有模糊目标的可靠性设计一个混合遗传算法

A Hybrid Genetic Algorithm of Reliability Design with Fuzzy Objectives

给出了求解带有模糊目标的可靠性设计问题的一个混合遗传算法.该方法利用模糊集隶属函数的特性将各子目标函数模糊化处理为能够反映各子目标相对重要程度、具有加权指数的协调满意度函数形式;通过加权函数再将各子目标的协调满意度函数综合成统一的单目标优化模型,然后用遗传算法求解模型,用Hooke-Jeeves方法改进遗传算法的群体,得到一个混合遗传算法,最后给出实例验证该方法的有效性.

A hybrid genetic algorithm of reliability design with fuzzy objectives was given. In this method where the property of membership function in fuzzy sets theory is utilized,each sub-objective function is transformed into coordinated satisfactory function-weighted exponent ; the weighted exponent shows the relative importance of each sub-objective function. The coordinated satisfactory functions are synthesized into a single-objective Optimization model. The genetic algorithm is used for its solution. The group of genetic algorithm is improved by using Hooke- Jeeves method. Finally,an example is given to illustrate the validity of the method.