| The research on pedestrian dynamics can not only provide reliable basis for pedestrian traffic facilities design, but also make pedestrian evacuation plan in emergency. Human’s social attribute makes the phenomenon that the pedestrians walk in group so common that it should not be ignored in the field of pedestrian dynamics research. In order to deeply understand group’s movement characteristics and its effect on pedestrian traffic system, in this paper, we studied the based human movement behavior by using a combination of experiments and simulation methods, meantime focus group behavior, and finally built an evacuation model coupling with the effects of group.To explore pedestrians’ psychology and movement characteristics when walking in groups, a questionnaire survey and a field observation were performed. Pedestrians’psychology characteristics were analyzed from five aspects, including group size, interpersonal distance, spatial relationship among group members, speed adjustment of group members, information transmission and acid action among group members.The survey results showed that group members had a closer interpersonal distance, faster information transmission and plenty of acid action. The field observation offered further insight into the movement similarity among group members and the diversity of different group types. Comparing with single pedestrian’s movement, pedestrian walking in group decreased pedestrian’s average speed.With regard to group movement similarity, it was found that group member’s speed,step frequency and offset angle hada good similarity.The effect of gender on single pedestrian’s speed was more obvious than on group’s speed, which was opposite to the situation of pedestrian’s step frequency. Owing to different social relationships, step frequency difference, offset angle difference and speed difference between the couple groups and the common-friend groups were relatively large.For a further detailed understanding of the interaction between groups and environment, four controlled experiments of group movement were carried out, including group movement in the channels, group bypassing obstacle, bi-directional group movement and group crossing movement. Through the analysis of group movement in the channels, the pattern of people walking side by side was the most favorable movement pattern for small groups if motion space was enough. With the increase of the channel width, groups’trajectory, groups’speed, groups’ formation and the diversity among group members would change. From the analysis of the bypassing obstacle experiment, group members’decision time and decision distance when they were faced with obstacle were got. Meantime the characteristic distance between group member and obstacle, and the characteristic distance between group member and wall were also obtained. Although group sizes were different, the above four characteristic parameters were roughly similar. The decision time was about1.5s, and the decision distance was about2m. In addition, the characteristic distance between pedestrian and obstacle was about0.5m, and the characteristic distance between pedestrian and wall was about0.3m. Through the research on bi-directional group movement, group members’decision time and decision distance were obtained. It is the time1.3s-2s before meeting and the place1.8m-3m from the meeting point that they decided to adopt measures to avoid potential collision. After meeting, group members needed0.37s-1s and0.5m-1.2m distance to restore to their original state of motion before meeting. Relative to large group, small groups’decision time and decision distance were a little larger, however, their recovery time and recovery distance were a little smaller, which were related with the flexibility and competitiveness of group movement. Through the analysis of group crossing experiment, it found several kinds of pedestrian conflict resolution mode, such as accelerating forward bypass, decelerating backward bypass, stopping to wait for bypassing and keeping low speed before meeting. The sequence of group achieving the crossing boundary had a significant influence on the advantage of group movement. Moreover, when the group sizes from the two directions were not the same, small groups were much easier to take place lateral offset. However, before and after the cross region, small groups’ speeds had a little change. With the increase of group size, the decreases of group members’speed were quite obvious.We put forward a group floor field to modify the traditional floor field model in order to explore the effect of groups on pedestrian dynamics. Referring to the method of setting static floor field, in this paper two methods of generating the group floor field were developed, i.e. group center generation (Method1) and group leader generation (Method2). Then the proposed group model was applied to simulate the bi-directional flow in a corridor. Comparing with the crowd without groups, the pattern that pedestrians walked in group more easily causes jamming at the same crowd density. Moreover, with the increase of the group size, the congestion density of the crowd showed the decrease trend. Comparing the results obtained from the two methods of generating group floor field, it seemed that method2could make the group model performance better than method1. In addition, we explored the effect of asymmetric interaction rate and people’s moving preference on group movement. It was obtained that the crowd with large group size easily suffered the impact of asymmetric interaction rate. However, the symmetrical injection rate of pedestrians made the crowd easier to congest.As the key component of the evacuation system, scientifically and reasonably describing the exit’s influence can guarantee movement model accuracy of predicting evacuation results. Based on the floor field model, we put forward an idea of "virtual reference point" and proposed a new method of building static floor field. Through plenty of simulations, then we conducted regression analysis to confirm the relationship between the virtual reference point and the exit width. Applying the three traditional methods and the new method to simulation the same scenario, the density profiles, velocity profiles and evacuation time were obtained. And the simulation results showed that the result of new method had a greater advantage than traditional methods. In addition, based on the results, we put forward a primary scheme to set the evacuation guide boards for wide exit. |
PDF Full Text Request