| In recent years,with the growth of urbanization,an increasing number of people have migrated to urban areas;thus,the urban population has increased sharply.Additionally,more emergency accidents have occurred in places where crowds were gathered,such as the metro.Once the metro accident occurred in tunnel,the passengers have to evacuate through the 0.6 m wide evacuation platform with single-file way.In the literature,research of pedestrian evacuation in an emergency condition has been rare due to ethics;thus,non-human biological systems can be adopted to investigate the evacuation behavior in emergency conditions.In this thesis,the simplest form of motion,single-file movement,was investigated by using ants as research subject in nonhuman biological systems,in which the ants moved under natural and high temperature stimulus.The comparison of ant behavior in different conditions can provide insights to human evacuation under emergency conditions.The second chapter describes the nest relocation experiment of ants and studies the single-file movement of ants in normal conditions.In this chapter,the fundamental diagram of ants moving using single-file movement was investigated by controlled experiments and modeling.From single-file movement experiments,ants prefer to follow the ant in front of them,and no overtaking behavior was observed.We also observed that the critical density in ant traffic,0.69,namely,the density corresponding to the maximum flow,was larger than vehicle and pedestrian traffic,0.5.The flow-density curve presented as an asymmetrical arch was different than those of pedestrian and vehicle traffic flow.To obtain insights into this phenomenon,we introduced a multi-grid NaSch(MG-NS)model,based on a multi-grid cellular automata model and traditional NaSch(NS)model to understand single-file ant trails.The multi-grid method is a convenient mean to refine time and space;therefore,one agent(ant)can occupy multiple grid sites,and its movement can be calculated in detail.Based on the model,we obtained the theoretical relationship between flow and density with varying delay time tj.The result was different from that of the traditional NS model and agreed well with the MG-NS model simulation results.In addition,the MG-NS model can match the result of the ants experiment well with tj=0.25(n=4).The shorter the delay time tj(larger n)was and the larger the flow,speed,and critical density were.Through stability analysis,the shorter delay time tj means agents react faster and can keep the flow in a steady state easier.Additionally,in the traditional NS model,because the size of the grid equals agent size,the delay time is always equal to or greater than one time step.However,for the agent with a sensitive response,the traditional NS model will be no more applicable.The proposed MG-NS model can overcome the disadvantages and,thus,is an improvement compared with the traditional model,thus,it can be used to reveal detailed mechanisms of traffic.In the third chapter,we further investigated the behavior of ants(Camponotus japonicus)in a single-file movement pattern under high temperature conditions.We focused on the motion of ants in a narrow channel that was 6 mm wide and 100 mm long.First,ants were placed into a chamber;then,the ants entered and passed through the narrow channel rapidly,stimulated by the high temperature.We observed that the speeds at the entrance and exit of the narrow channel were slower than that in the intermediate section.Under the stimulation of the high temperature,some ants moved in an abnormal manner,namely,moving backward through the narrow channel.Different from pedestrians,there was no significant relationship between adjacent ants at speed.Additionally,in the experiment,touching behavior was observed.When the preceding ants were touched,they increased their speed,whereas the following ants might decelerate,stop,or move backward.We found that touching behavior in ant traffic has a similar function to the effect of honking in vehicle traffic.Thus,touching behavior resulted in a more efficient evacuation in multiple ant experiments compared with single ant experiments.In fire accident fulling with smoke,the crawling posture is adopted for reducing the damage of toxic smoke.In chapter four,we conducted the single-file movement experiment of students moving with knee and hand crawling to simulate the occupant evacuation in fire and smoke environment.Thus,the knee and hand crawling motion was investigated mainly.Three groups of students with different ages,namely,elementary school students,middle school students,and graduate students,participated in this experiment.Notably,for upright walking and knee and hand crawling,the average speed of elementary and middle school students in individual motion was greater than that for graduate students.The possible reasons for this finding were related to obeying the experimental requirements,the perception of safety,and the limitation of the height of graduate students.Compared with the upright walking speed,the speed of knee and hand crawling was obviously slower.Thus,evacuation efficiency may decrease when evacuees escape from a building that is on fire occurred and filled with smoke.In the single-file movement of knee and hand crawling,middle school students had a higher speed in the free regime than elementary school students.Additionally,the comparison of adaptation time between elementary and middle school students demonstrated that middle school students had a shorter adaptation time and a stronger reaction to a short distance headway.Results of this study should be considered in evacuation models to improve the simulation of human evacuation behavior in an environment where there is fire and smoke.Combining the single-file movement of upright walking,we compare the similarities and differences between ants and pedestrians.The study exhibits the difference in the behavior of ants under different conditions,which can provide insights in to pedestrian evacuation in emergency conditions.Additionally,topics for further research are considered. |
PDF Full Text Request