隐性知识条件下变论域模糊控制器的设计

Design of Variable Universe Fuzzy Controller on Tacit Knowledge

  • 摘要: 阐述了变论域模糊控制器的基本结构,给出了在隐性知识条件下的设计方法,即把推理前件的隶属度函数峰点在论域内作平均分配,把推理后件的隶属度函数峰点视作待定参数,以递推最小二乘辨识算法作为其实现算法,并讨论了递推算法的截止条件.仿真实验表明,由该辨识算法设计的控制器能够较好地逼近原始的输入输出数据.

     

    Abstract: The basic structure of variable universe fuzzy controller is expatiated firstly.Under the condition of tacit information,a method is presented,i.e.,the peak-value-points of membership functions on the antecedent parts of inference are placed equidistantly,and the similar points on the consequent parts of inference are regarded as undetermined parameters.The recursive least square algorithm serves as a fulfillment algorithm,and the ending condition of this algorithm is discussed.Simulation experiment shows that a controller designed by the algorithm can approximate original input/output data with fine effect.

     

/

返回文章
返回