Diagnostic Monitoring of Electric Motors Based on Artificial Intelligence Methods

Abstract

In this article, the possibilities of application of artificial intelligence technology in diagnostics of faults of electric motors are examined. The application of traditional diagnostic monitoring systems in the diagnosis of motors faces problems in determining the normal and threshold values of several diagnostic parameters that cannot be measured in working conditions in cases of lack of certain information. To partially overcome these problems, it is also possible to provide a new hybrid control by applying artificial intelligence-based diagnostic systems together with traditional methods to increase the efficiency and accuracy of diagnostic control in the working condition. In the research work, fault diagnosis schemes using neural networks, fuzzy logic, neural-fuzzy and genetic algorithms were studied, and their advantages and disadvantages compared to traditional diagnostic methods were analyzed. From the conducted studies, we conclude that the methods of artificial intelligence technology have a great potential in the diagnosis of difficult-to-detect faults.