Abstract:  In order to overcome the defects of basic grey wolf optimization algorithm about the dependence on the initial population, the presence of premature convergence, and easily getting into local minima, this paper proposes an improved grey wolf optimization algorithm applied to solve the function optimization problem. In the proposed algorithm, firstly, chaotic Cat sequence is used to initiate individual position, which can strengthen the diversity of global searching. Secondly, the individual memory from PSO are applied to enhance the local search ability and convergence speed. Thirdly, a Gaussian disturbance based the rules of survival of the fittest will be given on the global optimum of each generation, thus the algorithm can effectively jump out of local minima. Experimental results based on the thirteen benchmark functions show that the proposed improved GWO algorithm is superior to the basic GWO, PSO, GA and ACO algorithm in both computational accuracy and convergence rate.

Key words: chaotic Cat map, grey wolf optimization algorithm, function optimization, Gaussian mutation, the rules of survival of the fittest

摘要： 针对基本灰狼优化算法在求解复杂问题时同样存在依赖初始种群、过早收敛、易陷入局部最优等缺点，提出一种改进的灰狼优化算法应用于求解函数优化问题中。该算法首先利用混沌Cat映射产生灰狼种群的初始位置，为算法全局搜索过程的种群多样性奠定基础；同时引入粒子群算法中的个体记忆功能以便增强算法的局部搜索能力和加快其收敛速度；最后采用高斯变异扰动和优胜劣汰选择规则对当前最优解进行变异操作以避免算法陷入局部最优。对13个基准测试函数进行仿真实验，结果表明，与基本GWO算法、PSO算法、GA算法以及ACO算法相比，该算法具有更好的求解精度和更快的收敛速度。

关键词: 混沌cat 映射, 灰狼优化算法, 函数优化, 高斯变异, 优胜劣汰选择