problems}, year = {2021}, volume = {66}, number = {2}, pages = {940-947}, url = {https://ieeexplore.ieee.org/abstract/document/9076872}, keywords = {Kernel;Symmetric matrices;Eigenvalues and eigenfunctions;Cost function;Dynamical systems;Manifolds;Laplace equations;Kernel-based learning;machine learning;numerical analysis;system identification}, doi = {10.1109/TAC.2020.2989769}, } - IFAC MECCNonparametric continuous-time identification of linear systems: theory, implementation and experimental resultsM. Mazzoleni, M. Scandella, S. Formentin, and 1 more authorIFAC-PapersOnLine, 2021Modeling, Estimation and Control Conference MECC 2021
This paper presents an algorithm for continuous-time identification of linear dynamical systems using kernel methods. When the system is asymptotically stable, also the identified model is guaranteed to share such a property. The approach embeds the selection of the model complexity through optimization of the marginal likelihood of the data thanks to its Bayesian interpretation. The output of the algorithm is the continuous-time transfer function of the estimated model. In this work, we show the algorithmic and computational details of the approach, and test it on real experimental data from an Electro Hydro-Static Actuator (EHSA).
@article{MAZZOLENI2021699, title = {Nonparametric continuous-time identification of linear systems: theory, implementation and experimental results}, journal = {IFAC-PapersOnLine}, volume = {54}, number = {20}, pages = {699-704}, year = {2021}, note = {Modeling, Estimation and Control Conference MECC 2021}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2021.11.253}, url = {https://www.sciencedirect.com/science/article/pii/S2405896321022977}, author = {Mazzoleni, M. and Scandella, M. and Formentin, S. and Previdi, F.}, keywords = {Software for system identification, Kernel methods}, } - EEAIInertial load classification of low-cost electro-mechanical systems under dataset shift with fast end of line testingNicholas Valceschini, Mirko Mazzoleni, and Fabio PrevidiEngineering Applications of Artificial Intelligence, 2021
This paper presents a rationale for designing a machine learning algorithm under dataset shift. In particular, we focus on the classification of the inertial load of low-cost Electro-Mechanical Actuators (EMAs) into several weight categories. In these low-cost settings, due to uncertainties in the manufacturing process, raw materials and usage, even if the EMA part number is the same, its serial numbers (i.e. items or exemplars) may show different physical behaviors. Thus, a learning model trained on data from a set of items can perform poorly when applied to other ones. The proposed solution comprises tailored normalization and cross validation procedures for training the classifier, along with suitable End Of Line (EOL) experiments for the characterization of a new produced EMA item. The approach is experimentally validated on the classification of the mass of sliding gates, using only measurements available on the gate EMA.
@article{VALCESCHINI2021104446, title = {Inertial load classification of low-cost electro-mechanical systems under dataset shift with fast end of line testing}, journal = {Engineering Applications of Artificial Intelligence}, volume = {105}, pages = {104446}, year = {2021}, issn = {0952-1976}, doi = {10.1016/j.engappai.2021.104446}, url = {https://www.sciencedirect.com/science/article/pii/S0952197621002943}, author = {Valceschini, Nicholas and Mazzoleni, Mirko and Previdi, Fabio}, keywords = {Classification, Dataset shift, EMA, End of line testing}, } - IFAC SYSIDModeling and simulation of bimetallic strips in industrial circuit breakers⁎⁎The research has been carried in the SMART4CPPS (Smart Solutions for Cyber-Physical Production Systems) project funded by FESR (Fondo Europeo di Sviluppo Regionale).L. Maurelli, M. Mazzoleni, and F. PrevidiIFAC-PapersOnLine, 202119th IFAC Symposium on System Identification SYSID 2021
This paper presents a dynamical model for the dynamics of the bimetallic strip in industrial circuit breakers. The strip acts as thermo-mechanical actuator that opens the circuit breaker in case of overloads. The overall model can be decomposed in two submodels: an electrothermal and a thermo-mechanical one. The first submodel is derived as a gray-box, while the second one as a black-box. Given the overall estimated model, the final aim is to determine appropriate calibration actions on the device prior to its delivery. The developed model is tested on experimental data of real industrial circuit-breakers.
@article{MAURELLI2021803, title = {Modeling and simulation of bimetallic strips in industrial circuit breakers⁎⁎The research has been carried in the SMART4CPPS (Smart Solutions for Cyber-Physical Production Systems) project funded by FESR (Fondo Europeo di Sviluppo Regionale).}, journal = {IFAC-PapersOnLine}, volume = {54}, number = {7}, pages = {803-808}, year = {2021}, note = {19th IFAC Symposium on System Identification SYSID 2021}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2021.08.460}, url = {https://www.sciencedirect.com/science/article/pii/S2405896321012350}, author = {Maurelli, L. and Mazzoleni, M. and Previdi, F.}, keywords = {Mechatronic systems}, } - IFAC SYSIDPiecewise nonlinear regression with data augmentationM. Mazzoleni, V. Breschi, and S. FormentinIFAC-PapersOnLine, 202119th IFAC Symposium on System Identification SYSID 2021
Piecewise regression represents a powerful tool to derive accurate yet modular models describing complex phenomena or physical systems. This paper presents an approach for learning PieceWise NonLinear (PWNL) functions in both a supervised and semi-supervised setting. We further equip the proposed technique with a method for the automatic generation of additional unsupervised data, which are leveraged to improve the overall accuracy of the estimate. The performance of the proposed approach is preliminarily assessed on two simple simulation examples, where we show the benefits of using nonlinear local models and artificially generated unsupervised data.
