A New Framework For Adaptive Control Adaptive Robust Control Approach

摘要

As a powerful technology of dealing with parametric uncertainty in feedback system design, adaptive control has been studied extensively since 1950's. The main stream of approaches is transfer function based and the fundamental scheme in the proof of stability is Lyapunov stability/hyperstability theories. This classic approach has achieved many fruitful results and has a good record of industrial applications. However, this classic approach also embodies some inherent limitations such as design complexity stemming from high relative degree and the minimal phase condition coming from model-matching. Meanwhile, the state space approach was explored since late 1970's as is represented by the research of Kreisselmier. In his landmark paper, Kreis-sekmier proposed a smart structure of full-order observer for SISO uncertain parametric systems and showed that the state of uncertain systems could be reconstructed by a set of filters and an adaptive law for uncertain parameters. This observer and the related state regulation problem were investigated in a series of papers. Unfortunately, only local results on state regulation problem were obtained except for that the PE requirement is put on the system input.