铁路站场往返取车作业优化模型及算法

Optimization Model and Algorithm of Wagon Placing-in and Taking-out Operations in a Marshaling Yard

为了实时优化铁路站场取车计划,给出调机处于不同起始位置时,单次连挂调机走行距离的计算公式。以调车机车总走行距离最小为优化目标,考虑车列最大长度限制和最晚到达约束,建立铁路站场多次往返取车优化模型,将多次往返取车模型抽象成权重动态变化的带权有向图的最小权重路径问题,采用改进的蚁群算法求解。以某铁路编组站树枝形站场取车作业为例,验证模型和算法的有效性。结果表明,随着停留车组数量的增加,改进蚁群算法收敛到最优解的迭代次数逐渐增加,在实际应用要求的计算时间限制内,可收敛并得到最优解。

To optimize the wagon taking-out plan of marshaling yard in real time, the calculation formula of running mileage of single coupling shunting locomotive is given when the shunting locomotive is at different starting positions. Taking the minimum total running distance of shunting locomotives as the optimization objective, considering the constraints of the maximum length of a train and the latest arrival time, the optimization model of wagon placing-in and taking-out shunting operations is formulated. The multiple-trip placing-in and taking-out operation model is abstracted into the minimum weight route problem of a weighted directed graph with dynamically changing weights, which then is solved by the improved ant colony algorithm. A taking-out shunting in a branchshaped railway shunting yard of a marshaling yard is taken as an example to validate the model and the algorithm. The results show that the iteration number of the modified ACA to be converged to its optimal solution gradually increases as the number of wagon groups rises. It is feasible to obtain the corresponding optimal solution within the permitted calculation time of practical applications.