控制参数值非线性调整策略的灰狼优化算法

A Grey Wolf Optimization Algorithm Based on Nonlinear Adjustment Strategy of Control Parameter

为了克服灰狼优化算法在解决函数优化问题时容易陷入局部最优的缺陷,提出采用正弦曲线、对数曲线、正切曲线、余弦曲线和2次曲线的非线性调整策略控制参数值。同时采用变异策略对智能个体位置进行处理,使其位置受适应度值大小影响。对3个标准测试函数进行仿真表明,余弦曲线和2次曲线调整策略优于线性调整策略,其他3种非线性调整策略劣于线性策略。

Aimed at the problem that the Grey Wolf Optimization （GWO） algorithm is easily bogged down in local optimization in solving function optimization, this paper proposes a nonlinear adjustment strategy by adopting sinusoid, logarithmic, tangential, cosine and quadratic curves. And at the same time a strate- gy on mutating position of the agents is presented, whose position is influenced by fitness value. The ex- perimental results for three standard test functions show that the proposed cosine and the quadratic curve strategies are superior to the classical linear strategy, and the others such as the sinusoid strategy, the logarithmic strategy, and the tangential curve strategy are inferior to the linear strategy.