求解连续函数优化的自适应布谷鸟搜索算法

为了提高布谷鸟搜索算法求解连续函数优化问题的性能,提出一种自适应布谷鸟搜索算法,改进算法利用解与当前最优解之间对应维上距离,实现随机游动步长的自适应调整。距离当前最优解对应维越远,维的随机游动步长越长,反之越短。利用解的适应度与群体平均适应度的关系自适应调整发现概率,使劣质解比优秀解更容易被淘汰。将自适应布谷鸟算法应用于8个典型测试函数,实验结果表明,改进算法有效改善求解连续函数优化问题的性能,尤其适合求解高维、多峰的复杂函数。与相关的布谷鸟搜索算法比较,自适应布谷鸟搜索算法更具竞争力。

Self-adaptive cuckoo search algorithmfor continuous function optimization problems

To improve the performance of cuckoo search algorithm for continuous function optimization problems, an improved algorithm based on self-adaptive cuckoo search algorithm was proposed. The distance between the corresponding dimension of the current solution and the current optimal solution was used, and the improved algorithm realized the adaptive adjustment of the random walk, with the results of the farther the distance to the current optimal solution at the corresponding dimension, the longer the random walk, and the closer the distance to the current optimal solution at the corresponding dimension, the shorter the random walk. The relationship of the solution's fitness and the population average fitness was used to adaptively adjust to find probability, with bad solutions more easily eliminated than good solution. The simulation experiment on 8 benchmark functions shows that the improved algorithm efficiently improves the performance on continuous function optimization problem, especially suitable for solving high-dimension and multimodal function optimization problems. Comparison with the results of the related cuckoo search algorithm shows that the improved algorithm is competitive.