遗传算法中约束的凝聚选择和复合形处理方法

SURROGATE REPRODUCTION AND COMPLEX METHOD FOR HANDLING CONSTRAINTS IN OPTIMIZATION BY GENETIC ALGORITHM

本文在实数编码的遗传算法中对约束条件提出一种凝聚选择和复合形法处理的方法。利用凝聚函数求出个体的约束违背值,在选择中不仅考虑适应度值而且考虑约束违背值,使有潜力的个体优先被选择。利用复合形法对违背约束的个体进行改进,从而改善整个种群的性能、增加可行个体数量,有利于最优解的搜索。算例表明本文方法是可行和有效的。

This paper proposes surrogate reproduction and complex method for handling constraints in optimization by real-coded genetic algorithm. This reproduction considers both the fitness values and constraint function values calculated by surrogate function, which ensures potential designs to be chosen preferentially. Meanwhile, this paper improves the individual violating constraints using complex method, so the equality of the whole population is enhanced and the numbers of feasible design are increased. It is advantageous to the search of optimum solution. The classical examples show that the method proposed in this paper is feasible and effective.