面向多目标优化问题的基于Species的遗传算法

为了能够快速准确地获得多目标优化问题的一组非支配解,提出了一种基于Species的多目标遗传算法.该算法采用Tchebycheff方法构建一定数量的子问题,进而基于Species机制构造多种群实现了对多个子问题的并行求解.这种采用多个体对一个最优解的搜索方式提高了算法的探索能力和开发能力.最后,对一组标准测试函数进行仿真实验,结果表明所提出的算法能够快速准确地获得一定数量的非支配解.

Species-Based Genetic Algorithm for Multiobjective Optimization Problems

In order to achieve a set of nondominated solutions for multiobjective optimization problems quickly and accurately， a Species-based genetic algorithm for multiobjecitve optimization problems was proposed. Firstly， a certain number of subproblems were developed with the Tchebycheff approach. Then multiple subpopulations were constructed based on the Species mechanism to solve all the subproblems simultaneously， which can improve the exploration and exploitation ability by using multiple individuals to search one optimal solution. Finally， a set of benchmark multiobjective functions were examined， and the experimental results showed that the proposed algorithm can obtain a certain number of nondominated solutions quickly and accurately.