Research On Evaluation Of Passenger Flow Distribution Service Level Of Three-line Transfer Station

With the rapid development of Chinese urban rail transit network,the series connection between urban rail transit lines and networks will become closer,and multi-line transfer stations will become an important carrier for connecting rail transit.As a typical representative of multi-line transfer stations,three-line transfer station has more practical significance for the study of three-line transfer station.At present,scholars have few researches on three-line transfer stations,and urgently need to carry out in-depth research on distribution of passenger flow on three-line transfer.The multi-line transfer station will cause a surge in passenger traffic at the transfer station,which is likely to cause congestion in the station hall.At the same time,the spatial structure of multi-line transfer station is complicated,and the passengers choose transfer form and the diversion ratio are also different.The distribution will face a test.This thesis summarizes domestic and foreign research literature,and establishes a service level indicator system of three-line transfer station passenger flow distribution.Then uses Anylogic simulation software to simulate and analyze the index data of each station layer of the TCM hospital station,and combines the T-S fuzzy neural network algorithm to the TCM hospital station carry out dynamic prediction and evaluation about service level of passenger flow at each platform level.The main research contents are as follows:(1)Combine with the design principles of three-line transfer station to describe and analyze the transfer pattern.According to inbound passenger flow,outbound passenger flow and passenger flow to explain the form of passenger flow.Combine passenger microscopic behavior and spatial and temporal distribution to classify and describe passenger flow characteristics Based on the effect of passenger flow,the distribution law of passenger flow at three-line transfer station is described.(2)Based on fuzzy rules and learning rules,this thesis gives a theoretical overview of fuzzy systems and neural networks.According to its advantages,combination methods,and structural classification,this chapter analyze fuzzy neural network algorithm.Combined with classification of fuzzy logic system,the T-S fuzzy neural network algorithm is selected as the research method for dynamic prediction and evaluation about passenger flow distribution service level of three-line transfer station.And give a detailed overview of structural characteristics of algorithm and parameter learning method.(3)Combining location information and its internal environment characteristics of the TCM hospital station,this thesis describes the basis for selecting the TCM hospital station as the research object of three-line transfer.According to characteristics of field scene structure about the TCM hospital station,and referring to layout of the facility environment about station hall of the TCM hospital station,the scene modeling of the TCM hospital station is performed.(4)Through field research on the TCM hospital station,combined with statistics of survey,Agent parameter variables are set for passenger distribution characteristics,passenger flow characteristics,transfer characteristics,service characteristics and train characteristics,and combined scene and logic flow to simulate and model passenger flow distribution in the TCM hospital station.(5)Establish a three-line transfer station passenger flow distribution service level index system to simulate and analyze passenger flow distribution service level index data of each platform layer of the TCM hospital station,and combined with the T-S fuzzy neural network algorithm,the dynamic prediction and evaluation about the passenger flow distribution service level at each platform of the TCM hospital station.Based on the simulation data and evaluation results,the corresponding technical measures are proposed for problems existing at the station level.