Temperature Estimation in Induction Motors using Machine Learning

-- The number of electrified powertrains is ever increasing today t owards a more sustainable future; thus, it is essential that unwanted failures are prevented, and a reliable operation is secured . Monitoring the internal temperatures of motors and keeping them under their thresholds is an important first step. Conventional modeling methods require expert knowledge and complicated mathematical appro aches . With all the data a modern electric drive collect s nowadays during the system operation, it is feasible to apply data - driven approach es for estimat ing thermal behaviors . In this paper, multiple machine - learning methods are investigated on their capa bility to approximate the temperatures of the stator winding and bearing in induction motors . The explored algorithms vary from linear to neural network s . For this reason, experimental lab data ha ve been captured from a powertrain under predetermined operating conditions. F or each approach, a hyperparameter search is then performed to find the optimal configuration. All the models are evaluated by various metrics, and i t has been found that neur al networks perform satisfactor ily even under transient c onditions. In [1], the percentage share of specific failures in induction machines has been presented .