论文部分内容阅读
研究随机线性反馈控制系统的结构辨识问题。在已知时滞的下界和模型阶的上界的假定下，通过使修改的Ｂａｙｅｓｉａｎ信息准则最小化，推导出由多输入多输出ＣＡＮ模型描述的系统的未知阶与时滞的估计算法，证明了算法是强一致收敛的，且能在有限步内达到其模型结构参数的真值。讨论了当模型的参数矩阵不满秩时减弱条件Ｈ’ｓ的强一致估计算法。
Research on structural identification of stochastic linear feedback control system. Under the assumption of the lower bound of the delay and the upper bound of the model order, the unknown order and delay estimation algorithm of the system described by the multiple-input-multiple-output CAN model is deduced by minimizing the modified Bayesian information criterion, The algorithm converges strongly and can reach the true value of its structural parameters in a finite step. We discuss a strong consensus estimation algorithm for H’s weakening condition when the model’s parameter matrix is not satisfied with rank.