基于模糊动态罚函数遗传算法的桁架结构拓扑优化  

Truss-structure topology optimization based on genetic algorithm with dynamic fuzzy penalty function

作　　者：郭丽华[1] 李华[1] 曹自洋[1] 杨建锋[1] 

机构地区：[1]苏州科技学院机械工程学院,江苏苏州215009

出　　处：《机械设计》2015年第2期61-65,共5页Journal of Machine Design

基　　金：江苏省高校自然科学研究面上资助项目(11KJB460009);江苏省自然科学基金资助项目(BK2011314)

摘　　要：针对罚函数遗传算法中惩罚系数难以确定的问题,建立了基于模糊数学的动态惩罚系数推理系统。首先根据惩罚系数对选择压力的影响,确定了惩罚系数的调整原则,定义了违反约束个体的比例和违反程度的表达式,并设计了一种惩罚系数的Mamdani型模糊控制器。这种模糊控制器可以根据每代群体中违反约束个体的比例和违反程度自动确定惩罚系数。最后,将此系统应用到改进遗传算法中,应用桁架拓扑优化的经典数值算例验证了算法的可行性。The penalty-coefficient-inference system based on fuzzy mathematics was proposed in order to decide the coefficients that was difficult to get in reality. Firstly, the adjustment principle of penalty coefficients was determined according to effects of penalty coefficients on selective pressure. And the expressions were defined which included the constraint-individual violation proportion and the violation degree. Then, a kind of fuzzy controller based on Mamdani was designed, determining the penalty coefficient automatically according to the constraint-individual violation proportion and the violation degree. Finally, the fuzzy controller is applied to improve genetic algorithm which is verified to be feasible by truss-structure topology optimization.

关 键 词：遗传算法 动态罚函数 Mamdani型模糊控制器 桁架结构拓扑优化 

分 类 号：TH12[机械工程—机械设计及理论]