ACP-based Passenger Behavior Analyses And Emergency Evacuation In Urban Rail Transit Stations

As a resource-intensive utilization,environmental protection,comfortable,fast,on time,and low energy consumption large-capacity transport service mode,urban rail transit has been a main transportation for many large and medium-sized cities in the world,which plays a critical role in meeting the masses’travel demand,optimizing urban layout,alleviating traffic jam and so on.As a passenger distribution center,the urban rail transit station(URTS)has the characteristics of complex internal structure,closed space,and poor connectivity with the outside world.Once a large passenger flow,fires,terrorist attacks,and other emergencies occur,it is difficult to predict passenger behaviors and evacuate and rescue passengers from hazardous areas.It is also easy to cause mass casualties.The research of passenger evacuation in URTS has become an important area of production safety risk prevention and control.The measures of understanding the movement behaviors and dynamical characteristics of passengers in normal and emergency situations,formulating reasonable passenger guidance and/or evacuation strategies,and making the effective guidance and control of passengers can significantly reduce the probability of incidents at URTS,reduce or even avoid passenger casualties,and finally ensure the safety of passengers.Based on the ACP method(Artificial Systems,Computational Experiments,Parallel Execution)and the modeling and analysis methods in the field of pedestrian and evacuation dynamics,this paper focuses on the analyses of passenger behaviors and emergency evacuation in URTS.The main contents of this paper can be summarized as follows:Firstly,a passenger-oriented framework of the artificial system for URTS is proposed.The data of stations and passengers are collected through field investigations.The basic attributes,distribution,movement and waiting behaviors,as well as characteristics of passengers,are statistically analyzed.The investigation results are used to model the artificial system,validate the models and design the scenarios of computational experiments.Two kinds of models are built based on fuzzy logic and social forces,respectively,which are further validated through simulation experiments and real data.Finally,the passenger-oriented artificial URTS system is improved.Secondly,the characteristics of passenger evacuation and bidirectional passenger flow under normal conditions are studied with the scenarios of station hall and passageway based on the artificial system of URTS.The movement behaviors of bidirectional passengers under the influence of preference behavior and the distribution of passengers’movement speed and direction in bidirectional passenger flow are studied by computational experiments.The dynamical characteristics of waiting passengers and passing passengers are analyzed by considering the behaviors and modes of waiting passengers.The effects of different waiting modes on the time delay,movement speed,and direction of passing passengers are analyzed.Finally,the movement behavior and dynamic characteristics of passengers are studied in case of an emergency.The effects of assailants on passengers’evacuation behaviors are considered.The effects of models and scenario parameters like desired speed,exit width,and duration time on the evacuation time and the number of attacked passengers are analyzed quantitatively.It reveals the characteristics and evolution mechanism of passenger movement behaviors under normal and emergency conditions.Thirdly,a passenger’s perceptual behavior model for emergency signs under a limited visual field is proposed.The dynamical behavior of passengers under the influence of emergency signs is analyzed and the passenger’s evacuation model is established.The proposed model i,s validated and verified by simulations and experiments.Finally,the evacuation characteristics of passengers under the condition of limited visual field caused by smog are studied in the hall and platform scenarios based on the artificial URTS.At the same time,three sign distribution schemes,i.e.,maximal cover,uniform,and random,are proposed based on the platform scenario of URTS.The influence of the number and distribution of emergency signs on the efficiency of passenger evacuation is studied quantitatively by computational experiments.It provides theoretical support for designing a reasonable sign distribution scheme.Fourthly,the leader-based crowd evacuation is presented as an optimization problem.A strict mathematical description is formulated for the first time,and the optimal evacuation strategy is proposed by combining numerical and simulation methods.Firstly,the concept of guidance demand is defined according to the location of passengers and their familiarity with the environment of the station.A hybrid two-layer model is proposed to optimize the number,initial locations and routes of leaders in the evacuation process by considering the passengers’ guidance demands and coordination mechanism among leaders.The social force model and its improvements are used to describe the dynamical behaviors of evacuees and leaders taking into account interactions between passengers and leaders as well as the surrounding environments.The number and locations of leaders are determined by the improved maximum coverage model,and the leader’s evacuation route is determined by the heuristic algorithm based on a co-simulation approach.Finally,three leader assignment strategies with a different number,locations,and routes of leaders are designed as comparison by taking a URTS in Beijing as a typical example.The performance of the comparison strategies and the optimal strategy are analyzed and evaluated through computational experiments based on the artificial URTS system.It provides an effective method to solve the problem of the influence of leader’s distributions and path planning on passenger evacuation ability on the URTS.Fifthly,the framework and function modules of the parallel URTS system are proposed based on the ACP method.The framework of parallel control and management system of URTS and the artificial URTS system,computational experiments and parallel execution modules which constitute the framework are introduced respectively.The parallel execution mechanism of real system and the artificial system is designed.Finally,a software and hardware platform for parallel control and management of a URTS is developed.Taking an RTS in Beijing as an example,this chapter describes the implemen,tation,evaluation and optimization process of computational experiments based on the artificial system as well as the parallel execution process of the real and artificial systems,and verifies the effectiveness of proposed models and methods for solving the management and emergency evacuation of passenger in URTS.