一种新的基于进化策略的多目标优化算法

用进化策略求解多目标优化问题时，为了提高解在决策变量空间中的搜索能力和保证Pareto前沿的多样性，提出了一种新的基于进化策略的多目标优化算法。运用自适应变异步长的进化策略，使解在决策变量空间中进行全局和局部搜索；并引入非劣解按一定比例进入下一代的方法，使完全被占优的个体有机会参与到下一代的繁殖，保持了解在Pareto前沿的多样性。该算法在保证解在决策空间多样性的同时，也保持了Pareto前沿的多样性。仿真实验表明，该算法具有良好的搜索性能。

Multiobjective Optimization Algorithm Based on Evolutionary Strategy

While using evolutionary strategy to solve multiobjective optimization,in order to improve exploration of the solutions in decision space and maintain the diversity of the pareto front,a multiobjective optimization algorithm based on evolutionary strategy is presented.The evolutionary strategy of self-adaptive mutation step is used to search solutions in the globle area and local area.And the non-dominated solution in certain ratio enters the next generation,so the dominated individual has opportunity to participate multiplying in the next generation,and the diversity of the pareto front is assured.The simulation results show the good performance of the proposed algorithm.