多车场集送一体化车辆调度问题及其遗传算法研究

Study on Multi-depots Integrated Vehicle Scheduling Problem and Its Genetic Algorithm

针对物流配送中的多车场一体化车辆调度问题提出了智能处理方法,采用了基于自然数的一体化配送对路径表示方式,用里程约束来控制车场的插入,以增加惩罚的方式加入时间窗约束。并且根据具体约束情况设计了改进的遗传算法,采用了动态染色体、改进的交叉和变异法、内部扰动和外部扰动等技术,提高了遗传算法的优化效率和优化效果。介绍了此算法的原理,给出了具有一个代表性算例试验结果和结果分析。试验结果表明了此方法对优化有里程和时间窗约束的多车场一体化车辆调度问题的有效性。

An intelligent technique is presented to the multi-depots integrated vehicle scheduling problem. The method is based on the delivery route on nature number. Because of the difference between delivery distance and time window limits, this method combines penalty function and the properties of the limits to control the distribution of depot. An improved genetic algorithm （GA） is proposed to optimize the multi-depots integrated VSP with delivery distance and time windows limits. The algorithm is also based on the chromosome on nature number, but some techniques such as dynamic chromosome, improved crossover and mutation, inside perturbation, outside perturbation are used in the algorithm. In this paper, the principium of the GA is introduced, a representative result and the analysis are given. The experiment indicates the validity of the improved GA to the VSP with the above-mentioned conditions.