New control method for high performance drive system of PM synchronous motor

Abstract

Permanent magnet (PM) synchronous motor has been gradually replacing DC motors in a wide range of drive applications such as machine tools and industrial robots. The advantage of using a PM synchronous motor is that many drawbacks caused by the brushes and commutators of a DC motor can be eliminated. Furthermore, the PM synchronous motor has the high power density, large torque to inertia ratio, and high efficiency as compared with a DC motor having the same output rating. The PM synchronous motor, however, has the nonlinear characteristics and inherent coupling problem. Therefore, to directly control the developed torque, the field-oriented control is usually employed. With the recent advances in general purpose microprocessors and digital signal processors (DSPs), the field-oriented control has made the PM synchronous motor drive possible for the high performance applications where traditionally only the DC motor drives were applied. However, there are some limitations to solve the disturbance and parameter variation effects using conventional controller especially in high resolution position control. One effort was done by compensating the load torque with a torque estimator. But, it is costly and includes noise effect caused by speed sensor. So a control approach using neural network for the robust position control of a PM synchronous motor is presented. The linear quadratic controller plus feedforward neural network is employed to obtain the robust PMSM drive system. The neural network is trained in on-line phases and this neural network is composed of a feedforward recall and an error back-propagation training. Since the total number of nodes is only eight, this system can be realized by the general microprocessor. During the normal operation, the input-output response is sampled and the weighting value is trained by error back-propagation at each sampling period to accommodate the possible variations in the parameters or load torque. And the state space...