多策略改进的蜜獾优化算法

Multi Strategy Improved Honey Badger Optimization Algorithm

针对标准蜜獾算法存在的易陷入局部最优值、目标精度低、局部搜索能力不足等问题,提出了一种多策略改进的蜜獾算法.在种群初始化阶段,使用Chebyshev混沌映射初始化种群,保证随机性的同时提高种群的均衡性;在局部挖掘阶段加入莱维飞行,使寻找最优值到确定最优值的过程更加平稳,避免算法陷入局部最优;在最优个体确定阶段上引入最优个体自适应变异策略,来提高算法局部搜索能力以及收敛精度,避免算法个体早熟.选取18个基准测试函数进行仿真,同时结合秩和检验、实际工程应用对该算法进行多维度评估,实验证明改进的蜜獾算法与标准蜜獾算法相比,在收敛速度、目标精度以及寻优搜索能力均有明显改善,表现出较好的鲁棒性.

Aiming at the problems of standard honeybadger algorithm,such as falling into local optimal value,low target accuracy and insufficient local search ability,a multi strategy improved honeybadger algorithm is proposed.In the phase of population initialization,Chebyshev chaotic map is used to initialize the population to ensure the randomness and improve the balance of the population;In the local mining stage,Levi′s flight is added to make the process from finding the optimal value to determining the optimal value more stable and avoid the algorithm falling into the local optimum;In order to improve the local search ability and convergence accuracy of the algorithm,the adaptive mutation strategy of the optimal individual is introduced in the stage of determining the optimal individual,and avoid the premature of the algorithm.18 benchmark functions are selected for simulation.At the same time,the multi-dimensional evaluation of the algorithm is carried out in combination with rank sum test and practical engineering application.The experiment shows that the improved honey badger algorithm has significantly improved convergence speed,target accuracy and optimization search ability compared with the standard honey badger algorithm,and shows better robustness.