Decentralized Adaptive Control And Application On The Power System

Most industry processes are multi-input multi-output systems. Despite the closed-loop performance advantages of multivariable controllers, decentralized controllers are widely used for its' simple structure, simplified design and good failure tolerance.Two decentralized PID controller design methods are proposed in this paper. The non-convex PID controller design method, which was proposed by Astrom, is extended into multivariable system. The novelty of this method is that interaction indices are introduced. This leads to the design method emphasizing the trade-off between the individual loop performances and interaction. Exact mathematics model is needed by this method. Another method is based on decentralized relay feedback. The definition of desired critical point is supplied. The controller parameters are determined by this point. Process model is not needed. In addition, the characteristics of multivariable systems and interaction are disused.Several simulation examples are provided. The validity of the two methods is identified by typical two-input two-out processes of chemistry and power industry.