摘要
通过对物流车辆配送过程的研究分析,建立了不带时间窗约束的物流车辆配送路径优化模型。针对普通遗传算法早熟和易陷入局部最优的缺点,通过引入小生境技术对遗传算法进行了改进。其中,选择操作采用了最优个体保留和轮盘赌结合的策略,交叉操作采用最优路径子路径保留策略;变异算子保证个体逐代进化。个体交叉前的配对选择以及交叉完毕后的调整都引入了小生境选择机制。这些方法大大增强了种群的多样性和全局寻优能力,加速了算法的收敛。最后通过实例验证了算法的可行性、实用性和高效性。
According to the research and analysis of logistics vehicles distribution process, a mathematical model of logistics distribution route without time window is proposed. In allusion to the genetic algorithm＇s disadvantage of premature and easy to fall into a local optimum, a niche technology and other improved methods are introduced. The selecting process adopts optimal individual reservation and rou- lette to select individuals of next generation. The crossover process uses a niche selecting mechanism in the pairing selecting and the sub-route of optimal individual is reserved when crossovering. Mutate operator can improve the individual generation by generation. The variety and ability for global optimization of populations are advanced greatly. Finally, the algorithm is proved to be feasible, practical and effective by testing data.
出处
《计算机工程与科学》
CSCD
北大核心
2012年第10期118-122,共5页
Computer Engineering & Science
基金
国家自然科学基金资助项目(71002106)
校青年培育基金资助(090112)
关键词
物流配送
路径优化
遗传算法
小生境技术
logistics distribution
routing optimization
genetic algorithm
niche technology
