基于遗传算法的新型模糊网络模型及其应用  

A Novel Fuzzy Net Based on Genetic Algorithm and Its Application

作　　者：俞洪[1] 俞金寿[1] 

机构地区：[1]华东理工大学自动化研究所,上海200237

出　　处：《石油化工自动化》2002年第6期30-33,共4页Automation in Petro-chemical Industry

摘　　要：提出一种基于遗传算法的新型模糊网络模型,此模糊模型采用一种新型隶属度函数,并且利用遗传算法的自然选择和遗传机制的寻优程序,进行选择、交叉、变异等遗传操作来模拟自然进化,完成了参数的寻优。在丙烯腈转化率预测应用中的实例表明,其性能较之以往的建模方法有了很大的提高。A novel fuzzy net based on genetic algorithm with a new membership function is proposed. It uses the natural selection of genetic algorithm and the optimization programming of genetic mechanism to simulate the natural evolution and to complete the optimization of parameters by the way of selection, intersection and mutation. And in the prediction of the transformation rate of acrylonitrile, it is concluded that the proposed net has a significant improvement of the performance over the other modeling methods.

关 键 词：遗传算法 模糊网络模型 隶属度函数 丙烯腈 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]