集成正态云和动态扰动的哈里斯鹰优化算法

Harris Hawk Optimization Algorithm Integrating Normal Clouds and Dynamic Perturbations

针对基本哈里斯鹰算法收敛速度慢和易陷入局部寻优等问题,提出一种集成正态云模型和动态扰动策略的改进哈里斯鹰算法.在全局搜索阶段,利用正态云模型的随机性和模糊性和随机反向学习思想对哈里斯鹰位置进行更新,从而丰富种群多样性和提高算法全局搜索能力.在局部开发阶段,引入动态扰动策略更新不同捕食策略下的哈里斯鹰位置,改善算法局部开发能力.选取不同维度的基准测试函数进行仿真,同时利用Wilcoxon秩和检验、Friedman检验、三杆桁架设计问题对改进哈里斯鹰算法和其它优化算法进行对比分析.实验结果表明:提出的改进哈里斯鹰算法收敛速度更快,寻优精度更高,验证了改进策略的有效性.

To address the problems of slow convergence speed and easy to fall into local search for excellence of the basic Harris Hawk algorithm,an improved Harris Hawk algorithm integrating the normal cloud model and dynamic perturbation strategy is proposed.In the global search phase,the randomness and fuzziness of the normal cloud model and the idea of stochastic backward learning are used to update the Harris Hawk positions,so as to enrich the population diversity and improve the global search ability of the algorithm.In the local exploitation phase,dynamic perturbation strategies are introduced to update the Harris hawk positions under different predation strategies to improve the local exploitation ability of the algorithm.Benchmark test functions of different dimensions are selected for simulation,while the improved Harris hawk algorithm and other optimization algorithms are compared and analyzed using Wilcoxon rank sum test,Friedman test,and three-rod truss design problem.The experimental results show that the proposed improved Harris Hawk algorithm converges faster and has higher accuracy of the optimization search,which verifies the effectiveness of the improved strategy.