Machine Learning for Sensorless Temperature Estimation of a BLDC Motor
In this article, the authors propose two models for BLDC motor winding temperature estimation using machine learning methods. For the purposes of the research, measurements were made for over 160 h of motor operation, and then, they were preprocessed. The algorithms of linear regression, ElasticNet,...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-07-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/14/4655 |
| _version_ | 1797526069952118784 |
|---|---|
| author | Dariusz Czerwinski Jakub Gęca Krzysztof Kolano |
| author_facet | Dariusz Czerwinski Jakub Gęca Krzysztof Kolano |
| author_sort | Dariusz Czerwinski |
| collection | DOAJ |
| description | In this article, the authors propose two models for BLDC motor winding temperature estimation using machine learning methods. For the purposes of the research, measurements were made for over 160 h of motor operation, and then, they were preprocessed. The algorithms of linear regression, ElasticNet, stochastic gradient descent regressor, support vector machines, decision trees, and AdaBoost were used for predictive modeling. The ability of the models to generalize was achieved by hyperparameter tuning with the use of cross-validation. The conducted research led to promising results of the winding temperature estimation accuracy. In the case of sensorless temperature prediction (model 1), the mean absolute percentage error MAPE was below 4.5% and the coefficient of determination R<sup>2</sup> was above 0.909. In addition, the extension of the model with the temperature measurement on the casing (model 2) allowed reducing the error value to about 1% and increasing R<sup>2</sup> to 0.990. The results obtained for the first proposed model show that the overheating protection of the motor can be ensured without direct temperature measurement. In addition, the introduction of a simple casing temperature measurement system allows for an estimation with accuracy suitable for compensating the motor output torque changes related to temperature. |
| first_indexed | 2024-03-10T09:24:56Z |
| format | Article |
| id | doaj.art-a3c06bec988e4c838b7358e3498f23ef |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T09:24:56Z |
| publishDate | 2021-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-a3c06bec988e4c838b7358e3498f23ef2023-11-22T04:54:06ZengMDPI AGSensors1424-82202021-07-012114465510.3390/s21144655Machine Learning for Sensorless Temperature Estimation of a BLDC MotorDariusz Czerwinski0Jakub Gęca1Krzysztof Kolano2Department of Computer Science, Lublin University of Technology, 20-618 Lublin, PolandDoctoral School, Lublin University of Technology, 20-618 Lublin, PolandDepartment of Electrical Drives and Machines, Lublin University of Technology, 20-618 Lublin, PolandIn this article, the authors propose two models for BLDC motor winding temperature estimation using machine learning methods. For the purposes of the research, measurements were made for over 160 h of motor operation, and then, they were preprocessed. The algorithms of linear regression, ElasticNet, stochastic gradient descent regressor, support vector machines, decision trees, and AdaBoost were used for predictive modeling. The ability of the models to generalize was achieved by hyperparameter tuning with the use of cross-validation. The conducted research led to promising results of the winding temperature estimation accuracy. In the case of sensorless temperature prediction (model 1), the mean absolute percentage error MAPE was below 4.5% and the coefficient of determination R<sup>2</sup> was above 0.909. In addition, the extension of the model with the temperature measurement on the casing (model 2) allowed reducing the error value to about 1% and increasing R<sup>2</sup> to 0.990. The results obtained for the first proposed model show that the overheating protection of the motor can be ensured without direct temperature measurement. In addition, the introduction of a simple casing temperature measurement system allows for an estimation with accuracy suitable for compensating the motor output torque changes related to temperature.https://www.mdpi.com/1424-8220/21/14/4655temperature estimationmachine learningBLDCelectric machine protection |
| spellingShingle | Dariusz Czerwinski Jakub Gęca Krzysztof Kolano Machine Learning for Sensorless Temperature Estimation of a BLDC Motor Sensors temperature estimation machine learning BLDC electric machine protection |
| title | Machine Learning for Sensorless Temperature Estimation of a BLDC Motor |
| title_full | Machine Learning for Sensorless Temperature Estimation of a BLDC Motor |
| title_fullStr | Machine Learning for Sensorless Temperature Estimation of a BLDC Motor |
| title_full_unstemmed | Machine Learning for Sensorless Temperature Estimation of a BLDC Motor |
| title_short | Machine Learning for Sensorless Temperature Estimation of a BLDC Motor |
| title_sort | machine learning for sensorless temperature estimation of a bldc motor |
| topic | temperature estimation machine learning BLDC electric machine protection |
| url | https://www.mdpi.com/1424-8220/21/14/4655 |
| work_keys_str_mv | AT dariuszczerwinski machinelearningforsensorlesstemperatureestimationofabldcmotor AT jakubgeca machinelearningforsensorlesstemperatureestimationofabldcmotor AT krzysztofkolano machinelearningforsensorlesstemperatureestimationofabldcmotor |