视野受限条件下的行人运动实验与模型研究

发布时间：2018-08-03 11:59

【摘要】：随着经济社会的快速发展,各种结构复杂的建筑设施大量涌现,同时伴随着行人出行频率的增加和大型公共活动的增多,在一些公共场所如学校、体育馆、地铁站、火车站等经常出现大规模人群的聚集。在高密度人群环境中,个体运动严重受阻,容易引发拥挤踩踏。当前针对行人疏散的研究绝大多数都集中在行人视野正常、外界能见度良好的情况下,然而当火灾发生时产生的烟尘或其它原因造成电力照明系统崩溃会导致外界能见度下降,此时行人的视野会受到不同程度的影响。所以本文采用实验和模型相结合的方法,对视野受限下行人微观运动和疏散过程进行了研究。为获取行人在视野受限条件下运动的基础数据,我们首先开展了基本的单列行人运动实验。后期基于实验视频和轨迹分析,发现行人在视野受限情况下倾向于寻找边界和跟随前方行人运动;随着人员密度的增加,行人轨迹的侧向摆幅增大;随着能见度的降低,高密度下走停现象开始占据交通主导,并逐渐向中等密度延伸;前向距离与运动速度关系可以分为运动受限和自由运动两个阶段:在运动受限阶段,前向距离与运动速度之间的相关性随着视野受限程度的加重而越来越弱;而在自由运动阶段,不同透光率下行人的自由运动速度都符合高斯分布(vLT=0.3%～N(1.31,0.072),vLT=0.1%～N(0.71,0.092)和vLT=0.0%～N(0.45,0.102));视野受限条件下密度与流量关系大致可以分为三个阶段:自由流阶段、最大流阶段和拥挤流阶段。透光率0.3%、0.1%和0.0%下的最大流量分别为1.3s-1、1.1s-s和0.9s-1左右。为研究视野受限情况下行人的疏散过程,我们分别针对视野部分受限和完全受限两种情况开展了疏散实验和模型模拟。在部分受限的超市疏散实验中,观察到行人的一些典型行为包括跟随行为、帮助行为和寻找依附物行为。随后又设计了调查问卷进一步对受限视野下的行人疏散进行研究,发现问卷结果与实验观察结果存在一定差异。最后构建了考虑视野受限情况下行人典型行为特征的疏散模型,不同场景下模拟结果与实验结果的高度吻合验证了模型的可靠性。针对完全受限的情况,实验观察到行人首先会沿着当前的朝向运动去寻找墙壁;在找到墙后此时行人有选择左手侧方向运动的倾向;之后行人会继续沿墙运动,运动过程中行人会采取不同的策略来解决与他人的冲突,直至最后找到出口离开。此外T检验结果表明行人沿墙运动速度大于未找到墙的运动速度。随后基于实验观察和分析结果,建立了多格子人员疏散模型,模拟重现了实验中行人的典型运动特征。最后通过对比正常视野和视野完全受限下的疏散,发现两者既有相同点也有不同点。相同点是:人员在均匀分布、随机分布和集中分布下,整体的疏散时间差别不大;两种视野情况下的行人疏散时间都随着初始人员密度的增加而上升;增加出口数量能降低人员疏散时间,视野完全受限情况下效果更显著;疏散过程中连续两人通过出口的时间间隔都呈现幂律关系。不同点是:正常视野下行人运动速度更快并且疏散距离更短,故其所需的疏散时间明显要短于视野完全受限下的疏散;增加出口宽度对于正常视野下的疏散十分有利而对于视野完全受限下的疏散几乎没有影响;正常视野条件下疏散中行人大多集中在出口附近,而在视野完全受限下墙的周边是人员高密度区域。通过以上研究我们对视野受限条件下的行人运动和疏散有了更深的认识,接下来将探讨该如何对视野受限环境中的行人进行疏散引导。首先,基于多格子模型我们建立了考虑引导员存在的人员疏散模型。模型中一共设置两种行人类型:引导员和跟随者。其次,针对视野受限情况下的引导员数量、类型、分布、运动速度及引导策略对疏散时间的影响开展了模拟研究。最终模拟结果表明:有标识的引导员在人群中容易被其他行人识别,从而能发挥更大引导作用,其疏散时间较无标识的引导员情况降低了 7%;动态引导中引导员通过自身运动扩大了引导范围,疏散时间较静态引导下减少了 20%;综合考虑人力成本和疏散时间的情况下存在一个相对较优的引导员数量;引导员的均匀分布能覆盖更大的引导面积从而更有利于整体疏散,但这跟实际中其他行人的分布有关;引导员的运动速度约为其他行人速度的75%时,疏散引导效果最优;对比不同的疏散引导策略,发现综合考虑与目标行人的距离和目标行人周围人数策略下的整体疏散时间最短;多个引导员之间的交流合作能避免在选择引导目标时产生冲突,此时整体疏散效率提高了 6%。
[Abstract]:With the rapid development of the economy and society, various complex construction facilities have springing up. At the same time, with the increase of pedestrian travel frequency and the increase of large public activities, large crowds often appear in some public places such as schools, gymnasiums, subway stations and railway stations. In the high density population environment, the individual movement is serious. At present, most of the studies on pedestrian evacuation are concentrated in the case of normal pedestrian vision and good visibility, however, when the smoke or other causes caused by the fire and other causes cause the breakdown of the power lighting system, the visibility of the pedestrians will be reduced, and the vision of pedestrians will be different at this time. In order to obtain the basic data of the pedestrian movement under the limited field of vision, we first carried out a basic single row pedestrian movement experiment. In the later period, we found pedestrians based on the experimental video and the trajectory analysis. With the increase of the density of personnel, the lateral pendulum of the pedestrian path increases with the increase of the density of people. With the decrease of visibility, the phenomenon of high density down stop begins to dominate the traffic and extends to the medium density, and the relationship between the forward distance and the velocity can be divided into limited motion. In the two stage of free movement, the correlation between the forward distance and the speed of motion is becoming weaker and weaker, while in the free movement stage, the free movement speed of pedestrians under different transmittance rates conforms to the Gauss distribution (vLT=0.3% to N (1.31,0.072), vLT=0.1% to N (0.71,0.092) and vLT=0.0%. N (0.45,0.102)); the relationship between density and flow can be roughly divided into three stages: the free flow stage, the maximum flow stage and the congestion flow stage. The transmittance rate is 0.3%, the maximum flow rate under 0.1% and 0% are 1.3s-1,1.1s-s and 0.9s-1 respectively. The evacuation experiment and model simulation were carried out in two kinds of restricted and completely restricted conditions. In some restricted supermarket evacuation experiments, some typical behaviors of pedestrians were observed, including following