针对一个Pareto局部搜索(PLS)算法在解决多目标组合优化问题中所得到的解集与初始点的选取有关,提出该算法的改进。改进算法从初始解开始进行PLS搜索产生一组改进解集VF,然后对VF中的所有解再进行PLS搜索,如此重复直到满足终止条件。实例计算表明,PLSⅠ算法和算法Ⅱ能得到很好的解且解的质量优于PLS算法。
The Pareto Local Search(PLS) algorithm is a novel method in multi-objective combinatorial optimization,which searches for a multi-objective local optimal solution set from an initial solution.To improve the algorithm,we take the solutions in a multi-objective local optimal solution set as initial solutions,and further search for a multi-objective local optimal solution set,until a stopping criterion is satisfied.Computational tests on several instances demonstrate the effectiveness of the improved algorithm.