论文部分内容阅读
解决多目标优化问题,并得到精确的、高质量的Pareto前沿解是非常具有挑战性的。将CS算法运用于多目标问题解的迭代更新过程,对传统的基于Pareto支配关系的适应度函数进行了改进,并提出基于小生境技术的逐步档案缩减法用于档案解的缩减与维护过程,设计出了多目标布谷鸟搜索算法(MOCS)。通过仿真实验验证以及相关性能指标的测试结果得出,MOCS算法与经典的NSGAII算法相比,在所得解的收敛性、多样性和均匀性方面均有所改善。
Solving multi-objective optimization problems and getting accurate, high-quality Pareto front solutions are challenging. The CS algorithm is applied to the iterative updating process of multi-objective problem solution, the traditional fitness function based on Pareto domination is improved, and the step-by-step file reduction method based on niche technology is proposed to reduce and maintain the file solution. Designed a multi-target cuckoo search algorithm (MOCS). The simulation results and the test results of relevant performance indexes show that the MOCS algorithm improves the convergence, diversity and uniformity of the solutions compared with the classical NSGAII algorithm.