Abstract: In order to solve the problem of Cuckoo Search（CS） algorithm, a new Elite Cuckoo Search algorithm with Gravitational search and Gauss perturbation（GGECS） is proposed. In this algorithm, a self-adaptive control strategy is put forward to adjust step size factor and discovery probability dynamically in cuckoo algorithm. Besides, the population is classified by using the Pareto principle. Gravity search and Gaussian perturbation are applied to the bird’s nests which belong to different population. By this method, the population diversity of the algorithm is improved. It can avoid falling into a local optimal solution, improve the searching precision and accelerate the convergence speed. This algorithm uses eight benchmark test functions to carry out simulation experiments. The results show GGECS algorithm has better global search capability than CS algorithm and Improved Cuckoo Search（ICS） algorithm. The results from GGECS algorithm are more close to theory optimal solution.

Key words: Cuckoo Search（CS） algorithm, self-adaptation, elite strategy, gravitational search, Gaussian perturbation

摘要： 针对布谷鸟算法（CS）的不足，提出了混合引力搜索与高斯扰动的精英布谷鸟搜索算法（GGECS）。该算法提出了自适应控制策略，将布谷鸟算法中的步长因子和发现概率进行动态地调整，并使用帕累托法则进行精英分类，分别对属于不同类别的鸟巢进行引力搜索和高斯扰动，从而提高算法的种群多样性，避免算法陷入局部最优解，提高了算法的寻优精度和收敛速度。使用8个标准测试函数进行仿真实验。结果表明，该算法较CS和ICS算法具有更好的全局搜索能力，其测试函数最优解也更为接近最优解的理论值。

关键词: 布谷鸟搜索算法, 自适应, 精英策略, 引力搜索, 高斯扰动