Study On Evaluation Model Of Passenger Flow Congestion Level In Urban Rail Stations

With the increased passenger flow and sharing rate, urban rail stations are experiencing greatly impacted, which brings many safety accidents. It has important theoretical and practical significance for safety management, design and maintenance of facilities, reducing accident rate to analyze passenger flow congestion bottleneck and evaluate congestion degree.Firstly, the paper analyzes factors affecting, distribution point, arrive and service time principle of passenger flow. Through K-S test of survey data, which concluded that arriving passenger flow obeys the Poisson distribution, service time interval obeys Negative Exponential distribution in gates or Normal distribution in the others. Then, queuing models have been identified as M/M/C or M/G/C type, meanwhile, the model of passenger flow arrival rate, service rate, queuing length, waiting time all are established.Secondly, based on the comparative analysis of passenger flow congestion degree evaluation index, queuing length and waiting time are selected. Then getting that queuing length and waiting time are divided into A, B, C, D, E, F six grades, also each level division standard, by using K-Means clustering analysis method to analyze the sample data. A type of congestion point’ congestion degree can be get by arithmetic average, then the overall congestion classification also be get by weighted average of each type of congestion degree.Finally, it is concluded that congestion classification of every congestion point and the whole station by using the evaluation model at zoo station. Moreover, some improved suggestions have been proposed and improved results have been evaluated for congestion heavy point.