基于种群密度的鲁棒多目标优化算法

An Improved Robust Multi-Objective Optimization Method Based on Population Density

为消除由于设计参数或环境因子扰动对多目标优化问题带来的影响,提出基于t分布构建有效目标函数,并在传统Pareto最优解评估策略基础上,一方面用种群在空间的分布密度替换NSGA2算法中基于距离的拥挤排序策略以维护外部档案;另一方面,引入基于种群分布密度的全局最优解概率选择策略和基于拉丁超立方的局部采样方法.实验结果表明,该算法能有效求解多目标问题的鲁棒Pareto最优解.

To solving the perturbed multi-objective optimization problems by design parameters or environmental factors,proposes the t distribution to construct effective objective function based on the Pareto optimal solution.On the one hand,the population density distribution replace the crowded ordering strategy of the NSGA2 algorithm to maintain external files.On the other hand,introduce the global optimal solution of population distribution density of probability selection strategy and local sampling method based on Latin hypercube.The experimental results show that,the algorithm can effectively solve the multi-objective robust Pareto optimal solution of the problem.