提出基于递归式神经网络的控制方法,对永磁同步电动机位置系统进行控制。控制器是由递归式神经网络控制器和补偿控制器组成,其中递归式神经网络控制器主要是为了跟踪输入信号,而补偿控制器是用来补偿最优控制法则和递归式神经网络控制器之间的误差。仿真结果表明,采用该种方法对位置系统进行控制,具有较好的跟踪性和鲁棒性。
Control based on diagonal recurrent neural network is proposed to control position system of the AC permanent-magnet synchronous motor,the diagonal recurrent neural network controller and the compensation controller make up the control system,the diagonal recurrent neural network is used to track input,compensation controller is used to compensate approximation error between the optimal control law and the diagonal recurrent neural network controller.The effectiveness of the proposed control schemes is demonstrated by the simulation results.