On-line fuzzy identification of thermal systems based on an improved T-S model

In view of fuzzy modeling of complicate nonlinear systems, an on-line identification algorithm based on an improved T-S model is presented. Differential equation structure of the model is obtained first, then the fuzzy cluster center is corrected on-line according to the close degree of the input sample and the cluster center, and the cluster radius is refreshed in real-time according to the distance between the input sample and the cluster center. Finally, consequent parameters of the model are identified by the recursive least squares algorithm. Compared with previous identification algorithms, the on-line identification algorithm presented in this paper requires less fuzzy rules, has higher identification precision, and is simple and easy to implement. Practicability and effectiveness of the method are proved by the simulation results of the Box-Jenkins data and the boiler overheated steam temperature system. It provides a method to identify model parameters on-line for many new control strategies, such as fuzzy predictive control, adaptive control, etc.