变结构神经网络控制在SRD中的应用

The Application of a Variable Structure Neural Network Control in SRD System

目的为了减小开关磁阻电机驱动系统参数变化对系统性能产生的影响,解决变结构控制的不连续控制特性问题.方法在常规变结构控制的基础上,引入神经网络控制技术,将普通滑模变结构与神经网络相结合,通过变结构神经网络控制策略进行控制器设计.结果经过对新的滑模变结构控制与传统的PID控制和普通变结构控制性能的对比分析表明,新的滑模变结构控制较好地克服了负载扰动及参数变化对系统转速的影响.结论新的控制方法快速性好、鲁棒性强,减少了开关磁阻电机驱动控制系统的输出抖动.

In order to improve the performances of Switched Reluctance Motor Drive system and to solve the characteristic of discontinuity of variable structure control, a new control scheme based on sliding variable structure and network control is presented. It is applied in switched reluctance motor to control drive speed. The experimental comparisons are made between a common sliding mode variable structure controller and a classical PID controller for stability and rapidity. The new method can weaken disturbances under varying load and parameter to some extent. The simulation results show that the new controller is effective in making the drive insensitive to parameter variations and disturbances.