Machine-Learning-Based Condition Monitoring of Power Electronics Modules in Modern Electric Drives

Integrating machine-learning (ML) models responsible for predicting the evolution of those directly collected or implicitly derived parameters enhances the smartness of industrial systems even further. In this article, data already residing in most modern electric drives has been used to establish a data-driven thermal model of power electronics modules. The developed method relies solely on existing information in the electric drive enabling its wide applicability. Adding more sensors to a product is a complicated task, thus it is not a desirable solution. For training and validating the thermal digital twin, a test bench has been designed specifically. Several approaches, from traditional linear models to deep neural networks, have been implemented to emanate the best ML model for estimating the case temperature of the module. Numerous evaluation metrics were then used to assess the investigated methods' performance and implementation in industrial embedded systems. The proposed solution performed satisfactorily while the powertrain underwent various static and dynamic operating profiles. The model identified a blockage in the air outlet of the drive by monitoring the deviation of the measured and estimated temperatures, thus it prevented the power module from experiencing a fatal failure.