A novel design of PID controller for multivariable control system

A single-neural identification-free adaptive intelligent controller based on grey prediction is developed for multi-input and multi-output(MIMO) system with time-delay. The basic method is that the MIMO system is decomposed into a set of multi-input-single-output subsystems. The decoupling problem of MIMO system is solved by a set of grey model GMC(1,n), which is used to approximate the MIMO system dynamic behavior. The GMC(1,n)model, obtained by integrating the convolution technology in the multi-variable-first-order grey model GM(1,n) model to establish the exact solution of grey model, greatly enhances the accuracy of prediction. In addition, each subsystem is controlled by PID controller, the parameters of which are regulated by using an identification-free adaptive intelligent algorithm. The intelligent algorithm uses performance criteria based on geometrical properties of the control error process, and needs no accurate mathematical model or no identification model of the controlled system. The simulation results show that the proposed controller minimize remarkably the interaction between every input and every output, and has fast transient response and strong robustness.