@article{MAZZOLENI2021421, title = {Piecewise nonlinear regression with data augmentation}, journal = {IFAC-PapersOnLine}, volume = {54}, number = {7}, pages = {421-426}, year = {2021}, note = {19th IFAC Symposium on System Identification SYSID 2021}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2021.08.396}, url = {https://www.sciencedirect.com/science/article/pii/S2405896321011708}, author = {Mazzoleni, M. and Breschi, V. and Formentin, S.}, keywords = {Hybrid System Identification, Nonparametric Methods, Nonlinear System Identification}, } - IFAC BMSA SIAT3HE model of the COVID-19 pandemic in Bergamo, ItalyMarco Polver, Fabio Previdi, Mirko Mazzoleni, and 1 more authorIFAC-PapersOnLine, 202111th IFAC Symposium on Biological and Medical Systems BMS 2021
The aim of this article is to give a better understanding of the dynamics of the SARS-CoV-2 pandemic in the Bergamo province (Italy), one of the most hit areas of the world, between February and April 2020. A new compartmental model, called SIAT3HE, was designed and fitted on accurate data about the pandemic provided by ATS Bergamo, the health protection agency of the Bergamo province. Our results show that SARS-CoV-2 reached Bergamo in January and infected 318,000 people, the 28.8% of the province population. The 43.1% of the infected individuals stayed asymptomatic. As 6,028 people died due to COVID-19 till April 30th, the infection fatality ratio of SARS-CoV-2 in the Bergamo province was 1.9%. These results are in very good agreement with available information: the number of infections is consistent with the results of recent serological surveys and the number of deaths due to COVID-19 is close to the excess mortality of the considered period.
@article{POLVER2021263, title = {A SIAT3HE model of the COVID-19 pandemic in Bergamo, Italy}, journal = {IFAC-PapersOnLine}, volume = {54}, number = {15}, pages = {263-268}, year = {2021}, note = {11th IFAC Symposium on Biological and Medical Systems BMS 2021}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2021.10.266}, url = {https://www.sciencedirect.com/science/article/pii/S2405896321016694}, author = {Polver, Marco and Previdi, Fabio and Mazzoleni, Mirko and Zucchi, Alberto}, keywords = {Healthcare management, Nonlinear system identification, System identification, validation, Compartmental models, COVID-19}, } - IEEE IECONA comparison of envelope and statistical analyses for bearing diagnosis in hot steel rolling mill linesKisan Sarda, Antonio Acernese, Luigi Russo, and 1 more authorIn IECON 2021 – 47th Annual Conference of the IEEE Industrial Electronics Society, 2021
Steel-working industries are characterized by high temperatures and pressures, elevated production speeds, and intense throughput, so that their sudden interruption leads to great money losses. Undoubtedly, they would extremely benefit from Industry 4.0 advancements in predicting anomalies and breakdowns. However, in these industries, the adoption of predictive maintenance methodologies based on the analysis of historical data is a challenging task. Indeed, to avoid costly and dangerous breakdowns, plant managers prefer to apply an early substitution of machine components long before the end of their useful life, making data on fault events, as well as trends on parts degradation, rarely available. This paper reports the outcome of an industrial research project on data-driven fault diagnosis in a steel making production process. The study aims to identify early stage degradations in rotating machines components in hot rolling mill lines. We compare two methodologies: a well-known frequency-domain analysis of vibrations data is correlated with an ad-hoc designed statistical analysis. The comparison has been conducted on experimental data collected in a steel making plant placed in the South of Italy.
@inproceedings{9589440, author = {Sarda, Kisan and Acernese, Antonio and Russo, Luigi and Mazzoleni, Mirko}, booktitle = {IECON 2021 – 47th Annual Conference of the IEEE Industrial Electronics Society}, title = {A comparison of envelope and statistical analyses for bearing diagnosis in hot steel rolling mill lines}, year = {2021}, volume = {}, url = {https://ieeexplore.ieee.org/abstract/document/9589440}, number = {}, pages = {1-6}, keywords = {Vibrations;Industries;Degradation;Statistical analysis;Electric breakdown;Frequency-domain analysis;Hidden Markov models;anomaly detection;bearing diagnosis;frequency-domain analysis;statistical methods}, doi = {10.1109/IECON48115.2021.9589440}, }
2020
- Data on the first endurance activity of a Brushless DC motor for aerospace applicationsMirko Mazzoleni, Matteo Scandella, Fabio Previdi, and 1 more authorData in Brief, 2020
This article describes the data acquired during the first test activity carried out in the Reliable Electromechanical actuator for PRImary SurfacE with health monitoring (REPRISE) H2020 project. The data consist of a set of measures from an Electro-Mechanical Actuator (EMA) employed in small aircrafts, such as phase currents, positions, temperature and loads. A test bench was developed to perform endurance sessions in various loads and working conditions. Specifically, two datasets are provided: (i) measurements used to monitor the EMA degradation through time; (ii) measurements that characterize the EMA closed-loop dynamic behaviour in healthy condition. The data are helpful to develop and test system identification methods and condition monitoring approaches.
@article{MAZZOLENI2020105153, title = {Data on the first endurance activity of a Brushless DC motor for aerospace applications}, journal = {Data in Brief}, volume = {29}, pages = {105153}, year = {2020}, issn = {2352-3409}, doi = {10.1016/j.dib.2020.105153}, url = {https://www.sciencedirect.com/science/article/pii/S2352340920300470}, author = {Mazzoleni, Mirko and Scandella, Matteo and Previdi, Fabio and Pispola, Giulio}, keywords = {Electro-mechanical actuators, Fault diagnosis, Condition monitoring, Predictive maintenance, Aerospace}, } - L4DCBlack-box continuous-time transfer function estimation with stability guarantees: a kernel-based approachMirko Mazzoleni, Matteo Scandella, Simone Formentin, and 1 more authorIn 2nd Conference on Learning for Dynamics and Control (L4DC), 2020
@inproceedings{mazzoleni2020black, title = {Black-box continuous-time transfer function estimation with stability guarantees: a kernel-based approach}, author = {Mazzoleni, Mirko and Scandella, Matteo and Formentin, Simone and Previdi, Fabio}, booktitle = {2nd Conference on Learning for Dynamics and Control (L4DC)}, volume = {120}, pages = {267--276}, year = {2020}, organization = {Proceedings of Machine Learning Research (PMLR)}, url = {https://proceedings.mlr.press/v120/mazzoleni20a.html}, } - ECCEnhanced kernels for nonparametric identification of a class of nonlinear systemsMirko Mazzoleni, Matteo Scandella, Simone Formentin, and 1 more authorIn 2020 European Control Conference (ECC), 2020
This paper deals with nonparametric nonlinear system identification via Gaussian process regression. We show that, when the system has a particular structure, the kernel recently proposed in [1] for nonlinear system identification can be enhanced to improve the overall modeling performance. More specifically, we modify the definition of the kernel by allowing different orders for the exogenous and the autoregressive parts of the model. We also show that all the hyperparameters can be estimated by means of marginal likelihood optimization. Numerical results on two benchmark simulation examples illustrate the effectiveness of the proposed approach.
