论文部分内容阅读
提出了一种用Pareto遗传算法来实施的带约束的多目标混合变量优化方法,得到Pareto最优解集,决策者从中可选出满足设计需要的解.该算法包括6个基本算子:选择、变异、交叉、离散变量圆整算子、小生境、Pareto集合过滤器.建立了用于多目标优化的适应度函数,使用模糊罚函数法将带约束的多目标优化问题转换为无约束优化问题,同时提出了处理混合变量多目标优化问题中离散变量的方法.最后用算例说明了该方法的应用
A Pareto-based optimization approach with constrained multi-objective mixed variables is proposed to get the Pareto optimal solution set, from which the decision maker can choose the solution that satisfies the design requirements. The algorithm includes six basic operators: , Mutation, crossover, discrete variable integral operator, niche, Pareto set filter. The fitness function for multi-objective optimization is established and the fuzzy multi-objective optimization problem is transformed to unconstrained optimization using fuzzy penalty function At the same time, a method to deal with the discrete variables in the multi-objective optimization problem of mixed variables is proposed.At last, an example is given to illustrate the application of this method