Research On Identification And Resolution Method Of Dynamic Bottleneck Of Railcar Flow Congestion In Railway Network

Under abnormal operation conditions,the operation of some stations or sections is interrupted,then the railcar flow passing through this station or section will be transferred to other stations.Afterwards it will lead to a dynamic bottleneck of railcar flow congestion in other stations,resulting in the phenomenon of railcar flow backlog.The phenomenon of railcar flow backlog has become an important problem in the daily operation of railway freight transportation.If the dynamic bottleneck of railcar flow congestion is not identified in time,the railcar flow backlog will spread to the upstream station and even the whole network,forming a large-scale blocking area that is difficult to control and seriously reducing the transport efficiency of the railway system.It is an important task of railway emergency management to identify and control the dynamic bottleneck of railcar flow congestion caused by emergencies.However,dynamic bottleneck of railcar flow congestion identification under abnormal operation conditions is subject to dynamic service network which is affected by many factors,such as network capacity,train diagram and transportation plan.Therefore,dynamic bottleneck of railcar flow congestion identification and resolution is difficult to some extent.In this paper,the concept of dynamic bottleneck node is defined.The main research contents are two aspects: one is the theoretical research of dynamic bottleneck node identification and resolution method,the other is the software development of prototype system.The theory and software can be applied to railway emergency dispatching management under abnormal conditions,providing basis for dispatching personnel to carry out transportation adjustment.Firstly,it is pointed out that the key problem of dynamic bottleneck node identification and resolution is dynamic railcar flow organization which needs to consider the the operation process of railcar flow.Based on the operation process of railcar flow,a continuous time-space network construction method under the train diagram is proposed to represent the dynamic transportation process of railcar flow.On the basis of the constructed time-space network,improved A* algorithm is used to generate the feasible path set of each railcar flow that lays a foundation for the assignment optimization model of dynamic railcar flow.Then,a dynamic optimization model of railcar flow organization under abnormal operation conditions is constructed to simulate the time-space distribution of railcar flow under abnormal operation conditions.On this basis,a dynamic bottleneck node identification algorithm is proposed to obtain the dynamic bottleneck node,and a dynamic bottleneck node resolution model based on the transportation plan adjustment strategy is constructed to obtain the dynamic bottleneck solution.Then based on C# language,this method is realized by combining CPLEX and Arc GIS Engine components.Thus a dynamic bottleneck node identification and resolution prototype system software is constructed,so as to achieve human-computer interaction and data visualization.In addition,the software can visually show the dynamic bottleneck node and provide the corresponding solution of the bottleneck intelligently that is conducive for the scheduler to timely grasp the distribution of railcar flow and make transportation adjustment plan in advance.Finally,taking a railway bureau as an example,the dynamic bottleneck node identification and resolution method is applied to a practical case.The results of the case show that :(1)the method of dynamic bottleneck identification is feasible;(2)abnormal operation conditions can significantly increase the overall loss of the transportation system;(3)the dynamic bottleneck node resolution method can be used to get the corresponding solution from the quantitative point of view.