关键词: 混合蛙跳算法, 邻近学习策略, 动态蛙跳策略, 差分进化, 并行计算

Abstract:

Aiming at the problems of slow convergence speed， low solution accuracy and low operation efficiency of Shuffled Frog-Leaping Algorithm （SFLA） in complex optimization problems， a Parallel SFLA based on Adaptive weight adjustment and Differential Evolution （DE） strategy （P-DE-ASFLA） was proposed. In the local search process， the optimal individual in the subgroup was updated by the proximity learning strategy to speed up the convergence of the algorithm； the worst individual in the subgroup was updated by the dynamic frog leaping rule to avoid premature convergence of the algorithm. In the global search process， the DE strategy was used to update the genes of the mixed population to enhance the global optimization ability of the algorithm. At the same time， based on the master-slave parallel architecture， the multi-process technology was used to parallelize the local search process of the subgroup， which greatly improved the operation efficiency of the algorithm. The experimental results show that the solution quality and operation efficiency of the algorithm in six standard test functions are far better than standard SFLA and DE algorithm.

Key words: Shuffled Frog-Leaping Algorithm (SFLA), proximity learning strategy, dynamic frog-leap strategy, Differential Evolution (DE), parallel computing