一种新型非线性收敛因子的灰狼优化算法

Novel grey wolf optimization algorithm based on nonlinear convergence factor

针对标准灰狼优化算法在求解复杂工程优化问题时存在求解精度不高和易陷入局部最优的缺点,提出一种新型灰狼优化算法用于求解无约束连续函数优化问题。该算法首先利用反向学习策略产生初始种群个体,为算法全局搜索奠定基础;受粒子群优化算法的启发,提出一种非线性递减收敛因子更新公式,其动态调整以平衡算法的全局搜索能力和局部搜索能力;为避免算法陷入局部最优,对当前最优灰狼个体进行变异操作。对10个测试函数进行仿真实验,结果表明,与标准灰狼优化算法相比,改进灰狼优化算法具有更好的求解精度和更快的收敛速度。

The classical grey wolf optimization （GWO） algorithm has a few disadvantages of low solving precision and high possibility of being trapped in local optimum. This paper proposed a novel grey wolf optimization （NGWO） algorithm for solving unconstrained optimization problems. The proposed algorithm used opposition-based learning strategy to initiate population, which strengthened the diversity of global searching. Inspired by particle swarm optimization （PSO）, this paper proposed an improved convergence factor update equation, which was based on that the values of parameter a are nonlinearly decreased over the course of iterations. The convergence factor was dynamically adjusted to maintain a better balance between global search and local search. Mutation operator was given on the current optimal individual of each generation, thus it could effectively jump out of local minima. Experiments are conducted on a set of 10 unconstrained benchmark functions. Based on the results,the proposed NGWO algorithm shows significantly better performance than the standard GWO algorithm.