基因遺傳演算法應用在模糊模型的建立之研究

Abstract

這篇論文主要研究如何運用基因遺傳演算法 (genetic algorithm)，去求 得一模糊模型 (fuzzy model) 的隸屬函數 (membership function)，再 經由一產生模糊法則 (fuzzy rule) 的方法，模糊模型便可以很容易地被 建立，一般來講，直接利用基因遺傳演算法去求隸屬函數並不是很有效率 。因此，我們提出邊界限制 (boundary constraint) 的觀念，遵循此一 概念，模糊模型的隸屬函數可以有效地被求得。模擬結果驗証了我們所提 出的方法。 This paper mainly investigates how to find a fuzzy model by employing genetic algorithms. With a method for generating fuzzy rules, a fuzzy model could be easily constructed. Generally speaking, it is not very efficient to seek membership functions directly by way of genetic algorithms. Thus, we propose a concept of boundary constraints. Following this concept, the membership functions of a fuzzy model can be effectively obtained. And the simulaton results validate the proposed method.