铁路编组站阶段计划动态调整方法研究

铁路编组站阶段计划执行过程中的不确定事件导致其无法持续最优,阶段计划动态调整方法对于提高编组站作业效率具有重要意义。研究提出包括时间预测、动态车流推算、计划调整、计划实施、实时信息采集反馈等步骤的阶段计划动态调整流程,通过计算残差相关系数进行数据属性相关性分析和降维处理,利用机器学习方法和神经网络模型预测各阶段作业过程用时,建立基于作业过程用时精准预测的动态车流推算模型,对动态车流推算过程进行符号化描述,提出模型的约束条件和目标函数,设计基于蚁群算法的编组站动态车流推算模型求解算法。结果表明,作业过程用时预测误差随训练集样本数量增大而逐渐减小,蚁群算法计算时间满足阶段计划动态调整实际应用的需要。

Dynamic Adjustment Method of Phase Plan for Railway Marshalling Station

The phase plan of railway marshalling stations cannot maintain optimal conditions due to uncertain events in the plan implementation,so the dynamic adjustment method of the phase plan for marshalling stations is very important for increasing the operation efficiency of marshalling stations.The dynamic adjustment process of the phase plan was researched and proposed,including time prediction,dynamic flow distribution,plan adjustment,plan implementation,and real-time information collection and feedback.Correlation analysis and dimension reduction of data attributes were carried out by the calculation of the residual correlation coefficient.Machine learning method and neural network model were utilized to predict the operation process time of each stage,and a dynamic flow distribution model based on the accurate prediction of operation process time was built.The dynamic flow distribution process was symbolically described,and the constraint condition and the objective function of the model were presented.Finally,a solution algorithm for the calculation model of dynamic flow distribution in marshalling stations based on the ant colony algorithm was designed.Results show that the prediction error of operation process time decreases gradually with an increase in the number of training samples,and the calculation time of the ant colony algorithm meets the needs of the practical application of phase plan dynamic adjustment.