Fuzzy Modeling, Fuzzy Control And Their Application In Automatic Control Of Power Plant Thermal Process

Fuzzy system modeling, fuzzy controller analyzing and design, and their application are the research focuses as well as the future developing direction of fuzzy control theory. The developing process and the present situation of fuzzy control theory and its application in power plant thermal process control are analyzed carefully and the key issues are studied precisely and deeply in this paper. The main points are as follows:1.Research on offline identification of T-S model. In accordance with the problem that sensitivity to initialization and noise, and some relative parameters must be determined beforehand during the fuzzy clustering process in the usual fuzzy cluster algorithm, and the existing competitive clustering algorithm have poor convergence properties, and make convergence to a local minimum more likely. A type of adaptive competitive cluster algorithm for structure identification is presented. The advantage of this kind of algorithm is that the fuzzy rules and initial value of relative parameters need not be preset and they can be identified only based on the input and output test data of the identified system. At the same time, orthogonal least squares (OLS) method algorithm is proposed to remove redundant or less important clusters during the clustering process in order to extract fuzzy rules that capture the important features of the systems. This kind of model offers a high semantic level, a good generalization capability and high precision.2.Research on the online identification of T-S model. In practical process, because of random disturbance, time-variant or uncertainty presentting, in many case, it is difficult to get effective model by offline modeling. Thereby, the online identification method was researched. At first, Measurement noise and outliers are removed from sample datum by data filtering. The new potential function is proposed to identify model structure and parameters, which in next step are fine tuned using the well-known gradient-descend algorithm. The influence of fuzzy rule is defined, according to that fuzzy rule can be update and optimal online. The consequent parameters of the T-S model are identified and optimized by extended kalman filter. The approach has been successfully applied to T-S...