摘要
针对传统直接转矩控制中存在电流、磁链和转矩脉动较大及速度传感器的使用降低了系统的可靠性,增加了系统的成本等问题,提出了利用遗传算法(GA)优化的BP网络电机速度辨识方法,实现了异步电机无速度传感器直接转矩控制。该方法保持了直接转矩控制固有的转矩响应快和系统鲁棒性强的优点,降低了磁链、转矩脉动,加快了系统的响应速度,并对负载的扰动具有较强的鲁棒性,有效地改善了系统的动、静态性能,实验结果证实了该方法的可行性和有效性。
There are big ripples in motor drive on current and flux linkage and. torque when using traditional direct torque control （DTC） . System with speed sensor has lower reliability and higher system cost. To solve these problems, a strategy has been put forward that using Genetic algorithm （GA） optimize BP neural network motor speed identifier, thus speed sensor-less direct torque control for induction motor is realized. Rapid torque response and strong robustness of direct torque control method are still maintained. Ripples on current and flux linkage and torque are dramatically reduced. In the mean time, the system based on BP and GA is robust to motor load disturbance. The system dynamic and static performance are dramatically improved. The experimental results show the feasibility and effectiveness of this method.
出处
《控制工程》
CSCD
北大核心
2012年第4期733-736,共4页
Control Engineering of China
基金
湖南省教育厅科研项目(09C261)
关键词
异步电机
无速度传感器控制
直接转矩控制
BP网络
遗传算法
鲁棒性
induction motor
speed sensor-less control
direct torque control
BP neural-network
genetic algorithm
Robustness
