Passenger Behavior Modeling In Metro Station And Emergency Simulation System Development

With the acceleration of urbanization and the improvement of mobility of urban traffic in China,the contradiction between supply and demand of large-scale urban traffic has become increasingly prominent.Urban rail transit,with its large transport capacity,safety and many other advantages,has become the main choice to solve the problem of urban traffic congestion,has been vigorously developed in China.After 53 years of development,subway,as the most important mode of urban rail transit,has reached 7,253.73 kilometers.In spite of this,the congestion problem of urban rail transit system is still increasingly serious,and the high passenger flow of metro in big cities such as Beijing,Shanghai,Guangzhou and Shenzhen has become the norm.Large passenger flow in underground confined space is a great potential safety hazard.Major cities have issued emergency plans related to sudden large passenger flow to ensure operation safety.At the present stage,it is difficult to carry out the emergency drill of large passenger flow in practical operation,so that the operation personnel generally lack the opportunity to practice the emergency handling ability of large passenger flow,unable to master the law of the impact of the emergency handling measures of large passenger flow,and lack the evaluation means of emergency plan.On this paper,the institute of the Ministry of Transport subject subway emergency disposition of large passenger flow simulation,from the perspective of safe operation,the research and development based on Agent and VR subway station passenger emergency pedestrian simulation system,is used to simulate and evaluate large passenger emergency disposal measures pedestrian behavior and the influence of the overall traffic to the station,and provide realistic safety training.This paper mainly studies the following aspects:(1)Analyze the traffic characteristics and pedestrian behavior characteristics of large passenger flow in subway stations.References and field investigation,and analysis the station platform inside the station hall,automatic ticket gate,stairs and escalators,characteristics of pedestrian facilities,including the queuing behavior,counter choice behavior,more specific facilities passenger flow speed and maximum capacity,etc.,based on the platforms of queuing behavior and self-service check-in dynamic local path planning model;(2)Construct a multi-level interactive micro-pedestrian model of subway station based on Agent.The decision-making framework of the four-layer pedestrian model at the task layer,strategic layer,tactical layer and operational layer is designed,and the Agent sensors,detectors,controllers and effectors are constructed based on the concept of Agent perception,judgment,decision and action,and the algorithm is realized by object-oriented programming technology.At the initial moment of the simulation,the task layer assigns the final travel task for the intelligent pedestrians within the station,formulates the travel purpose state transfer rules,arranges the travel sequence of the pedestrians at specific facilities in multiple sections,and realizes the pedestrian travel simulation of all OD chains in the paying area of subway stations.The strategy layer realizes the global path planning by graph search shortest path algorithm,and realizes the dynamic global path planning based on the asynchronous refresh of various plane domain types,and proposes an inflection point optimization algorithm.Tactical layer to realize the local path planning,dynamic build queue list data structure,to research the dynamic counter choice Logit model based on queuing behavior abstract for the agent evaluation function based on the queue list data selection algorithm,to realize the local path planning,the station facilities more dynamic counter and achieving the dynamic collision pedestrian model to predict local path planning and obstacle avoidance.At the operation level,the microscopic travel behavior of pedestrians with specific travel purposes in subway stations is completed,and the queuing list data structure is used to realize the behavior of different travel purposes based on the queuing behavior.Different area types are defined and the travel behavior is associated with area types.On this basis,the model parameters are calibrated and verified.The model is used to simulate the phenomenon of self-organization of pedestrian flow,queuing system and passing capacity of each facility,and the model is verified qualitatively and quantitatively.(3)Develop the simulation system for emergency treatment of large passenger flow in subway stations.Investigate the emergency response plans for large passenger flows in major cities,select typical emergency response measures for large passenger flows and design algorithms to realize their impact on agents,and improve the construction of interactive intervention pedestrian models in subway stations.In a high degree of reduction environment model based on BIM and VR,this pedestrian simulation model algorithm simulation model and emergency measures in the system implementation,emersion the pedestrian behavior,realize the disposal measures of VR interaction and its intervention on virtual pedestrians,built inference system,and to give big passenger flow within the system emergency disposal measures of quantitative and visual evaluation standard,A railway station is taken as an example to analyze the impact of different disposal measures.