Non‐linear fuzzy predictive control of hydroelectric system

Abstract

This study brings out a brand new non-linear predictive control method, a fuzzy generalised predictive control, for hydroelectric plant systems. Takagi-Sugeno (T-S) fuzzy method and generalised predictive control are implemented in the non-linear control method. The basic concept of the method is generalised predictive control, which is a linear system based method. However, the authors novelty adopt fuzzy model, especially the T-S model, as the prediction model to broaden the applications of the predictive control to non-linear range. Therefore, the proposed method inherits the advantages of generalised predictive control, but it can also effectively control a complex non-linear system. Its effectiveness and efficiency are verified by Lyapunov stability theory. Applying the control method on the non-linear hydroelectric system, the different experiments show that this control method succeeds in maintaining the hydroelectric system stable. Those experimental results accord with theoretic analysis. Thus, the proposed control method enjoys the capacity to govern a non-linear system. A comparison experiment between the proposed method and PID method illustrates the advantages of the authors' method, which are its smooth treatment, less overshoot and fast to be stable of response. In addition, discussions about the tuning control parameters facilitate the usages of the proposed method.