论文部分内容阅读
针对线性未知时变系统控制器设计中离线虚拟参考反馈校正法的不足,提出了基于遗忘梯度算法的自适应虚拟参考反馈校正方法.首先,通过离线虚拟参考反馈校正法初始化控制器参数;然后,在系统运行过程中实时采集控制对象两端的输入/输出数据,构造滤波器及虚拟参考信号;最后,利用滤波数据在线辨识并实时更新控制器参数.对控制器参数收敛性进行了理论分析,并仿真验证了所提方法在控制对象特性改变情况下,具有良好的自适应性.
Aiming at the deficiency of the offline virtual reference feedback correction method in the design of linear unknown time-varying system controller, an adaptive virtual reference feedback correction method based on forgetting gradient algorithm is proposed. Firstly, controller parameters are initialized by offline virtual reference feedback correction method. In the process of system operation, the input / output data of both ends of the control object are collected in real time to construct the filter and the virtual reference signal. Finally, the controller parameters are identified and updated in real time by using the filter data. The convergence of the controller parameters is analyzed theoretically. The simulation results show that the proposed method has good adaptability under the condition of changing the characteristics of the control object.