一种自适应交替的差分混合蛙跳优化算法

An Adaptive Alternating Optimization Algorithm of Differential Shuffled Frog Leaping

针对混合蛙跳算法在解决高维优化问题时易早熟收敛、求解精度低等问题,提出一种自适应交替的差分混合蛙跳优化算法。采用粒子群算法在短时间内产生一组满足约束条件的初始解,以提高初始解的质量。在此基础上,利用差分进化算法全局搜索能力强、种群多样性好等优点,设计一种自适应选择机制,动态地交替使用混合蛙跳算法和差分进化算法,使两者有机融合、优势互补。对6个经典函数的仿真测试结果表明,该算法可以丰富粒子的多样性,使算法前期和后期都具有较好的寻优能力,且寻优速率、求解精度、稳定性都优于混合蛙跳算法、差分进化算法和差分混合蛙跳算法。

Because of the problems of Shuffled Frog Leaping Algorithm （SFLA） such as premature convergence and low accuracy for hard high-dimensional optimization problems,an adaptive alternating optimization algorithm of differential shuffled frog leaping called ADE-SFLA is presented.In order to improve the quality of the initial solution,this algorithm uses Particle Swarm Optimization （PSO） to generate a group of initial solution that satisfies the constraints.On this basics,it draws on that Differential Evolution （DE） algorithm has strong global search capability,better population diversity,etc.It designs an adaptive selection mechanism to dynamically alternate SFLA and DE,and builds a win-win relationship between them and complementary advantages.Six classic functions of the simulation results show that the algorithm not only can enrich the diversity of particle,but also can make the algorithm have a better pre-and postoptimization ability.Its optimization rate,solution accuracy,stability are better than SFLA and DE,and are also better than the differential SFLA which is compared.