Asymptotically Stable Multiple Input-Multiple Output Direct Model Reference Adaptive Controllers for Processes Not Necessarily Satisfying a Positive Real Constraint

提出了一种改进，对于非常大的一类过程，应该保证渐近输出模型遵循一个非常容易实现的直接自适应控制算法，该算法可以应用于多输入多输出系统。该算法为了满足正实约束，不得不加入补充前馈动力学，从而导致稳定但不完美的渐近跟踪。通过将部分前馈动力学纳入模型的新修改使得对增强对象和模型输出的完美跟踪意味着对原始输出的完美跟踪。提出的研究包括新改进算法的稳定性证明，设计鲁棒前馈动力学的程序研究和实验验证研究。

A modification is proposed that should, for a very large class of processes, guarantee asymptotic output model following for a very easily implementable direct adaptive control algorithm that can be applied to multiple input multiple output systems. This algorithm in order to satisfy a positive real constraint, has had to incorporate supplementary feedforward dynamics, thus resulting in stability but not perfect asymptotic tracking. The new modification by incorporating part of the feedforward dynamics into the model is such that perfect tracking of the augmented plant and model outputs implies perfect tracking of the original outputs. The proposed research involves the stability proof of the new modified algorithm, a study of procedures for designing robust feedforward dynamics, and an experimental validation study.