Research on Neural Network Vector Control System for Induction Motor

Abstract

The purpose of paper is deals with development of neural network control system for induction motor. Such system could be in use to control motor with possible unpredicted behavior, which could be caused due to appearance of unexpected faults during its operation or other factors, when it is essential to keep operational process without stoppage, while prevent motor further damage development and its eventual failure. Proposed control systems are built on the basis of the artificial neural network and provides a list of possibilities for creating control systems that are resistant to the uncertainty of the object and the control process. It was chosen field-oriented control algorithm as a basis for neural controller due to its advantages that can be improved in future. The basic equations and elements of the indirect field oriented control scheme are given. The control scheme is developed basing on feedforward neural network which have two separate input layers: one divided hidden layer and one general hidden layer. Therefore, it is worth exploring such a combination of control methods and its effect on induction motor operation.