@inproceedings{9143785, author = {Mazzoleni, Mirko and Scandella, Matteo and Formentin, Simone and Previdi, Fabio}, booktitle = {2020 European Control Conference (ECC)}, title = {Enhanced kernels for nonparametric identification of a class of nonlinear systems}, year = {2020}, url = {https://ieeexplore.ieee.org/abstract/document/9143785}, volume = {}, number = {}, pages = {540-545}, keywords = {Kernel;Gaussian processes;Numerical models;Standards;Bayes methods;Nonlinear dynamical systems}, doi = {10.23919/ECC51009.2020.9143785}, } - SPRINGERElectro-Mechanical Actuators for the More Electric AircraftM Mazzoleni, G Di Rito, and F Previdi2020
@misc{mazzoleni2020electro, title = {Electro-Mechanical Actuators for the More Electric Aircraft}, author = {Mazzoleni, M and Di Rito, G and Previdi, F}, year = {2020}, publisher = {Springer Nature}, doi = {10.1007/978-3-030-61799-8}, url = {https://link.springer.com/book/10.1007/978-3-030-61799-8}, } - IFAC WCIdentification of dynamic textures using Dynamic Mode DecompositionD. Previtali, N. Valceschini, M. Mazzoleni, and 1 more authorIFAC-PapersOnLine, 202021st IFAC World Congress
Dynamic Textures (DTs) are image sequences of moving scenes that present stationary properties in time. In this paper, we apply Dynamic Mode Decomposition (DMD) and Dynamic Mode Decomposition with Control (DMDc) to identify a parametric model of dynamic textures. The identification results are compared with a benchmark method from the dynamic texture literature, both from a mathematical and from a computational complexity point of view. Extensive simulations are carried out to assess the performance of the proposed algorithms with regards to synthesis and denoising purposes, with different types of dynamic textures. Results show that DMD and DMDc present lower error, lower residual noise and lower variance compared to the benchmark approach.
@article{PREVITALI20202423, title = {Identification of dynamic textures using Dynamic Mode Decomposition}, journal = {IFAC-PapersOnLine}, volume = {53}, number = {2}, pages = {2423-2428}, year = {2020}, note = {21st IFAC World Congress}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2020.12.045}, url = {https://www.sciencedirect.com/science/article/pii/S2405896320302974}, author = {Previtali, D. and Valceschini, N. and Mazzoleni, M. and Previdi, F.}, keywords = {Dynamic textures, System Identification, Texture Synthesis, Dynamic Mode Decomposition}, } - IFAC WCMechatronics applications of condition monitoring using a statistical change detection methodM. Mazzoleni, M. Scandella, L. Maurelli, and 1 more authorIFAC-PapersOnLine, 202021st IFAC World Congress
In this paper, we propose the use of a change detection strategy to perform condition monitoring of mechanical components. The method looks for statistical changes in the distribution of features extracted from raw measurements, such as Root Mean Square or Crest Factor indicators. The proposed method works in a batch fashion, comparing data from one experiment to another. When these distributions differ by a specified amount, a degradation score is increased. The approach is tested on three experimental applications: (i) an ElectroMechanical Actuator (EMA) employed in flight applications, where the focus of the monitoring is on the ballscrew transmission; (ii) a CNC workbench, where the focus is on the vertical shaft bearing, (iii) an industrial EMA with focus on the ballscrew bearing. All components have undergone a severe experimental degradation process, that ultimately led to their failure. Results show how the proposed method is able to assess component degradation prior to their failure.
@article{MAZZOLENI202092, title = {Mechatronics applications of condition monitoring using a statistical change detection method}, journal = {IFAC-PapersOnLine}, volume = {53}, number = {2}, pages = {92-97}, year = {2020}, note = {21st IFAC World Congress}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2020.12.100}, url = {https://www.sciencedirect.com/science/article/pii/S2405896320303566}, author = {Mazzoleni, M. and Scandella, M. and Maurelli, L. and Previdi, F.}, keywords = {Predictive maintenance, condition monitoring, actuators, bearings}, } - IFAC WCKBERG: A MatLab toolbox for nonlinear kernel-based regularization and system identificationM. Mazzoleni, M. Scandella, and F. PrevidiIFAC-PapersOnLine, 202021st IFAC World Congress
We present KBERG, a MatLab package for nonlinear Kernel-BasEd ReGularization and system identification. The toolbox provides a complete environment for running experiments on simulated and experimental data from both static and dynamical systems. The whole identification procedure is supported: (i) data generation, (ii) excitation signals design; (iii) kernel-based estimation and (iv) evaluation of the results. One of the main differences of the proposed package with respect to existing frameworks lies in the possibility to separately define experiments, algorithms and test, then combining them as desired by the user. Once these three quantities are defined, the user can simply run all the computations with only a command, waiting for results to be analyzed. As additional noticeable feature, the toolbox fully supports the manifold regularization rationale, in addition to the standard Tikhonov one, and the possibility to compute different (but equivalent) types of solutions other than the standard one.
