多阶段动态扰动和动态惯性权重的布谷鸟算法

针对布谷鸟仿生智能优化算法存在着的易陷入局部最优、求解精度低以及收敛速度慢等问题，提出了基于多阶段动态扰动和动态惯性权重的布谷鸟搜索算法（MACS）。利用多阶段动态扰动策略对布谷鸟算法的全局位置的最优鸟巢位置根据方差可调的正态随机分布进行扰动，有利于增加种群的多样性和鸟窝位置的灵活性，提高算法全局搜索能力。在局部位置处引入动态惯性权重，使得算法有效克服易陷入局部最优的缺陷，提高局部寻优搜索能力。引入了动态切换概率p]代替固定概率，可以动态平衡全局搜索和局部搜索。通过与4种算法相比和11个测试函数的仿真结果表明：改进布谷鸟算法（MACS）的寻优性能明显提高，收敛速度更快，求解精度更高，具有更强的全局搜索能力和跳出局部最优能力。

Cuckoo Algorithm for Multi-stage Dynamic Disturbance and Dynamic Inertia Weight

The bionic intelligent optimization algorithm of cuckoo bird is prone to fall into the problems of low local optimal solution accuracy and slow convergence speed, This paper presents a cuckoo search algorithm based on multi-stage dynamic disturbance and dynamic inertia weight（MACS）. Firstly, the multi-stage dynamic disturbance strategy is used to disturb the optimal nest position of the global position of cuckoo algorithm according to the normal random distribution with adjustable variance. It is beneficial to increase the diversity of the population and the flexibility of the nest location, and improve the global searching ability of the algorithm. Secondly, the dynamic inertia weight is introduced in the local position, which makes the algorithm overcome the defect of falling into the local optimum effectively and improve the local optimization search ability. Finally, dynamic switching probability p] is introduced instead of fixed probability, which can dynamically balance global searching and local searching. This paper compares with 4 algorithms and 11 test functions. The simulation results show that the improved cuckoo algorithm（MACS） has better optimization performance, faster convergence speed, higher solution accuracy, stronger global searching ability and jumping out of local optimal ability.