Abstract: Cuckoo Search algorithm（CS） is a swarm intelligence optimization algorithm which is neoteric and inspired by biology. To overcome the defections that standard algorithm has slow convergence rate and the search for algorithm optimization ability is weak, a new Adaptive Probability of Cuckoo Search algorithm（APCS） is proposed. Firstly, a parameter called state discriminant [Ps] is proposed and calculated by using Pareto optimal solution. Secondly, an equilibrium parameters [Peb] is created and calculated by exploration and exploitation equilibrium state. Finally, the adaptive dynamic adjustment of the discovery probability [Pa] is realized. In the end, eight benchmark functions are used to compare and analyze the performance of the two algorithms in 10 and 30 dimensions. Experiment results show that the convergence speed, optimization ability, stability and calculation time of APCS algorithm are better than those of CS algorithm.

Key words: cuckoo search algorithm, convergence, dynamic parameters, global search, benchmarks

摘要： 布谷鸟搜索算法（CS）是一种受生物启发的新型群智能优化算法。针对CS算法在搜索后期收敛速度慢并且寻优能力弱的问题，提出一种发现概率参数自适应调节的布谷鸟改进算法（APCS）。首先利用Pareto最优解计算出状态判别参数[Ps]，其次通过探索-开发平衡状态计算出平衡参数[Peb]，最终实现鸟蛋的被发现概率[Pa]的自适应动态调整。最后通过8个基准函数对两种算法的性能在10维和30维的情况下分别进行了对比与分析，结果表明，APCS算法的收敛速度、寻优能力、稳定性和计算时间都优于CS算法。

关键词: 布谷鸟搜索算法, 收敛性, 动态参数, 全局搜索, 基准测试