摘要
针对基本混合蛙跳算法随机性强,在处理复杂函数优化问题时容易陷入局部最优、收敛速度慢的缺点,提出了一种改进的混合蛙跳算法,该算法利用高斯变异算子对子群最差青蛙进行适当的扰动,修正了其更新策略,从而维持了群体的多样性。用典型测试函数对粒子群优化算法、基本混合蛙跳算法及改进算法进行对比实验,仿真结果验证了新算法的有效性和鲁棒性。
With strong random, basic Shuffled Frog Leaping Algorithm （SFLA） algorithm easily traps into local optima and has a slow convergence speed when it is used to address complex functions, in order to overcome the shortcomings, an improved SFLA is proposed. The worst frog in the subpopulation is properly disturbed by the use of gauss mutation operator, and updating strategy is modified, and thus maintains the population diversity. Particle Swarm Optimization （PSO） algorithm, SFLA and its improved algorithm are compared by using four benchmark test functions. Also, simulation results demonstrate the effectiveness and robustness of the improved SFLA.
出处
《商洛学院学报》
2009年第2期19-22,共4页
Journal of Shangluo University
基金
国家自然科学基金项目(60674108)
关键词
仿生优化
混合蛙跳算法
智能优化
复杂函数
bionic optimization
shuffled frog leaping algorithm
intelligent optimization
complex functions
