混合型货物作业点取送车作业优化通用模型及算法

General Optimization Model and Algorithm for Placing-in and Taking-out Wagons in Mixed Freight Operation Sites

针对混合型货物作业点取送车作业问题,以最小化计划时段内调机取送车作业总时间(包括调机总走行时间、调机总等待时间和接发列车对调车作业的总影响时间)为优化目标函数,考虑不同货物作业点间调移作业要求的调机访问次序、同一货物作业点车组先送后取顺序、调机牵引能力限制、货物作业点容车数、批次作业的开始及结束时分等约束条件,建立混合型货物作业点取送车作业优化的通用模型。运用模拟退火算法作为主函数,嵌入逆序判断及处理函数、批次划分函数、批次作业车数判断函数和取送作业时间计算函数,对模型进行求解,并以作业形式具有广泛代表性的岳阳南站案例,对模型及算法进行验证。结果表明:模型能较好地对计划阶段内具有多种形式的取送车作业做出安排;优化后岳阳南站取送车方案的总等待时间为8 min,总影响时间为0 min,因而所提出的模型和算法是可行的和有效的。

For the placing-in and taking-out wagon systems in mixed freight operation sites,with minimizing the total time for shunting locomotive placing-in and taking-out wagons in the planned period(including the total running time,total waiting time of shunting locomotive and the total influence time of train receiving and departure on shunting operation)as the optimization objective function,this paper establishes a general model for placing-in and taking-out wagons in mixed freight operation sites,considers the following factors:constraints of shunting locomotive visiting sequence required by transferring operation between different freight operation sites,the ordering of the first placing-in and then taking-out in the same freight operation site,the traction capacity limit of shunting locomotive,the wagon-accommodation capacity of freight operation site,the start and finish time for batch operation.Using simulated annealing algorithm as the main function,the model is solved by embedding reverse-order judgment and processing functions,the batch division function,the wagon-numberjudging function in batch operation,and the function of computing time for placing-in and taking-out wagons.Yueyang South Station is taken as an example to verify the model and algorithm.The results show that the model can well arrange the placing-in and taking-out wagons with various forms in the planned period.The total waiting time of placing-in and taking-out wagons obtained in the optimization scheme of Yueyang South Station is 8 min and the total influence time is 0 min,which shows that the proposed model and algorithm are feasible and effective.