నైరూప్య

Designing and Simulation of Dual Loop Speed Controller For Efficient Speed Control Of Permanent Magnet Synchronous Machine Using Neural Network

Ritu Tiwari, Manoj Nigam

In the field of power electronics and industrial instrumentation electrical drives plays a very crucial role. In recent years, rapid growth in manufacturing industries has been recorded. Such fast growth requires equipments of high accuracy and efficiency. There are a variety of ac servo drives on the market with both the brush machine and others ac servo drives. Two types of permanent-magnet ac motor drives are available in the drives industry. These are the permanent magnet synchronous motor (PMSM) drive with a sinusoidal flux distribution, and the brushless dc motor (BDCM) drive with a trapezoidal flux distribution. This paper deals with the designing and simulation of Neural Network (NN) based speed controller for efficient speed control of PMSM drives. The basic idea came from inability of PI controllers observed during practical situations. It has been observed that PI controllers are widely used for speed control of PMSM drives, although a PI controller provides a good speed control once motor reaches in the steady state. However it shows good performance in steady state but highly suffered from speed ir-regulation problem in transition state. To overcome this problem this paper proposed a dual loop controller, in which inner loop is designed for regulation of motor stator current using Pulse Width Modulation (PWM) technique and a neural network based controller is proposed in the outer loop to provide efficient speed regulation to replace conventional PI controller. The results obtained after designing and simulation of developed technique shows the high efficiency of speed control of NN based controller.