基于人机社会力模型的人群疏散算法

针对公共场合紧急情况下人群疏散困难和效果有限的问题,提出一种基于人机社会力模型的机器人疏散人群的方法。首先,基于原始社会力模型提出了一种新的人机社会力模型,该模型在原始社会力模型的基础上加入了机器人对人作用的人机作用力;然后,基于人机社会力模型提出一种新的利用机器人疏散人群的方法,该方法在人群疏散场景中加入运动机器人,通过机器人自身的运动,利用人机作用力影响周围行人的运动状态,减小行人之间的压力,从而达到加快人群运动速度、提高人群疏散效率的目的。在室内封闭场景人群逃生、两群行人交错这两种典型的疏散场景中分别进行仿真实验,并将实验结果与未加入机器人的人群疏散结果进行对比分析,实验结果表明,基于人机社会力模型的机器人疏散人群的方法能够明显加快人群的运动,提高人群的疏散效率。     

A social force evacuation model driven by video data

This paper proposes a video data-driven social force model for simulating crowd evacuation. The initialization of pedestrian position, path navigation, and goal selection in the improved social force model was guided by real video data. To initialize pedestrian position and determine path navigation, the distribution of the pedestrians is set according to the real video. We also extracted the trajectories of pedestrian movement from the videos, and these trajectories were stored into a path set to guide the evacuation of pedestrians. Moreover, a fitness function was defined to model the behavior of a pedestrian goal selection. The fitness function could process the evacuation parameters, which were extracted from the video, and consider the degree and distance of exit congestion. Furthermore, we quantified the relationship values among pedestrians, and a new force called “group force” was added to the primary social force model. Pedestrians with close relationship gathered into one group and walked together. To validate the effectiveness of the proposed method, the video data-driven model was applied to simulate campus halls and roads. Simulation results show that the proposed approach is consistent with real-world situations and can assist in analyzing emergency evacuation scenarios.     

基于突发事件影响力传播的双向人流疏散仿真

当突发事件发生时,行人行走行为会因为突发事件本身以及突发事件在人群中的传播而改变。对于突发事件影响下的行人行走特征的研究能够提高人流疏散的效率。针对现有研究中数据获取方式的不足,对真实行人场景视频进行图像处理,提取相关数据后分析了无突发事件时行人一般行走特性。针对突发事件下的行人流,利用k-邻近算法和合力的思想描述了突发事件的影响传播和突发事件下行人流的自组织现象,并由此提出一种新的元胞自动机模型,该模型中的行人元胞会受到正常行走、突发事件、安全标识这三个因素所抽象产生的三个作用力的影响。利用模型对突发情况下的双向人流疏散进行仿真,实验结果表明,当安全标识的距离为0、10、20个元胞时,在小范围行人通道中安全标识分布的距离对人群疏散作用不明显;通过对人群间是否存在影响力的研究发现,疏散的效果主要受到附近行人对突发事件传播的影响;突发事件的影响程度太大或影响范围过小都会引发拥堵,不利于人群的疏散。仿真结果与真实世界中的双向行人流疏散情况基本吻合。     

基于深度Q网络的人群疏散机器人运动规划算法

针对公共场合密集人群在紧急情况下疏散的危险性和效果不理想的问题，提出一种基于深度Q网络（DQN）的人群疏散机器人的运动规划算法。首先通过在原始的社会力模型中加入人机作用力构建出人机社会力模型，从而利用机器人对行人的作用力来影响人群的运动状态；然后基于DQN设计机器人运动规划算法，将原始行人运动状态的图像输入该网络并输出机器人的运动行为，在这个过程中将设计的奖励函数反馈给网络使机器人能够在"环境-行为-奖励"的闭环过程中自主学习；最后经过多次迭代，机器人能够学习在不同初始位置下的最优运动策略，最大限度地提高总疏散人数。在构建的仿真环境里对算法进行训练和评估。实验结果表明，与无机器人的人群疏散算法相比，基于DQN的人群疏散机器人运动规划算法使机器人在三种不同初始位置下将人群疏散效率分别增加了16.41%、10.69%和21.76%，说明该算法能够明显提高单位时间内人群疏散的数量，具有灵活性和有效性。     

考虑分类人群行为的建筑疏散模型研究

人员行为决定了应急疏散时人群的时空分布,是研究疏散动力学的关键。考虑疏散时人员的心理特性与身份状态,将人群分为恐慌人群、易感人群、冷静人群和管理人群四类,基于社会力模型表达各类人群的疏散行为特征,并开展不同情境的疏散动力学过程分析。研究发现行人的恐慌心理具有传播作用,对其他行人的疏散行为有明显的影响,而管理人员的引导作用对疏散有积极影响,当其比例在10%~15%的时候效果显著,且合适的位置更易提高疏散效率;人员的服从水平越大,疏散效率越高。提出的分类人群疏散行为模型能为建筑安全疏散评估与优化提供理论支持。     