@article{MAZZOLENI20201231, title = {KBERG: A MatLab toolbox for nonlinear kernel-based regularization and system identification}, journal = {IFAC-PapersOnLine}, volume = {53}, number = {2}, pages = {1231-1236}, year = {2020}, note = {21st IFAC World Congress}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2020.12.1340}, url = {https://www.sciencedirect.com/science/article/pii/S2405896320317468}, author = {Mazzoleni, M. and Scandella, M. and Previdi, F.}, keywords = {Kernel methods, System Identification}, }
2019
- SCLNonlinear system identification via data augmentationSimone Formentin, Mirko Mazzoleni, Matteo Scandella, and 1 more authorSystems & Control Letters, 2019
This paper presents a novel nonparametric approach to the identification of nonlinear dynamical systems. The proposed methodology exploits the potential of manifold learning on an artificially augmented dataset, obtained without running new experiments on the plant. The additional data are employed for approximating the manifold where input regressors lie. The knowledge of the manifold acts as a prior information on the system, that induces a proper regularization term on the identification cost. The new regularization term, as opposite to the standard Tikhonov one, enforces local smoothness of the function along the manifold. A graph-based algorithm tailored to dynamical systems is proposed to generate the augmented dataset. The hyperparameters of the method, along with the order of the system, are estimated from the available data. Numerical results on a benchmark Nonlinear Finite Impulse Response (NFIR) system show that the proposed approach may outperform the state of the art nonparametric methods.
@article{FORMENTIN201956, title = {Nonlinear system identification via data augmentation}, journal = {Systems & Control Letters}, volume = {128}, pages = {56-63}, year = {2019}, issn = {0167-6911}, doi = {10.1016/j.sysconle.2019.04.004}, url = {https://www.sciencedirect.com/science/article/pii/S0167691119300532}, author = {Formentin, Simone and Mazzoleni, Mirko and Scandella, Matteo and Previdi, Fabio}, keywords = {System identification, Semi-supervised learning}, } - BSPCClassification algorithms analysis for brain–computer interface in drug craving therapyMirko Mazzoleni, Fabio Previdi, and Natale Salvatore BonfiglioBiomedical Signal Processing and Control, 2019
This paper presents a novel therapy to recover patients from drug craving diseases, with the use of brain–computer interfaces (BCIs). The clinical protocol consists of trying to mentally repel drug-related images, and a Stroop test is used to evaluate the blue therapy effect. The method requires a BCI hardware package and a software program which communicates with the device. In order to improve the BCI detection rates, data were collected from five different healthy subjects during the training. These measurements are then used to design a better classification algorithm with respect to the default BCI classifier. The investigated algorithms are logistic regression, support vector machines, decision trees, k-nearest neighbors and Naive Bayes. Although the low number of participants is not enough to guarantee statistically significant results, the designed algorithms perform better than the default one, in terms of accuracy, F1-score and area under the curve (AUC). The Naive Bayes method has been chosen as the best classifier between the tested ones, giving a +12.21% performance boost as concerns the F1-score metric. The presented methodology can be extended to other types of craving problems, such as food, pornography and alcohol. Results relative to the effectiveness of the proposed approach are reported on a set of patients with drug craving problems.
@article{MAZZOLENI2019463, title = {Classification algorithms analysis for brain–computer interface in drug craving therapy}, journal = {Biomedical Signal Processing and Control}, volume = {52}, pages = {463-472}, year = {2019}, issn = {1746-8094}, doi = {10.1016/j.bspc.2017.01.011}, url = {https://www.sciencedirect.com/science/article/pii/S1746809417300198}, author = {Mazzoleni, Mirko and Previdi, Fabio and Bonfiglio, Natale Salvatore}, keywords = {Machine learning, Pattern recognition, Brain–computer interface, Signal processing}, } - Experimental Development of a Health Monitoring Method for Electro-Mechanical Actuators of Flight Control Primary Surfaces in More Electric AircraftsMirko Mazzoleni, Fabio Previdi, Matteo Scandella, and 1 more authorIEEE Access, 2019
This paper presents a health monitoring approach for Electro-Mechanical Actuators (EMA). We define four different indicators to continuously evaluate the health state of the system. The four indicators are computed by leveraging the output from a Statistical Process Monitoring (SPM) method based on multivariate statistics, such as the Hotelling’s T 2 statistic and the Q statistic. SPM approaches give a dichotomous answer, i.e. the presence/absence of a fault. In this work, we propose four ways to compute a continuous indicator starting from the discrete SPM output, that is better suited for health monitoring. We test the approach using a dataset collected from a large experimental campaign on a 1:1 scale EMA for primary flight controls of small aircrafts, that led to EMA failure. Results show the effectiveness of the method.
@article{8878102, author = {Mazzoleni, Mirko and Previdi, Fabio and Scandella, Matteo and Pispola, Giulio}, journal = {IEEE Access}, title = {Experimental Development of a Health Monitoring Method for Electro-Mechanical Actuators of Flight Control Primary Surfaces in More Electric Aircrafts}, year = {2019}, volume = {7}, url = {https://ieeexplore.ieee.org/abstract/document/8878102}, number = {}, pages = {153618-153634}, keywords = {Actuators;Monitoring;Aircraft;Fault diagnosis;Aerospace control;Fault detection;Safety;Actuators;aerospace components;aerospace safety;condition monitoring;electromechanical systems;fault detection;predictive maintenance;statistical process monitoring}, doi = {10.1109/ACCESS.2019.2948781}, } - IFAC ALCOSA comparison of manifold regularization approaches for kernel-based system identificationM. Mazzoleni, M. Scandella, and F. PrevidiIFAC-PapersOnLine, 201913th IFAC Workshop on Adaptive and Learning Control Systems ALCOS 2019
In this paper, we present a simulation study to investigate the role of manifold regularization in kernel-based approaches for nonparametric nonlinear SISO (Single-Input Single-Output) system identification. This problem is tackled as the estimation of a static nonlinear function that maps regressors (that contain past values of both input and output of the dynamic system) to the system outputs. Manifold regularization, as opposite to the Tikhonov one, enforces a local smoothing constraint on the estimated function. It is based on the assumption that the regressors lie on a manifold in the regressors space. This manifold is usually approximated with a weighted graph that connects the regressors. The present work analyzes the performance of kernel-based methods estimates when different choices are made for the graph connections and their respective weights. The approach is tested on benchmark nonlinear systems models, for different connections and weights strategies. Results give an intuition about the most promising choices in order to adopt manifold regularization for system identification.
