Metaheuristic algorithms are important methods to solve optimization problems and maintaining a balance between the global exploration and local exploitation is crucial to the performance of such algorithms. We propose a self-adaptive multi strategy cuckoo search algorithm (MSACS) based on the cuckoo search algorithm (CS). First, five different search strategies were proposed to calculate the use probability and control parameters by using adaptive strategies to ensure that the algorithm can autonomously adjust according to the change in the functions and iteration times. Second, the performance of the MSACS was tested on 28 common benchmark functions and compared with the performance of several CS algorithms, particle swarm optimization (PSO) algorithms and difference evolution algorithms (DE). MSACS achieved the best results on 17 of these functions and performed well on the remaining 11 functions. Finally, the improved algorithm was applied to the optimization of a ball screw driving system model. By adjusting the dimensionless input velocity function, the peak acceleration of screw is reduced and the peak acceleration of crank angle is reasonable.

元启发式算法是解决优化问题的重要方法，在全局探索和局部开发之间保持平衡对此类算法的性能至关重要。我们提出了一种基于布谷鸟搜索算法（CS）的自适应多策略布谷鸟搜索算法（MSACS）。首先，提出了五种不同的搜索策略，通过使用自适应策略来计算使用概率和控制参数，以确保算法可以根据函数和迭代时间的变化来自主调整。其次，在28个通用基准功能上测试了MSACS的性能，并与几种CS算法，粒子群优化（PSO）算法和差异演化算法（DE）的性能进行了比较。MSACS在其中的17个功能上获得了最佳结果，而在其余11个功能上则表现出色。最后，将改进算法应用于滚珠丝杠传动系统模型的优化。通过调整无量纲输入速度函数，可以减小螺杆的峰值加速度，并且曲柄角的峰值加速度是合理的。