基于混合策略改进的樽海鞘群算法

Improved salp swarm algorithm based on hybrid strategy

针对樽海鞘群算法(SSA)在求解复杂优化问题时存在的易陷入局部最优、收敛精度低等缺点,提出一种基于混合策略改进的樽海鞘群算法(ISSA).首先,采用Sobol序列实现樽海鞘种群的初始化,使初始种群在解空间中分布更加均匀,进而提高算法的全局寻优能力;其次,在领导者位置更新阶段引入步长控制因子,根据不同寻优时期自动调节领导者的搜索范围,有效平衡算法的全局搜索与局部搜索;然后,采用改进的透镜成像策略对领导者进行映射,避免算法陷入局部最优;此外,在追随者位置更新阶段,引入一种自主选择追随机制,改善追随者的盲从性,以提高算法的收敛精度;最后,与其他几种代表性优化算法在12个基准测试函数上进行仿真实验对比,并进行Wilcoxon秩和检验,实验结果表明所提出ISSA在收敛速度和精度上有明显提升,相较于其他优化算法具有更好的寻优效果和稳定性.另外,通过两个工程设计案例实验进行测试,进一步验证了所提出ISSA的可行性和适用性.

Aiming at the shortcomings of salp swarm algorithm(SSA)in solving complex optimization problems,such as easy to fall into local optimization and low convergence accuracy,an improved salp swarm algorithm(ISSA)based on hybrid strategy is proposed.Firstly,Sobol sequence is used to initialize the population of Tarpa scabbard,making the initial population more evenly distributed in the solution space,thereby improving the global optimization ability of the algorithm.Secondly,in the leader position update stage,a step size control factor is introduced to automatically adjust the leader’s search range according to different optimization periods,effectively balancing the global search and local search of the algorithm.Then,an improved lens imaging strategy is used to map leaders to avoid the algorithm falling into local optimization.In addition,in the follower position update stage,an autonomous selection following mechanism is introduced to improve the blind obedience of the follower to improve the convergence accuracy of the algorithm.Finally,compared with other representative optimization algorithms,simulation experiments are conducted on 12 benchmark test functions,and Wilcoxon rank sum test is conducted.The experimental results show that the ISSA proposed has significant improvement in convergence speed and accuracy,and has better optimization effect and stability compared with other optimization algorithms.In addition,two engineering design case experiments are conducted to further verify the feasibility and applicability of the proposed ISSA.