behavior, helping behavior and finding dependent behavior. Then a questionnaire was designed to further study the pedestrian evacuation under the limited view. There are some differences between the results of the volume and the experimental results. Finally, an evacuation model of pedestrian typical behavior is constructed. The reliability of the model is verified by the high agreement between the simulation results and the experimental results in different scenes. The experimental results show that pedestrians will first follow the current orientation. The pedestrians will move along the left side and move along the left side of the wall when the wall is found. Then the pedestrians will continue to move along the wall. In the course of the movement, the pedestrian will take different strategies to solve the conflict with others until the exit is found. In addition, the T test results show that the speed of pedestrians moving along the wall is greater than that of the unfound wall. Then, based on the experimental observation and analysis results, the multiple grid personnel evacuation model was established, and the typical motion characteristics of the rows in the experiment were simulated. Finally, the same points and different points were found by comparing the evacuation of the normal field of vision and the completely limited field of vision. In the medium distribution, the whole evacuation time is not very different; the evacuation time of pedestrians in the two field of vision increases with the increase of the initial personnel density; the increase of the number of exits can reduce the time of evacuation and the effect is more obvious in the case of completely limited field of vision; the time interval of the evacuation process shows a power law relationship through the time interval of the exit. The difference is that the pedestrian movement is faster and the evacuation distance is shorter in the normal field of vision, so the evacuation time is obviously shorter than that under the complete confinement of the field of vision; the increase of the exit width is very favorable to the evacuation in the normal field of vision and has little effect on the sparse dispersion under the completely limited field of vision; in the normal field of vision evacuation Most of the pedestrians are concentrated in the vicinity of the exit, and the high density area around the wall under the full limit of vision. Through the above study we have a deeper understanding of the pedestrian movement and evacuation under the limited view of vision. Then we will discuss how to conduct evacuation guidance to the pedestrian in the restricted environment. We set up a personnel evacuation model considering the presence of the guide. There are two types of pedestrians in the model: the guide and the followers. Secondly, the effects of the number, type, distribution, velocity and guidance strategy on the evacuation time are simulated. The guide can be easily identified by other pedestrians in the crowd, and thus can play a greater guiding role. The evacuation time has been reduced by 7% than that of the unidentified guide. The guide operator in the dynamic guidance expands the guide range through its own movement, and the evacuation time is reduced by 20% under the static guidance; the manpower cost and the evacuation time are considered comprehensively. There is a relatively superior number of guides; the uniform distribution of the guide can cover a larger guide area and is more conducive to the whole evacuation, but it is related to the distribution of other pedestrians in practice; the speed of the guide is about 75% of the other pedestrian speed, and the effect of the evacuation Guide is optimal; the different evacuation guidance strategies are compared. The shortest evacuation time is taken into consideration of the distance between the target pedestrians and the number of pedestrians around the target, and the communication and cooperation between the multiple guides can avoid the conflict when choosing the guide target, and the overall evacuation efficiency is improved by 6%.
【学位授予单位】：中国科学技术大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：TU998.1

【相似文献】