MIMO自校正控制器的全局收敛算法

A GLOBAL CONVERGENT ALGORITHM OF MIMO SELF-TUNING CONTROLLER

中文摘要

      本文为随机多变量线性系统建立了全局收敛的MIMO自校正控制器算法。使用这种随机适应控制算法，不论系统和算法的初始条件如何，总能保证系统参数未知条件下的适应控制渐近地达到系统参数已知条件下的最优控制效果。文献[1]、[2]的绝大部分结果和[3]中的随机结果大体上可视为本文结果的特殊情况。使用这种算法，不用在线解方程组，避免了麻烦的矩阵求逆运算。

英文摘要

      This paper established a global convergent algorithm of MIMO self-tuning controller for stochastic multivariable linear systems. Whatever the initial conditions of the systems and the initial values of the algorithm may be, this stochastic adaptive control algorithm ensures the adaptive control with the system parameters unknown to achieve asymptotically the optimal control with the system parameters known. The major results in [1] and [2] and the stochastic result in [3] may be regarded roughly as special cases of the results of this paper. This algorithm does not need an on-line solution of linear simultaneous equations and avoids the evaluation matrix inversion as well.