单车辆物流配送决策模型及其遗传算法

Decision-making Model and Its Genetic Algorithm for Single-vehicle Distribution

从提高装配运输系统运作效率及效益的角度出发,综合考虑配送时窗、配送重量、体积及车辆装载力等约束因素,建立了单车辆配装运输决策模型.通过遗传算法应用在模型中的适应性分析及可能存在的问题或缺陷分析,采用自然数编码、最大保留交叉及自适应交叉变异等技术以遗传父代优良特性、改善算法的早熟及进化停滞现象并增强其寻优性能.算法中嵌入配送重量、体积及配送时窗约束检验子程序以减小编码难度并提高解的精准度.最后以实例说明模型的应用,并指出进一步的研究方向.

To improve efficiency and benefit for loading-transportation system,a decision-making model for single vehicle loading-transportation,under comprehensive consideration of the restraints of goods weight and volume as well as loading capacity of vehicle and time-window,is developed.Based on an analysis of the adaptability for genetic algorithm application to the model as well as its possible problems or defects,technologies such as decimal coding and maximum preserved crossover operator as well as adaptive crossover mutation are adopted to inherit good properties of parents and improve the phenomenon of algorithm premature and stagnation as well as enhance its optimization performance.Constraint test subroutine for distribution weight,volume and delivery time window embedded in genetic algorithm in order to reduce the coding difficulty and improve solution＇s accuracy.An example is given to explain the process.In the end,further research orientation is discussed.