An adaptive block-based matching algorithm for crowd motion sequences

For crowd analytics and surveillance systems, motion estimation is an essential first step. Lots of crowd motion estimation algorithms have been presented in the last years comprising pedestrian motion. However, algorithms based on optical flow and background subtraction have numerous limitations such as the complexity of the computation in the presence of high dense crowd and sudden motion changes. Therefore, a novel estimation algorithm is proposed to measure the motion of crowd with less computational complexity and satisfy the real time requirements. The proposed algorithm is based on block-based matching, particle advection, and social force model. By the block-based matching, the motion is estimated in each frame, and the corresponding motion field is created. The particle advection process provides more information about the behavior of pedestrians groups, their tracked trajectories and the boundary of each group segment. Relying on the social force model, a predicted direction of the motion vectors (MV) could be measured significantly. Subsequently, the block-based technique is combined with the social force model to obtain the accurate motion vector with the less possible number of search points. The experimental results indicate that the proposed method achieves high performance by reducing the search points, particularly when many collision situations or obstacles exist in the scenes. Considering the reduction in the computational complexity, the quality of degradation is very low. In all cases, average PSNR degradation of the proposed algorithm is only 0.09.     

改进社会力的双向行人流模型研究

在行人运动和行人行为观察的基础上,结合社会力模型,把行人运动看作一种自驱动的个体在连续空间中移动的过程。为了解决行人间相互作用力的不足,通过引入行人运动感知域,提出一种行人碰撞避免的解决机制,这种机制能较好解决行人间的相互碰撞行为,防止行人间出现大面积的堵塞现象。仿真实验结果显示改进后的模型能再现正常条件下行人流基本图,并很好地再现了行人流“通道形成”自组织现象。     

结合最短路径改进的社会力人群疏散仿真模型

社会力模型广泛应用于人群疏散仿真,针对该模型在仿真过程中存在行人停滞不前、无法通过非凸边形障碍物和疏散路径与行人实际选择的路径不相符等问题,提出了一种社会力改进模型。该模型基于场景中的障碍物生成路径节点,利用这些节点生成无向图,同时考虑了节点的安全系数和拥挤系数对节点通行性的影响生成最短疏散路径。通过改进后的社会力模型进行了多种场景的仿真实验,实验结果显示行人在复杂障碍物场景中能有效绕过障碍物,生成合理的疏散路径,表明该模型有效改善社会力模型,使人群疏散仿真更加真实。     

社会力模型中恐慌度概念的分析和改进

社会力模型是人群仿真领域的经典模型,自1995年提出后就被广为应用和调整。该模型2000年又推出一个改进版,增加了恐慌度概念。虽然目前基于社会力模型的研究已经很多,但是针对恐慌度概念分析的研究尚不多见。因此梳理了社会力模型中关键参数物理意义和恐慌度的概念,用恐慌度的变化来解释人群疏散中的“快即是慢”与“从众行为”现象。指出了社会力模型中由于对行人感知描述得不够细致,导致在一定条件下,人群仿真结果存在个别行人不从众与到达出口却不撤离的问题。针对所提问题,通过增加行人视野范围描述以及重定义行人自运动状态等方式来完善社会力模型。实验结果显示,改进后的模型能够较好地模拟人群的从众现象,并且有助于加深对社会力恐慌度概念的理解。     

Modeling pedestrian flow accounting for collision avoidance during evacuation

Collision avoidance behavior has become an open challenging problem since it is one of critical factors that influence the pedestrian flow dynamics. In this paper, a cellular automaton (CA) model is developed to depict the pedestrian movements when collision avoidance behaviors exist during evacuation. Then, we utilize the proposed model to simulate the influences of the collision avoidance on the pedestrian movements during the evacuation in a classroom with two exits. The numerical results indicate that more collision avoidance behaviors have negative influences on the evacuation efficiency, and that more competition behaviors generate more collisions while have no prominent positive impacts on the evacuation efficiency. Moreover, the evacuation time increases with the decreasing number of aisles in the classroom and the number of collisions increases with the increasing number of parts in the classroom divided by aisles. The above results are helpful to develop effective evacuation strategies and design the internal layouts of buildings.     

Tracking pedestrians using local spatio-temporal motion patterns in extremely crowded scenes

Tracking pedestrians is a vital component of many computer vision applications, including surveillance, scene understanding, and behavior analysis. Videos of crowded scenes present significant challenges to tracking due to the large number of pedestrians and the frequent partial occlusions that they produce. The movement of each pedestrian, however, contributes to the overall crowd motion (i.e., the collective motions of the scene's constituents over the entire video) that exhibits an underlying spatially and temporally varying structured pattern. In this paper, we present a novel Bayesian framework for tracking pedestrians in videos of crowded scenes using a space-time model of the crowd motion. We represent the crowd motion with a collection of hidden Markov models trained on local spatio-temporal motion patterns, i.e., the motion patterns exhibited by pedestrians as they move through local space-time regions of the video. Using this unique representation, we predict the next local spatio-temporal motion pattern a tracked pedestrian will exhibit based on the observed frames of the video. We then use this prediction as a prior for tracking the movement of an individual in videos of extremely crowded scenes. We show that our approach of leveraging the crowd motion enables tracking in videos of complex scenes that present unique difficulty to other approaches.     

