Abstract: Aiming at the problems of uneven initial population distribution, slow convergence speed, weak global search ability and fall into easily local optimum of standard WOA algorithm, an improved whale optimization algorithm based on mix strategy is proposed. Firstly, Sobol sequence is used to initialize the population to make distribution of initial population more even in solution space. Then, the global search and local development capabilities are balanced and improved by nonlinear time-varying factors and inertia weights, and random learning strategy is combined to increase the population diversity in the iterative process. Finally, the Cauchy mutation algorithm is introduced to improve the ability to jump out of the local optimum. Through the optimization experiment of 12 benchmark functions and the parameter estimation of a water resource demand forecasting model, the results show that the optimization accuracy and convergence speed of the improved whale optimization algorithm based on the mix strategy are significantly improved.

Key words: whale optimization algorithm, Sobol sequence, nonlinear strategy, inertia weight, stochastic learning, Cauchy mutation

摘要： 针对标准WOA算法初始种群分布不均、收敛速度较慢、全局搜索能力弱且易陷入局部最优等问题，提出一种混合策略改进的鲸鱼优化算法。采用Sobol序列初始化种群以使初始解在解空间分布更均匀；通过非线性时变因子和惯性权重平衡并提高全局搜索及局部开发能力，并结合随机性学习策略增加迭代过程中种群的多样性；引入柯西变异提升算法跳出局部最优的能力。通过对12个基准函数和一个水资源需求预测模型的参数估计进行优化实验，结果表明，基于混合策略改进的鲸鱼优化算法在寻优精度及收敛速度上均有明显提升。

关键词: 鲸鱼优化算法, Sobol序列, 非线性策略, 惯性权重, 随机性学习, 柯西变异