基于时间序列多品种网络流的高铁车站站改期间行车组织优化研究

Study on Optimization of Train Operation Organization during theReconstruction of High-speed Railway Hub Station Basedon Time Series Multi-commodity Network Flow Mode

将高铁车站站场基本布局抽象为站场网络图,借助时间序列多品种网络流模式进行行车组织优化。基于行车进路规则构建生成进路数据集算法,将既有行车方案与构建的站改期间站场网络图融合,推断既有行车方案中到发线是否可用以及是否存在可用进路,据此将列车分为无需优化列车和需要优化列车两类。针对无需优化列车,构建计算公式和算法,推算到发线占用时间序列和咽喉占用时间序列。针对需要优化列车,构建算法并依据到发线和咽喉的占用时间序列,推断是否存在可用到发线以及是否同时存在可用进路,将需要优化列车分为可以优化列车和无法优化列车。通过构建可行股道推断算法以及可行股道进出站进路可用推断算法,形成行车组织优化方案。

The basic layout of high-speed railway station yard is abstracted into station and field network diagram,and the time series multi-commodity network flow model is used to optimize the driving organization.The algorithm of generating approach data set is constructed based on the rules of approach,and the existing train scheme is integrated with the station network diagram constructed during the station reform period,so as to infer whether the arrival and departure lines in the existing train scheme are available or not and whether there is an available approach or not,and then the trains are classified into two categories of trains that do not need to be optimized and those that need to be optimized according to the algorithm.For trains that do not need to be optimized,formulas and algorithms are constructed to derive the occupation time series of the arrival and departure lines and the occupation time series of the throats.For trains that need to be optimized,the algorithm is constructed and based on the occupation time series of arrival and departure lines and throats,it infers whether there are available arrival and departure lines and whether there are available approaches at the same time,so that the trains that need to be optimized are classified into trains that can be optimized and trains that cannot be optimized.By constructing the feasible lane inference algorithm and the feasible lane inlet/outlet approach available inference algorithm,the train organization optimization scheme is formed.