@article{MAZZOLENI2019180, title = {A comparison of manifold regularization approaches for kernel-based system identification}, journal = {IFAC-PapersOnLine}, volume = {52}, number = {29}, pages = {180-185}, year = {2019}, note = {13th IFAC Workshop on Adaptive and Learning Control Systems ALCOS 2019}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2019.12.641}, url = {https://www.sciencedirect.com/science/article/pii/S2405896319325868}, author = {Mazzoleni, M. and Scandella, M. and Previdi, F.}, keywords = {Kernel methods, System Identification, Manifold regularization}, }
2018
- Learning meets control. Data analytics for dynamical systemsMirko Mazzoleni2018
- IEEE MEDDevelopment and Experimental Testing of a Health Monitoring System of Electro-Mechanical Actuators for Small AirplanesFabio Previdi, Yamuna Maccarana, Mirko Mazzoleni, and 3 more authorsIn 2018 26th Mediterranean Conference on Control and Automation (MED), 2018
This paper reports the preliminary results of the REPRISE (Reliable Electromechanical actuator for PRImary SurfacE with health monitoring) project, which aims to design a novel Electro-Mechanical Actuator (EMA) to be used on primary flight surfaces of small aircrafts. An important element of the actuator control system is a Health Monitoring (HM) module. This component is an algorithm able to detect anomalies in the device even if there is no evident loss of ability in pursuing its main function (position tracking). In particular, the project aim is to identify any degradation in the mechanical transmission elements, the ballscrew and other components such as bearings. Moreover, it is strongly advisable that the HM algorithm is based on a feature whose value can be easily computed and monitored during the actuator life. In this work, a large experimental activity has been carried out with the purpose of bringing the actuator close to failure, by progressive fault injection in overload operating conditions. A feature named Σ, that is, the mean of the RMS of the three phase currents (the input to the electric motor), is proposed as a parameter for HM. The effectiveness of this parameter in detecting the mechanical transmission degradation is experimentally tested. The degradation has been confirmed by visual inspection and screw thread profile measurements. In spite of this, the actuator is still able to perform position tracking in an effective way.
@inproceedings{8442734, author = {Previdi, Fabio and Maccarana, Yamuna and Mazzoleni, Mirko and Scandella, Matteo and Pispola, Giulio and Porzi, Nicola}, booktitle = {2018 26th Mediterranean Conference on Control and Automation (MED)}, title = {Development and Experimental Testing of a Health Monitoring System of Electro-Mechanical Actuators for Small Airplanes}, year = {2018}, volume = {}, number = {}, pages = {673-678}, url = {https://ieeexplore.ieee.org/abstract/document/8442734}, keywords = {Actuators;Monitoring;Degradation;Performance evaluation;Current measurement;Aircraft;Fasteners}, doi = {10.1109/MED.2018.8442734}, } - IFAC SYSIDCondition assessment of electro-mechanical actuators for aerospace using relative density-ratio estimationM. Mazzoleni, M. Scandella, Y. Maccarana, and 3 more authorsIFAC-PapersOnLine, 201818th IFAC Symposium on System Identification SYSID 2018
This paper faces the problem of developing an effective Condition Monitoring algorithm (CM) for Electro-Mechanical Actuators (EMA) in aerospace applications. In this view, a test campaign has been carried out in order to progressively bring the EMA near to failure, by means of a test bench suitably developed. Various indicators have been computed from measured data, for a set of the EMA’s working regimes. The statistical distribution of the computed features is assessed and tracked over time. We propose an online statistical approach, based on density estimation techniques, in order to detect potential changes in the data distribution. The discovered changes are then interpreted as a modification of the EMA’s health state, leading to a first building block for a complete condition assessment strategy.
@article{MAZZOLENI2018957, title = {Condition assessment of electro-mechanical actuators for aerospace using relative density-ratio estimation}, journal = {IFAC-PapersOnLine}, volume = {51}, number = {15}, pages = {957-962}, year = {2018}, note = {18th IFAC Symposium on System Identification SYSID 2018}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2018.09.070}, url = {https://www.sciencedirect.com/science/article/pii/S2405896318317300}, author = {Mazzoleni, M. and Scandella, M. and Maccarana, Y. and Previdi, F. and Pispola, G. and Porzi, N.}, keywords = {Condition monitoring, Change-point detection, Kernel methods, Time-series}, } - IFAC SYSIDIdentification of nonlinear dynamical system with synthetic data: a preliminary investigationM. Mazzoleni, M. Scandella, S. Formentin, and 1 more authorIFAC-PapersOnLine, 201818th IFAC Symposium on System Identification SYSID 2018
This paper introduces a new rationale for learning nonlinear dynamical systems. The method makes use of an additional identification dataset, obtained without performing a new experiment on the system under study. The data are generated in an automatical manner, starting from a set of experimentally acquired measurements. In order to leverage the additional generated information, fundamental techniques from the machine learning field known as Semi-Supervised Learning (SSL) are employed and adapted. The problem is then cast as a regularized parametric learning problem. The effectiveness of the proposed approach is assessed on various nonlinear benchmark systems via repeated simulations, comparing the obtained results with a standard regularization method for learning parametric models.
@article{MAZZOLENI2018622, title = {Identification of nonlinear dynamical system with synthetic data: a preliminary investigation}, journal = {IFAC-PapersOnLine}, volume = {51}, number = {15}, pages = {622-627}, year = {2018}, note = {18th IFAC Symposium on System Identification SYSID 2018}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2018.09.227}, url = {https://www.sciencedirect.com/science/article/pii/S2405896318318998}, author = {Mazzoleni, M. and Scandella, M. and Formentin, S. and Previdi, F.}, keywords = {System Identification, Semi-Supervised Learning, Regularization}, } - IEEE CCTACondition Monitoring of Electro-Mechanical Actuators for Aerospace Using Batch Change Detection AlgorithmsMirko Mazzoleni, Matteo Scandella, Yamuna Maccarana, and 3 more authorsIn 2018 IEEE Conference on Control Technology and Applications (CCTA), 2018
This paper proposes the use of a change detection algorithm to monitor the degradation of mechanical components of Electro-Mechanical Actuators (EMA) employed in the aerospace industry. Contrary to the standard on-line application of change detection methods, the presented approach can be applied in a batch mode, leveraging on the knowledge of when the data were collected. The methodology is applied to data measured during an endurance test campaign on a real EMA employed in aerospace, by means of a developed test bench, progressively bringing the EMA to failure. Three rationales for building an indicator of degradation are tested. Results show how the method is able to assess the degradation of the actuator over time, constituting a first step towards a condition monitoring solution for the more-electric-aircraft of the future.
