模糊神经网络要点分析.pptVIP

模糊神经网络 输入/输出变量隶属度函数曲线 模糊神经网络 用遗传算法训练网络，将网络可变阈值Wb、可变权值Wc、We、Wf 作为基因组成染色体。群体中第i条染色体的适应度为: S(i)=1/E(i) 其中 初始染色体数：200 每次遗传操作选取100对染色体进行交叉变异操作 变异概率为0.1，变异步长0.01 模糊神经网络 天气预报输出结果 模糊神经网络 如： rule1：if X1 is small and X2 is small then Y is H1 …1.0,0.66,0.48,0.50 rule45：if X1 is big and X2 is big then Y is H5…1.02,0.84,1.05,1.27 按此方法，可提取出重要性提高的9条规则作为实际应用的参考如下所示： rule1：if X1 is small and X2 is small then Y is H3 with 1.49 rule2：if X1 is small and X2 is mid then Y is H4 with 1.69 rule3：if X1 is small and X2 is bigs then Y is H5 with 2.09 rule4：if X1 is mid and X2 is small then Y is H3 w