基于TS模糊模型的热工过程建模方法  

Fuzzy modeling Based on TS model and Its Application for Thermal Process

作　　者：郭颖[1] 吕剑虹[1] 吴波[1] 向文国[1] 

机构地区：[1]东南大学能源与环境学院,南京210096

出　　处：《系统仿真学报》2010年第1期210-215,240,共7页Journal of System Simulation

基　　金：国家863高技术基金(2006AA05A107;2008AA041407);2007年度江苏省科技成果转化专项资金资助(BA2007008)

摘　　要：热工过程对象通常具有复杂的动态特性和非线性、强耦合、不确定性等特征,从而使得常规建模方法难以取得满意的效果,因此提出一种改进型TS模糊模型在线建模方法。此建模方法基于以下思想,首先提出基于中心粒群算法的截集模糊C-均值聚类算法并对TS模糊模型进行模型结构离线辨识,确定模型的结构和前件部分参数的初始值;然后应用解耦扩展卡尔曼滤波算法进行后件参数在线辨识,同时对前件辨识结果进行精确修正。最后,对煤气炉、500MW机组及气化炉等热工过程进行仿真计算表明本方法具有精度高、计算量小等优点。Considering the complex dynamic behaviors, nonlinearities, strong coupling, and time-varying characteristics of the thermal process, a modified online modeling method based on TS fuzzy model was proposed and applied. The modeling method is on the base of the following ideas： firstly, the sectional set fuzzy c-means clustering based on center particle swarm optimization was proposed to identify the model structure off line and the initial values of the antecedent parameters. Secondly, the decoupled extended Kalman filter was used to identify the consequent parameters of the fuzzy rules on line and modify the antecedent parameters exactitudely. Finally, advantages of this method such as high precision and less calculation were proved by simulation results on the gas stove, 500MW power plant, and the gasifier.

关 键 词：TS模糊模型 截集模糊C-均值聚类 解耦扩展卡尔曼滤波 热工过程 辨识建模 

分 类 号：TK3[动力工程及工程热物理—热能工程]