多策略融合的蛇优化算法及其应用

针对蛇算法寻优阶段交互性差，初始种群随机程度严重，易陷入局部最优解等问题，提出了一种多策略融合的蛇优化算法(multi-strategy snake optimizer, MSSO)。首先，利用正交矩阵对蛇种群进行初始化，使个体分布更加均匀；其次，设计探索开发阶段切换的自适应方程，用以替换原有的食物量与温度阈值，使算法进行自适应阶段切换；最后，使用联合反向选择策略替换算法原有的新个体孵化方法，提高算法收敛精度的同时加快算法收敛效率。选取10个基准测试函数从不同角度对MSSO算法进行实验，测试算法性能，分析各策略的有效性，并使用Wilcoxon秩和检验来证明算法显著性，通过两个工程应用仿真实验来验证MSSO的实用性。各实验结果表明MSSO较比较算法综合表现更优，证明MSSO算法改进在寻优能力、鲁棒性、实用性等方面均有所提升。

Multi-strategy fusion snake optimizer and its application

This paper proposed a multi-strategy snake optimizer to address the problems of poor interactivity in the optimization-seeking phase of the snake algorithm, serious randomness of the initial population, and the tendency to fall into local optimal solutions. Firstly, it used an orthogonal matrix to initialize the snake population to make the individuals more uniformly distributed; secondly, it designed an adaptive equation to explore the development phase switching to replace the original food quantity and temperature threshold to make the algorithm perform adaptive phase switching; finally, it used a joint reverse selection strategy to replace the original new individual hatching method of the algorithm to improve the convergence accuracy of the algorithm while accelerating the convergence efficiency of the algorithm. It selected ten benchmark test functions to experiment the MSSO algorithm from different perspectives to test the algorithm performance, analyzed the effectiveness of each strategy, and used the Wilcoxon rank sum test to prove the algorithm significance, and verified the practicality of MSSO through two engineering application simulation experiments. The results of each experiment show that MSSO performs better than the comparative algorithm comprehensively, which proves that the MSSO algorithm improvement has improved in the aspects of the search ability, robustness and practicality.