@inproceedings{8511334, author = {Mazzoleni, Mirko and Scandella, Matteo and Maccarana, Yamuna and Previdi, Fabio and Pispola, Giulio and Porzi, Nicola}, booktitle = {2018 IEEE Conference on Control Technology and Applications (CCTA)}, title = {Condition Monitoring of Electro-Mechanical Actuators for Aerospace Using Batch Change Detection Algorithms}, year = {2018}, volume = {}, number = {}, pages = {1747-1752}, url = {https://ieeexplore.ieee.org/abstract/document/8511334}, keywords = {Degradation;Actuators;Monitoring;Condition monitoring;Estimation;Aerospace industry;Standards}, doi = {10.1109/CCTA.2018.8511334}, } - ECCSemi-supervised learning of dynamical systems: a preliminary studyMirko Mazzoleni, Simone Formentin, Matteo Scandella, and 1 more authorIn 2018 European Control Conference (ECC), 2018
System identification has, in recent years, drawn insightful inspirations from techniques and concepts of the statistical learning research area. Examples of this consist in the widely adoption of regularization and kernels methods, in order to better condition the identification problem. By pursuing the same purpose, we introduce the concept of semi-supervised learning to tackle the system identification challenge. The problem, casted into the framework of the Reproducing Kernel Hilbert Spaces, leads to a new regularization technique, called manifold regularization. An application to the identification of a NFIR model is carried out, and a comparison with the standard Tikhonov regularization technique is shown.
@inproceedings{8550550, author = {Mazzoleni, Mirko and Formentin, Simone and Scandella, Matteo and Previdi, Fabio}, booktitle = {2018 European Control Conference (ECC)}, title = {Semi-supervised learning of dynamical systems: a preliminary study}, year = {2018}, volume = {}, number = {}, pages = {2824-2829}, url = {https://ieeexplore.ieee.org/abstract/document/8550550}, keywords = {Manifolds;Kernel;Semisupervised learning;Statistical learning;Symmetric matrices;Standards;Hilbert space}, doi = {10.23919/ECC.2018.8550550}, } - IEEE CDCClassification of Light Charged Particles Via Learning-Based System IdentificationMirko Mazzoleni, Matteo Scandella, Simone Formentin, and 1 more authorIn 2018 IEEE Conference on Decision and Control (CDC), 2018
This paper presents a nonparametric learning approach for the automatic classification of particles produced by the collision of a heavy ion beam on a target, by focusing on the identification of isotopes of the most energic light charged particles (LCP). In particular, it is shown that the measurement of the particle collision can be traced back to the impulse response of a linear dynamical system and, by employing recent kernel-based approaches, a nonparametric model is found that effectively trades off bias and variance of the model estimate. Then, the smoothened signals can be employed to classify the different types of particles. Experimental results show that the proposed method outperforms the state of the art approaches. All the experiments are carried out with the large detector array CHIMERA (Charge Heavy Ions Mass and Energy Resolving Array) in Catania, Italy.
@inproceedings{8618946, author = {Mazzoleni, Mirko and Scandella, Matteo and Formentin, Simone and Previdi, Fabio}, booktitle = {2018 IEEE Conference on Decision and Control (CDC)}, title = {Classification of Light Charged Particles Via Learning-Based System Identification}, year = {2018}, volume = {}, number = {}, url = {https://ieeexplore.ieee.org/abstract/document/8618946}, pages = {6053-6058}, keywords = {Kernel;Detectors;Atmospheric measurements;Particle measurements;Ions;Covariance matrices;Atomic measurements}, doi = {10.1109/CDC.2018.8618946}, }
2017
- IEEE AIMDevelopment of a reliable electro-mechanical actuator for primary control surfaces in small aircraftsM. Mazzoleni, Y. Maccarana, F. Previdi, and 4 more authorsIn 2017 IEEE International Conference on Advanced Intelligent Mechatronics (AIM), 2017
This paper lays the foundation for the development of an innovative electro-mechanical actuator for flight-control surfaces. The main features of the enhanced system will be the introduction of new sensor types and health monitoring capabilities. A dedicated test bench has been developed in order to perform endurance tests, leading the mechanical components to failure. In this view, a Condition Monitoring (CM) algorithm is expected to assess the progressive faults degradation, estimating their progression and the Remaining Useful Life (RUL) of related subsystems. Based on the development of new hardware and software components, the REPRISE project is expected to deliver a significant contribution to the More Electric Aircraft mission.
@inproceedings{8014172, author = {Mazzoleni, M. and Maccarana, Y. and Previdi, F. and Pispola, G. and Nardi, M. and Perni, F. and Toro, S.}, booktitle = {2017 IEEE International Conference on Advanced Intelligent Mechatronics (AIM)}, title = {Development of a reliable electro-mechanical actuator for primary control surfaces in small aircrafts}, year = {2017}, volume = {}, number = {}, pages = {1142-1147}, url = {https://ieeexplore.ieee.org/abstract/document/8014172}, keywords = {Actuators;Monitoring;Aircraft;Reliability;Hydraulic systems;Software;Condition monitoring}, doi = {10.1109/AIM.2017.8014172}, } - IFAC WCControl-oriented modeling of SKU-level demand in retail food marketM. Mazzoleni, S. Formentin, F. Previdi, and 1 more authorIFAC-PapersOnLine, 201720th IFAC World Congress
In food market, modeling the dynamics of Stock-Keeping Unit (SKU) requests is of fundamental importance, not only to understand the market but also for optimization and control purposes. In fact, standing on model-based predictions of future demand, an efficient planning of the promotional calendar can be devised. Moreover, better inventory management can be achieved, by reducing losses due to expired aliments remained unsold and improving distribution operations. In this work, data-driven control-oriented modeling of such a demand is discussed and a novel switching dynamical strategy is proposed. When applied to experimental data from a real food company, the above strategy is shown to accurately predict future sales under fixed promotion events and outperform the state-of-the-art modeling methods.
