Abstract: Aiming at the problems that the bird swarm algorithm has insufficient population diversity in the initial iterations, slow convergence speed in the later iterations and the tendency to fall into local optimum solution, a hybrid algorithm of bird swarm algorithm and arithmetic optimization algorithm based on Cauchy mutation（HBSAAOA） is proposed. Firstly, the location of producers in BSA is updated by using the high distribution of multiplication and division operators in the arithmetic optimization algorithm to improve the diversity of population and hence to enhance the global search ability. Then, a random search strategy and a Cauchy mutation strategy are introduced to generate candidate solutions, which will disturb the local exploitation in the later stage to enhance the ability of the algorithm jumping out of local optimum solution and to improve the convergence speed. Finally, the greedy strategy is used to select the best individual to replace the poor and thereby to improve the quality of the solution. Through the simulation experiments of 23 classical test functions and some CEC2014 benchmark functions, and applying HBSAAOA to two engineering application problems, the results show that the improved strategies are effective, and the improved algorithm has faster convergence speed, higher optimization accuracy and better robustness.

Key words: bird swarm optimization algorithm, arithmetic optimization algorithm, Cauchy mutation

摘要： 针对鸟群优化算法迭代初期种群多样性不足、迭代后期收敛速度慢、易陷入局部最优解等问题，提出一种融合柯西变异的鸟群与算术混合优化算法（hybrid algorithm of bird swarm algorithm and arithmetic optimization algorithm based on Cauchy mutation，HBSAAOA）。利用算术优化算法中乘除算子的高分布性对BSA中生产者的位置进行更新，以提高种群多样性，增强全局搜索能力。引入随机搜索策略和柯西变异策略来生成候选解，对后期局部开发阶段进行扰动，以增强算法跳出局部最优解的能力并提高收敛速度。利用贪婪策略对最优个体进行选择并替代较差的个体，从而提高解的质量。通过对23个经典测试函数以及部分CEC2014基准函数进行仿真实验，并将HBSAAOA应用到两个工程应用问题上，结果表明改进策略有效，改进算法的收敛速度更快、寻优精度更高，并且鲁棒性更好。

关键词: 鸟群优化算法, 算术优化算法, 柯西变异