一种基于T-S云模型的非线性系统控制  

T-S cloud model for nonlinear systems control

作　　者：黄景春[1,2] 肖建[1] 周聪[1] 

机构地区：[1]西南交通大学电气工程学院,四川成都610031 [2]兰州交通大学机电与动力工程学院,甘肃兰州730070

出　　处：《电机与控制学报》2009年第A01期149-151,156,共4页Electric Machines and Control

基　　金：国家自然科学基金(60674057);教育部博士点基金(20060613003);四川省应用基础研究基金(05JY029-006-4)

摘　　要：将云模型与T-S模糊系统结合,利用隶属云代替模糊系统的前件,提出一种T-S云模型。T-S云模型综合考虑模型的精确性和可解释性,不但可以利用专家的知识和经验建立系统模型,而且还可以通过输入/输出数据设计云模型系统。详细分析T-S云模型的系统结构。基于云模型和模糊系统对模糊概念表述的一致性,在不考虑超熵的情况下,使用梯度下降法辨识T-S云模型前件参数。利用递推最小二乘法辨识T-S云模型后件参数。设计了基于T-S云模型的控制器,实现了将卡车后倒至指定的卸车位置,体现了T-S云模型的不确定处理能力。仿真研究验证了算法的有效性。The integration of cloud model and T - S fuzzy system was considered. The antecedents of the fuzzy system Were replaced by the membership clouds. And a T - S cloud model was proposed. The T - S cloud model balances the trade-off between accuracy and interpretability. The system model can be designed not only through the expert knowledge and experience but also the input/output data. The structure of cloud model system was analyzed. The concept of fuzzy is consistent in cloud model and fuzzy systems. Therefore without considering the hyper-entropy, an iterative gradient-descent algorithm was used to update the antecedent parameters of the T - S cloud model. Then the consequent parameters were calculated by recursive least square method. A controller based on the T -S cloud model was developed to back up the truck. Simulation results show the effectiveness of the proposed method.

关 键 词：云模型 T—S模糊系统 非线性系统控制 梯度下降法 最小二乘法 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]