基于MATLAB的改进海马算法

Improved Seahorse Algorithm Based on MATLAB

摘要: 针对海马算法寻优精度不足、易陷入局部最优的问题，本文提出一种基于Singer混沌及失败者放逐的海马算法。在初始化种群阶段，Singer混沌映射被引入用以生成遍历搜索空间的初始海马个体，增强了初始种群的多样性，有利于提高算法的搜索精度；在海马捕食阶段，引入失败者放逐策略，将捕食失败的海马个体放逐到搜索空间内，有利于算法跳出局部最优；在海马繁殖阶段，引入动态繁殖策略，动态影响父本和母本的权重，有利于防止算法过早收敛到局部极值。在算法性能测试实验中选用了10个基准测试函数和10个CEC2013测试函数，实验结果表明本文所提改进海马算法在寻优精度和收敛速度上都有较大提升，是一种优化能力强、鲁棒性好的算法。

Abstract: This paper proposes a multi-strategy improved seahorse algorithm to address the problem that the seahorse algorithm is not sufficiently accurate and easily falls into local optimum. In the initialization phase, Singer chaos mapping is introduced to generate the initial seahorse individuals traversing the search space, which enhances the diversity of the initial population and helps to improve the search accuracy of the algorithm; in the seahorse predation phase, a loser banishment strategy was introduced to banish the seahorse individuals that failed to feed into the search space, which is conducive to the algorithm jumping out of the local optimum; in the seahorse reproduction phase, a dynamic reproduction strategy was introduced to dynamically influence the weights of the fathers and the mothers, which is conducive to preventing the algorithm from converging to the local extreme prematurely. Ten benchmark test functions and ten CEC2013 test functions were used to test the performance of the algorithm. The experimental results show that the improved seahorse algorithm proposed in this paper has a large improvement in both the search accuracy and convergence speed, and is an algorithm with strong optimization capability and good robustness.