小波网络在直接转矩控制系统低速运行中的应用

针对传统的异步电动机直接转矩控制系统低速运行时存在较大转矩脉动的问题,详细分析了定子电阻变化对系统控制性能的影响,提出了基于小波网络的定子电阻辨识方法.将定子电流的误差和定子电流误差的变化量作为小波网络的输入,网络输出为定子电阻误差的动态估计值;综合应用递推正交最小二乘法与改进的Givens变换训练小波网络参数,利用小波网络良好的时频局部特性可以准确的观测出定子磁链和转矩,优化了逆变器的控制策略.仿真结果对比表明该方法可以有效得改善电动机的低速运行性能,优于采用BP(Backward Propagation)神经网络的方法. 

Application of wavelet network for induction motor in direct torque control system operating at low speed

To improve the low-speed dynamic performance of induction motor DTC(direct torque control),a novel method of stator resistance identification based on WN(wavelet network) was presented.The inputs of the WN were the current error and the change in the current error and the output of the WN was the stator resistance error.The synthesized method of ROLSA(recursive orthogonal least squares algorithm) and improved Givens transform was used to fulfill the network structure and parameter identification.By the use of wavelet transform that accurately localized the characteristics of a signal both in the time and frequency domains,the occurring instants of the stator resistance change could be identified by the multi-scale representation of the signal.Once the instants were detected,the accurate stator flux vector and electromagnetic torque were acquired by the parameter estimator,which made the DTC applicable in the low region,and the inverter control strategy was optimized.The simulation results show that the proposed method can efficiently reduce the torque ripple and current ripple,and is superior to the BP(backward propagation) neural network.