| 摘要: |
| 大多数现有的多目标进化算法（MOEA－Multiobjective Evolutionary Algorthm)都是基于Pareto机制的，如NPGA（Niched Pareto Genetic Alogrithm ,NSGA(Non-dominated SOrting Genetic Agorithm)等，这些算法的每一个循环都要对种群中的部分或全部个体进行排序或比较，计算量很大，文中介绍了一种基于变权重线性加权的Pareto轨迹法－WSTPEA（Weighted sum Approach and Traching Pareto Method)，该算法不是同时求得所有可能的非劣解，而是每执行一个循环步骤求得一个非劣解，通过权重变化次数控制算法循环的次数，从而使整个种群遍历Pareto曲线（面),文中给出了算法的详细措述和流程图，并且对两个实验测试问题进行了计算，最后对结果进行了分析。 |
| 关键词: 多目标优化 线性加权法 进化算法 |
| DOI： |
| 分类号: |
| 基金项目: |
|
| A multiobjective evolutionary algorithm based on weighted sum approach and tracing pareto method |
|
|
| Abstract: |
| The most existing multiobjective evolutionary algorithms (MOEAs) such as NPGA (Niched Pareto Genetic Algorithm),NSGA (Non-dominated Sorting Genetic Algorithm) etc.are based on Pareto mechanical.Each step for checking Pareto optimality requires sorting and pair-wise comparison of at least a subset of the population,thus increasing the computational needs.This paper introduces a new algorithm based on weighted sum approach and tracing Pareto method WSTPEA.The WSTPEA achieves a noninferior solution at each intermediate step which is not like the existing MOEAs that generate the total Pareto set in one run.In this paper the WSTPEA is described in detail and the flow chart of the algorithm is given.Two multiobjective problems are calculated and the solutions are analyzed. |
| Key words: multiobjective optimization weighted sum approach evolutionary algorithm |
