在更加一般的条件下对黄金分割反馈控制器在单输入单输出和多输入多输出特征模型中的稳定性和鲁棒性进行了系统分析,并在混合辨识背景下研究了投影梯度法在特征模型框架中应用时的输入状态稳定(ISS)性和参数估计误差的L2界.最终的结果进一步扩展了已有的结论,对整个系统的稳定性分析具有重要作用.
Under more general conditions, the stability and the robustness of golden-section feedback control law are analyzed in this paper, and in the framework of hybrid identification, the input-state stable (ISS) property of the gradient project method used in SISO and MIMO characteristic models is studied, and the L2 bound of estimated errors of characteristic parameters is obtained. The final results further gen-eralize the former results and play very important roles in the stability analysis of entire hybrid system.