@article{MAZZOLENI201713003, title = {Control-oriented modeling of SKU-level demand in retail food market}, journal = {IFAC-PapersOnLine}, volume = {50}, number = {1}, pages = {13003-13008}, year = {2017}, note = {20th IFAC World Congress}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2017.08.1951}, url = {https://www.sciencedirect.com/science/article/pii/S2405896317325818}, author = {Mazzoleni, M. and Formentin, S. and Previdi, F. and Savaresi, S.M.}, keywords = {Knowledge discover (data mining), Identification, model reduction, Production planning, control}, } - IFAC WCA comparison of data-driven fault detection methods with application to aerospace electro-mechanical actuatorsM. Mazzoleni, Y. Maccarana, and F. PrevidiIFAC-PapersOnLine, 201720th IFAC World Congress
In this paper, a model-free framework is proposed in order to equip electromechanical actuators, deployed in aerospace applications, with health-monitoring capabilities. A large experimental activity has been carried out to perform acquisitions with both healthy and faulty components, taking into consideration the standard regulations for environmental testing of avionics hardware. The injected faults followed a Fault Tree Analysis and Failure Mode and Effect Analysis. Features, belonging to different domains, have been extracted from the measured signals. These indexes are based largely on the motor driving currents, in order to avoid the installation of new sensors. Finally, a Gradient Tree Boosting algorithm has been chosen to detect the system status: the choice has been dictated by a comparison with other known classification algorithms. Furthermore, the most promising features for a classification point of view are reported.
@article{MAZZOLENI201712797, title = {A comparison of data-driven fault detection methods with application to aerospace electro-mechanical actuators}, journal = {IFAC-PapersOnLine}, volume = {50}, number = {1}, pages = {12797-12802}, year = {2017}, note = {20th IFAC World Congress}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2017.08.1837}, url = {https://www.sciencedirect.com/science/article/pii/S2405896317324606}, author = {Mazzoleni, M. and Maccarana, Y. and Previdi, F.}, keywords = {Fault Detection, Machine Learning, Electro-Mechanical Actuators}, } - IFAC WCFault detection in airliner electro-mechanical actuators via hybrid particle filteringM. Mazzoleni, G. Maroni, Y. Maccarana, and 2 more authorsIFAC-PapersOnLine, 201720th IFAC World Congress
In this paper, a modification of the standard particle filter algorithm is applied to face the fault detection issue, on an electro-mechanical actuator. The variant, based on a hybrid system interpretation of the health monitoring problem, is known as OTPF (Observation and Transition Particle Filter). By modeling each fault condition as a hybrid system mode, the method is able to assess the most likely regime for each time stamp. Following this approach, data were acquired from an electro-mechanical actuator, used in aerospace environment, under various fault conditions. The injected mechanical defects consisted in damages undergone by steel spheres, inside a ballscrew transmission system. Then, a model for each condition was identified and the proposed methodology applied. Simulation results show the superiority of the method with respect to the EKF (Extended Kalman Filter), especially because the distribution of the disturbances which affect the system is usually not Gaussian.
@article{MAZZOLENI20172860, title = {Fault detection in airliner electro-mechanical actuators via hybrid particle filtering}, journal = {IFAC-PapersOnLine}, volume = {50}, number = {1}, pages = {2860-2865}, year = {2017}, note = {20th IFAC World Congress}, issn = {2405-8963}, doi = {doi.org/10.1016/j.ifacol.2017.08.640}, url = {https://www.sciencedirect.com/science/article/pii/S240589631731025X}, author = {Mazzoleni, M. and Maroni, G. and Maccarana, Y. and Formentin, S. and Previdi, F.}, keywords = {Fault detection, diagnosis, Particle filtering, Diagnosis of discrete event, hybrid systems}, } - Unsupervised learning of fundamental emotional states via word embeddingsMirko Mazzoleni, Gabriele Maroni, and Fabio PrevidiIn 2017 IEEE Symposium Series on Computational Intelligence (SSCI), 2017
This paper presents a novel approach for the detection of emotional states from textual data. The considered sentiments are those known as Ekman’s basic emotions (Anger, Disgust, Sadness, Happiness, Fear, Surprise). The method is completely unsupervised and it is based on the concept of word embeddings. This technique permits to represent a single word through a vector, giving a methematical representation of the word’s semantic. The focus of the work is to assign the percentage of the aforementioned emotions to short sentences. The method has been tested on a collection of Twitter messages and on the SemEval 2007 news headlines dataset. The entire period is expressed as the mean of the word’s vectors that compose the phrase, after preprocessing steps. The sentence representation is finally compared with each emotion’s word vector, to find the most representative with respect to the sentence’s vector.
@inproceedings{8280819, author = {Mazzoleni, Mirko and Maroni, Gabriele and Previdi, Fabio}, booktitle = {2017 IEEE Symposium Series on Computational Intelligence (SSCI)}, title = {Unsupervised learning of fundamental emotional states via word embeddings}, year = {2017}, volume = {}, number = {}, pages = {1-6}, url = {https://ieeexplore.ieee.org/abstract/document/8280819}, keywords = {Semantics;Sentiment analysis;Dictionaries;Social network services;Data mining;Computer architecture}, doi = {10.1109/SSCI.2017.8280819}, }
2016
- IEEE ICCAModeling and identification of an Electro-Hydraulic ActuatorA. L. Cologni, M. Mazzoleni, and F. PrevidiIn 2016 12th IEEE International Conference on Control and Automation (ICCA), 2016
n this work, a physical non-linear model of an Electro-Hydraulic Actuator has been developed. Each system component (valves, pipes, cylinders) and their interactions have been modeled by means of conservation and constitutive laws. The actuator dynamics, with lumped-parameter element models, have been treated accurately, with special attention given to modeling friction. Finally, a global parametric-identification procedure has been performed for all unknown parameters. Throughout this paper all the modeling assumptions and results from system identification are verified with experimental data.
