GA优化的感应电机无速度传感器控制

Speed-sensorless control for induction motor based on GA optimization

针对感应电机无速度传感器直接转矩控制系统低频脉动问题,在中、低频段分别采用两种不同的模型参考自适应辨识方法,即在中频段以电压、电流两种磁链辨识模型为参考模型和可调模型;在低频段则以电压、电流两种反电势辨识模型代替,以消除低频段的积分累积误差;同时充分利用低频段周期长的特点,将遗传算法引入系统.动态优化自适应律不仅很好地解决了模型间的无扰切换问题,而且降低了电机参数非线性对转速辨识精度的影响,提高了系统的鲁棒性.仿真结果证明,该方法降低了低频脉动,有效地扩大了系统调速范围.

A novel method is presented to improve low-frequency ripple for speed-sensorless direct torque control（DTC） of induction motor.In middle and low frequency ranges,two kinds of model reference adaptive system（MRAS） speed identification methods were employeed.Flux identification models based on voltage and current were used as reference and adjustable models in middle frequency range,while back electromotive force identification model wais adopted in low frequency range to eliminate accumulative error caused by integrator.Genetic algorithm（GA） was introduced for optimizing adaptive law dynamically,which not only can solve no-disturbance switch between the models,but also improve speed identification precision in spite of non-linear variable parameters of motor.Therefore,the robustness of the system is heightened.Simulation results show that the proposed scheme reduces low-frequency ripple and enlarges speed regulative scope.