Smooth speed control of electric motor drive based on neural network method

Abstract

This project aims to regulate the speed of permanent magnet synchronous motor (PMSM) by using the Neural Network. PMSM has been increasingly employed these days. However, there are large amounts of disturbances and uncertainties existing in PMSM speed regulation system, such as parameters variation, inverter nonlinearity, cogging torque and so on. Most of these factors may cause torque ripple, which is one of the most critical issues in PMSM system. This problem is expected to be solved by Neural Network because of its strong nonlinear fitting ability and self-learning ability. Hence, the disturbances can be obtained accurately and suppressed. This project allows student to develop an advanced electric motor control strategy from analysis, simulation, and evaluation.