模糊控制中的参数模糊化新方法研究

模糊控制中的参数模糊化新方法研究 A New Fuzzification Method in Fuzzy Control System 周穗华 ZHOU Sui-hua 1 2 first-author 张小兵 ZHANG Xiao-bing 1 2 皮汉文 PI Han-wen 1 2 海军工程大学兵器工程系 海军工程大学兵器工程系 Department of Weaponry Engineering, Naval University of Engineering Department of Weaponry Engineering, Naval University of Engineering 海军工程大学兵器工程系,湖北,武汉,430033 海军工程大学兵器工程系,湖北,武汉,430033 提出了一种基于参数正态分布的模糊化新方法 ,并建立了模型 .该模型首先将参数正态分布化 ,然后确定各模糊集对应的积分区间 ,最后通过积分求出参数属于各模糊集的隶属度 ,从而实现了参数的模糊化 .这种模糊化方法数学物理意义明确 ,算法统一 ,便于计算 ,可操作性强 ,解决了模糊控制系统中参数模糊化时隶属度难于确定的实际问题 .特别适用于对那些有一定正态离散性的参数 （如误差型或概率型参数 ）进行模糊化 .通过模糊化实例计算表明该方法有效、可行 . A new fuzzification method based on the normal distribution of the parameter is presented and the fuzzification calculating model is built. Such model first makes the parameter normalization distribution, then determines the corresponding integral interval of fuzzy set, last calculates the memberships relative to the fuzzy set. It can be used in fuzzy control system and an other area to calculate the membership which is difficult to confirm before, especially in the fuzzification of the parameters such as error and probability. The physical and mathematic meaning is explicit in this method. This new method and its calculation model are illustrated to be efficient and feasible. 模糊控制 模糊化 正态分布 fuzzy control fuzzification normal distribution O231 国防科技重点实验室基金资助项目 （ 0 0JS2 2 .8.1.JB110 4） 2003-03-01 2021-04-01 3