Development of a New System for the Asynchronous Motor Parameters Identification based on Neural Networks

The article deals with the problems of identifying the asynchronous motor (AM) parameters. AM Parameters Identification Identification is an important aspect for setting up and synthesizing control systems, also parameter identification algorithms are complex mathematical models protected as an object of intellectual property. The purpose of the article is to modernize the existing methodology for asynchronous motor testing, regulated by state standard (GOST), through the introduction of neural networks for further parameters’ identification based on them. The practical value lies directly in the neural networks that identify the AM parameters. Theoretical studies were carried out using the Matlab Simulink software, data processing was carried out in the Matlab and MS Excel. The article provides a rationale for choosing the structure of neural networks, algorithms for its training, and also experimentally selects the hidden layer optimal configuration for each of the experiments. Neural networks were trained on motors with a power of 10-100 kW with a rated voltage of 380 V, a rated frequency of 50 Hz, and a synchronous rotation speed of 1500 rpm. The developed neural networks can be used to identify motor parameters with similar passport data. It is also possible to train new neural networks for motors with other capacities and passport data, following the article methods.