多目标布谷鸟搜索算法

Multi-objective Cuckoo Search Algorithm

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。

It is challenging to solve multi-objective optimization problems with getting high-quality Pareto fronts accurately. The multi-objective Cuckoo Search algorithm(MOCS) was designed by firstly applying the recently developed Cuckoo Search Algorithm(CS) in solving Multi-objective optimization problems, and the fitness function based Pareto definiteness was improved, and the Gradual archive reduction method based on niche technology was proposed to improve the Archive solutions quality. The simulation test results and related performance indicators of nine test problems show that, MOCS algorithm is obviously improved in the aspect of the convergence, the diversity and the uniformity compared with the classic NSGA-II algorithm.