基于神经网络PI控制器的矢量控制系统速度估算  

Speed Estimator of Sensorless Vector Control System Based on Hybrid PI and Neural Networks Controller

作　　者：沈安文[1] 朱晓琳[1] 范红勇[2] 

机构地区：[1]华中科技大学控制科学与工程系,武汉430074 [2]北京经纬纺机新技术有限公司,北京100176

出　　处：《微电机》2008年第5期9-12,共4页Micromotors

摘　　要：利用静止坐标系下电压电流模型组合磁通观测器推算转子磁通的感应电动机无速度传感器矢量控制系统,在利用推算的转子磁通q轴分量的PI控制作用进行转速估算的基础上,提出一种利用神经网络PI控制器替代PI控制器进行转速估算的方法,提高了无速度传感器矢量控制系统的动态性能。仿真和实验结果验证了该方法的有效性。Based on the speed estimation using the PI control action of q-axis estimated rotor flux, this paper presents a speed estimator using hybrid PI and neural networks controller instead of PI controller for a speed sensorless vector control system of induction motor. The system is based on the hybrid flux ob- server in rest frame which is associated with voltage model and current model. The dynamic performance of the system is improved. The effectiveness of the proposed algorithm has been verified by simulation and experimentation.

关 键 词：神经网络PI控制器 感应电动机 无速度传感器矢量控制 实验 

分 类 号：TM346[电气工程—电机]