内置式永磁同步电机(interior permanent magnet synchronous motor,IPMSM)矢量伺服控制系统采用的比例积分(proportional integral,PI)调节器存在抗扰性能差、稳态精度低、跟踪参考信号难以达到理想的控制效果等问题,考虑到自抗扰控制具有参数适应性广、鲁棒性强等优点,结合神经网络强大的自学习和自适应能力,提出双神经网络自学习的改进自抗扰控制器.基于位置误差和转速误差构建目标函数,分别利用径向基函数(radial basis function,RBF)神经网络和误差反向传播(back propagation,BP)神经网络对非线性扩张状态观测器(nonlinear extended state observer,NLESO)和非线性状态误差反馈(nonlinear state error feedback,NLSEF)中的参数进行在线整定.仿真实验表明:通过RBF和BP神经网络对自抗扰控制器中关键参数进行实时整定可以找出最优控制量,该控制策略具有更好的位置、转速跟踪效果,且抗负载扰动能力和自适应能力更强.
Active disturbance rejection control of IPMSM based on double neural network self-learning
The proportional integral(PI)regulator used in the interior permanent magnet synchronous motor(IPMSM)vector servo control system has some problems,such as poor anti-interference performance,low steady-state accuracy,difficult to achieve the desired control effect of tracking reference signal,and so on.Considering the active disturbance rejection controller(ADRC)has the advantages of wide parameter adaptability and strong robustness,combined with the strong self-learning and adaptive ability of neural network,an improved ADRC based on double neural network is proposed.Based on the position error and rotational speed error,the objective function is constructed,and the parameters of nonlinear extended state observer(NLESO)and nonlinear state error feedback(NLSEF)are adjusted online by radial basis function(RBF)neural network and error back propagation(BP)neural network respectively.The simulation results show that adjusting the key parameters of ADRC in real time by RBF and BP neural network can find out the optimal control quantity.The control strategy has better position and speed tracking effect,and stronger anti-load disturbance ability and adaptive ability.
NETL
NSTL
万方数据
维普
