基于遗传算法的集装箱装车配载方案的优化

GA-Based Matching Scheme Optimization for Container Loading onto Railway Vehicle

针对手工编制集装箱装车配载方案费时、配载不合理等问题,基于集装箱和铁路车辆的参数,推导集装箱装车后铁路车辆的转向架承重、左右两转向架承重差和合重心横向偏离量的计算公式。以1组待装车的左右两转向架承重差之和最小为目标函数,以车辆的载重、转向架承重、左右两转向架承重差和合重心横向偏离量以及集装箱的箱重限制为约束条件,建立集装箱装车配载方案的优化模型。基于遗传算法原理,设计启发式遗传算法对该模型求解,采用罚函数方法和修复策略处理约束条件,并针对编码提出特殊交叉算子和变异算子,以保证生成可行解。以将42个20英尺集装箱和16个40英尺集装箱配载在37辆铁路车辆上为例,采用建立的模型和启发式遗传算法得到集装箱装车的配载方案。该方案不仅满足铁路车辆装载技术要求,而且优于手工编制的配载方案,验证了该优化模型及其求解算法的有效性和合理性,可以实现集装箱装车配载的智能化。

As to the time-consuming and unreasonable matching scheme worked out manually for container loading onto railway vehicle, based on container parameters and railway vehicle parameters, the computa- tional formulas were deduced for the bearing of bogie, the loading difference between left and right bogies as well as the lateral deviation of the combined center of gravity for a loaded vehicle. With minimizing the sum of the bearing differences between the left and right bogies of a vehicle to be loaded, taking the load- ing capacity of vehicle, the bearing of bogie, the loading difference between left and right bogies, the later- al deviation of the combined center of gravity and the limit for the weight of container as constraint condi- tions, the optimization model of the matching scheme for containers loading onto railway vehicles was con- structed. Based on the principle of genetic algorithm, a heuristic genetic algorithm was designed to solve this model, which used the penalty function method and the repair strategy to handle the constraint condi- tions, designed special crossover and mutation operators to ensure the generation of feasible solution. Taking 42 20-foot containers and 16 40-foot containers loading onto 37 railway vehicles as example, the matching scheme of containers was obtained using the proposed model and heuristic genetic algorithm. The matching scheme not only has met the technical requirements for loading railway vehicles, but is also superior to the manually worked out one. The computational example has verified the effectiveness and rationality of the proposed optimization model and its solving algorithm, and can realize the intelligent matching of containers loading onto railway vehicles.