New decentralized MRAC algorithms for large-scale uncertain dynamic systems

Abstract

In this paper, we present the first complete solution to the decentralized model reference adaptive control (MRAC) problem for general interconnected uncertain dynamic systems with arbitrarily large interconnection strengths. Each local subsystem may be perturbed by significant and unmodeled dynamics as well as contaminated by bounded output disturbances. Despite these, excellent properties of the closed-loop controlled systems can be proven, which are hardly achieved by conventional decentralized MRAC. The tracking errors can be made as small as possible by simply decreasing the design time constant /spl tau/. It should also be noted that the presented algorithms form the decentralized counterpart of our earlier work (1992, 1993), where the variable structure design (VSD) for the MRAC of general uncertain plants is investigated. A simple example is also provided to explain the advantage of reducing the controller complexity when the proposed decentralized MRAC design is incorporated.