（ 重庆大学数统学院，重庆市 404100； ）
摘要： 铁路区段内列车运行调整问题作为铁路行车调度指挥工作中的核心技术, 决定着区段内行车秩序的优劣。列车运行调整的目的就是最终使列车实际运行图与计划运行图的差异达到最小化。因此建立有效的列车运行调整模型，设计良好的算法解决该模型成为关键。本文将混合蛙跳算法的分组思想引入到粒子群算法中这样就保证了各个小组内粒子间的差异性，有利于粒子位置的更新。改进的算法能够有效地避免早熟收敛问题，并能较大幅度地提高收敛速度和收敛精度。实验结果表明：当运行趋于稳定状态时，蛙跳简化粒子群算法比普通粒子群算法、差分进化算法和混合蛙跳算法收敛速度快、能更有效地求得全局最优解。
关键词： 列车运行调整；粒子群算法；差分进化算法；混合蛙跳算法; 蛙跳简化粒子群算法
WANG Juntong， HU Xiaobing*
（ School of mathematics and statistics, Chongqing University, Chongqing 404100； ）
Railway train operation adjustment within section that works as the core technology of the Railway Dispathwork determines the merits of the order in rode sector.The purpose of the train operation adjustment is the final plane to make the difference between the actual and planned diagram minimized. Therefore, the establishment of effective train operation adjustment model and a well-designed algorithm to solve the model become critical.Grouping Ideas of the Shuffled Frog Leaping Optimization in this article which is introduced into the particle swarm algorithm so as to ensure the difference between the groups within the particle ,is advantageous to the particle position update.The new algorithm not only outperforms PSO in terms of accuracy and convergence rate but also avoids effectively being trapped in local minima.The experimental results show that: while running stable, the convergence speed of Shuffled Frog Leaping Simplified Particle Swarm Optimization is faster and the global optimal solution can be more efficiently than the Particle Swarm Optimization ,the differential evolution algorithm and the Shuffled Frog Leaping Optimization.