The Predictive Diagnosis Method of Electric Drive State Via an Artificial Neural Network

This paper proposes a method of diagnosing the state of electric drive and/or its elements via an artificial neural network in real-time. The proposed method used pre-installed sensors by manufacturers. Additional sensors or devices are not needed to implement this method. The paper considers the relevance of this topic, describes the advantages of the method, explains the algorithm of work. The necessity of using the artificial neural network (ANN) is justified. The obtained ANN trained on data taken from real objects. The data of the training sample are trends values get from sensors during commissioning. Theoretical calculations are confirmed by experimental data. Special attention is paid to the scope proposed method.