【摘要】 针对鲸鱼优化算法收敛速度慢和寻优精度低的缺点，提出一种基于概率精英差分和自适应黄金正弦的鲸鱼优化算法。基于最大最小思想优化拉丁超立方体抽样来初始化鲸鱼种群，使初始种群分布更加均匀，拥有更好的全局搜索能力；提出融合余弦自适应算子的黄金正弦算法改进鲸鱼的螺旋更新，加快收敛速度，提高收敛精度；设计概率精英差分变异方法并进行贪婪选择，优化算法流程，增强算法跳出陷入局部最优的能力。选取4个单峰测试函数、4个多峰测试函数和5个多最优解的多模态测试函数与主流优化算法进行对比实验，实验结果表明，该算法具有更高的寻优精度、更快的收敛速度以及更优的全局搜索能力，通过消融实验验证了该算法改进策略的有效性。更多还原

【Abstract】 A whale optimization algorithm based on probabilistic elite difference mutation and adaptive golden sine was proposed to solve the shortcomings of low convergence speed and low optimization accuracy. The Latin hypercube sampling based on the idea of maximum and minimum was optimized to initialize the whale population, which made the initial population distribution more uniform with better global search ability. The golden sine algorithm with cosine adaptive operator was proposed to improve the spiral updating of whale, so as to accelerate the convergence speed and improve the convergence accuracy. The probabilistic elite difference variation method was designed and the greedy selection was carried out to optimize the algorithm flow and enhance the ability of the algorithm to jump out of the local optimal. Five single-peak test functions, five multi-peak test functions and five multi-modal test functions with multiple optimal solutions were selected to conduct comparison experiments with the mainstream optimization algorithms. Experimental results show that the proposed algorithm has higher optimization accuracy, higher convergence speed and better global search ability, and the effectiveness of the improved strategy of the algorithm is verified by the ablation experiment.更多还原