@inproceedings{7505299, author = {Cologni, A. L. and Mazzoleni, M. and Previdi, F.}, booktitle = {2016 12th IEEE International Conference on Control and Automation (ICCA)}, title = {Modeling and identification of an Electro-Hydraulic Actuator}, year = {2016}, volume = {}, number = {}, pages = {335-340}, keywords = {Conferences;Automation}, url = {https://ieeexplore.ieee.org/abstract/document/7505299}, doi = {10.1109/ICCA.2016.7505299}, } - An application of the remote maintenance paradigm to semi-automated machinesFabio Previdi, Mirko Mazzoleni, Alberto Luigi Cologni, and 1 more authorIn 14th IMEKO TC10 Workshop on Technical Diagnostics 2016: New Perspectives in Measurements, Tools and Techniques for Systems Reliability, Maintainability and Safety, 2016
@inproceedings{previdi2016application, title = {An application of the remote maintenance paradigm to semi-automated machines}, author = {Previdi, Fabio and Mazzoleni, Mirko and Cologni, Alberto Luigi and Ermidoro, Michele}, booktitle = {14th IMEKO TC10 Workshop on Technical Diagnostics 2016: New Perspectives in Measurements, Tools and Techniques for Systems Reliability, Maintainability and Safety}, pages = {285--289}, year = {2016}, } - IEEE IECONLow computational complexity control of a three-phases open-windings AC brushless motorAlberto Cologni, Mirko Mazzoleni, and Fabio PrevidiIn IECON 2016 - 42nd Annual Conference of the IEEE Industrial Electronics Society, 2016
The computational complexity of a modulating control is always one of the biggest limitation in the real control of brushless motor. This paper presents a low computational complexity control for a three phases open-windings AC brushless motor. The control strategy is based on a modified version of the D-Q frame: this solution allows to exploit the motor control libraries, already available on some motor control microprocessors. The PWM is generated in accordance to a novel algorithm based on a combination of two standard SVM algorithms, one for each inverter. The current controller has been tested on a real open-windings motor, installed as an active vibrations damper for aerospace applications. The experiments demonstrate the good quality of the current tracking and the high efficiency of the inverter, controlled with the proposed approach.
@inproceedings{7793011, author = {Cologni, Alberto and Mazzoleni, Mirko and Previdi, Fabio}, booktitle = {IECON 2016 - 42nd Annual Conference of the IEEE Industrial Electronics Society}, title = {Low computational complexity control of a three-phases open-windings AC brushless motor}, year = {2016}, volume = {}, number = {}, pages = {577-582}, keywords = {Brushless motors;Actuators;Support vector machines;Inverters;Damping;Standards;Windings;Current control;Brushless motor;Open windings;High efficiency inverter;Space vector pulse width modulation}, url = {https://ieeexplore.ieee.org/abstract/document/7793011}, doi = {10.1109/IECON.2016.7793011}, }
2015
- IFAC BMSA Comparison of Classification Algorithms for Brain Computer Interface in Drug Craving TreatmentM. Mazzoleni, and F. PrevidiIFAC-PapersOnLine, 20159th IFAC Symposium on Biological and Medical Systems BMS 2015
In this paper, the use of Brain Computer Interfaces (BCIs) is proposed as a means to recover patients from craving diseases, with the aim of a clinical protocol. In order to understanding the BCI messages, a classification algorithm based on logistic regression has been developed. The choice was dictated by a comparison with other known classification techniques of different reasoning type, highlighting the pros and cons of them. Finally, a result regarding the brain areas which are more involved during the activity is reported.
@article{MAZZOLENI2015487, title = {A Comparison of Classification Algorithms for Brain Computer Interface in Drug Craving Treatment}, journal = {IFAC-PapersOnLine}, volume = {48}, number = {20}, pages = {487-492}, year = {2015}, note = {9th IFAC Symposium on Biological and Medical Systems BMS 2015}, issn = {2405-8963}, doi = {10.1016/j.ifacol.2015.10.188}, url = {https://www.sciencedirect.com/science/article/pii/S2405896315020790}, author = {Mazzoleni, M. and Previdi, F.}, keywords = {Brain Computer Interface, Neural activity Machine Learning, Pattern Recognition}, }
2014
- IEEE CDCFault Detection via modified Principal Direction Divisive Partitioning and application to aerospace electro-mechanical actuatorsMirko Mazzoleni, Simone Formentin, Fabio Previdi, and 1 more authorIn 53rd IEEE Conference on Decision and Control, 2014
In this paper, the use of the Principal Direction Divisive Partitioning (PDDP) method for unsupervised learning is discussed and analyzed with a focus on fault detection applications. Specifically, a geometric limit of the standard algorithm is highlighted by means of a simulation example and a modified version of PDDP is introduced. Such a method is shown to correctly perform data clustering also when the standard algorithm fails. The modified strategy is based on the use of a Chi-squared statistical test and offers more guarantees in terms of detection of a wrong functioning of the system. The proposed algorithm is finally experimentally tested on a fault detection application for aerospace electro-mechanical actuators, for which a comparison with k-means and fuzzy k-means approaches is also provided.
@inproceedings{7040292, author = {Mazzoleni, Mirko and Formentin, Simone and Previdi, Fabio and Savaresi, Sergio M.}, booktitle = {53rd IEEE Conference on Decision and Control}, title = {Fault Detection via modified Principal Direction Divisive Partitioning and application to aerospace electro-mechanical actuators}, year = {2014}, volume = {}, number = {}, pages = {5770-5775}, keywords = {Clustering algorithms;Standards;Actuators;Matrix decomposition;Partitioning algorithms;Vectors;Frequency-domain analysis}, doi = {10.1109/CDC.2014.7040292}, url = {https://ieeexplore.ieee.org/abstract/document/7040292}, }