Extended route choice model based on available evacuation route set and its application in crowd evacuation simulation

In this study, we propose an extended route choice model based on an available evacuation route set to simulate the selection of pedestrians in selecting an appropriate route during evacuation in emergency situations. In this model, four parameters (i.e., distance to available route, length of available route, level of congestion in available route, and capacity of available exit) affect the route choice of the pedestrian and the evacuation route set. In this study, the evacuation route set is created and optimized by a modified social force model and a route learning method. Experimental results show that the extended model can effectively reproduce crowd behavior in an emergency situation, which can assist in analyzing emergency evacuation scenarios. Moreover, two important conclusions regarding increasing evacuation efficiency show that the proposed model is in line with real-world situations.     

Modeling the impact of interaction on pedestrian group motion

Mobile social robots aimed at interacting with and assisting humans in pedestrian areas need to understand the dynamics of pedestrian social interaction. In this work, we investigate the effect of interaction on pedestrian group motion by defining three motion models to represent (1) interpersonal-distance, (2) relative orientation and (3) absolute difference of velocities; and model them using a dataset of 12000+ pedestrian trajectories recorded in uncontrolled settings. Our contributions include: (i) Demonstrating that interaction has a prominent effect on the empirical distributions of the proposed joint motion attributes, where increasing levels of interaction lead to more regular behavior (ii) Developing analytic motion models of such distributions and reflect the effect of interaction on model parameters (iii) Detecting the social groups in a crowd with almost perfect accuracy utilizing the proposed models, despite the constant flow direction in the environment which causes unrelated pedestrians to move in a correlated way, and thus makes group recognition more difficult (iv) Estimating the level of intensity with considerable rates utilizing the proposed models     

初始分布非均匀的行人流疏散仿真研究

为了有效模拟研究初始分布非均匀的行人流疏散问题,通过定义行人方向模糊可视域,改进了场域元胞自动机模型。模型中,行人的目标位置选择受到方向模糊可视域内行人之间的排斥力和吸引力、出口处行人分布、距可选位置相对距离三种因素共同作用。研究表明,改进模型能够有效地实现初始分布非均匀的行人流在疏散过程中的动态平衡;改进模型不依赖于各因素的影响系数,从而避免了影响系数量化过程的主观性和疏散系统的限制性;在疏散过程中,如果行人保持一个较大的视野半径,疏散系统能够实时提供出口处行人分布状态,就可以有效地提高行人流疏散效率。该研究有助于相关行人流疏散策略和方案的制定。     

CLUST: Simulating Realistic Crowd Behaviour by Mining Pattern from Crowd Videos

In this paper, we present a data‐driven approach to simulate realistic locomotion of virtual pedestrians. We focus on simulating low‐level pedestrians' motion, where a pedestrian's motion is mainly affected by other pedestrians and static obstacles nearby, and the preferred velocities of agents (direction and speed) are obtained from higher level path planning models. Before the simulation, collision avoidance processes (i.e. examples) are extracted from videos to describe how pedestrians avoid collisions, which are then clustered using hierarchical clustering algorithm with a novel distance function to find similar patterns of pedestrians' collision avoidance behaviours. During the simulation, at each time step, the perceived state of each agent is classified into one cluster using a neural network trained before the simulation. A sequence of velocity vectors, representing the agent's future motion, is selected among the examples corresponding to the chosen cluster. The proposed CLUST model is trained and applied to different real‐world datasets to evaluate its generality and effectiveness both qualitatively and quantitatively. The simulation results demonstrate that the proposed model can generate realistic crowd behaviours with comparable computational cost.     

Technology for simulating crowd evacuation behaviors

This paper presents a model for simulating crowd evacuation and investigates three widely recognized problems. For the space continuity problem, this paper presents two computation algorithms: one uses grid space to evaluate the coordinates of the obstacle's bounding box and the other employs the geometry rule to establish individual evacuation routes. For the problem of collision, avoidance, and excess among the individuals, this paper computes the generalized force and friction force and then modifies the direction of march to obtain a speed model based on the crowd density and real time speed. For the exit selection problem, this paper establishes a method of selecting the exits by combining the exit's crowd state with the individuals. Finally, a particle system is used to simulate the behavior of crowd evacuation and produces useful test results.     

Control of swarm behavior in crossing pedestrians based on temporal/spatial frequencies

In the densely-populated urban areas, pedestrian flows often cross each other and congestion is caused. The congestion makes us feel uncomfortable and sometimes leads to pedestrian accidents. To reduce the congestion or the risk of accidents, it is required to control the swarm behavior of pedestrian flows. This paper proposes modeling and controlling method of the crossing pedestrian flows. In the social/urban engineering, it is well known that the swarm behavior with a diagonal stripe pattern emerges in the crossing area of the flows. This is a self-organized phenomenon caused by the local collision avoidance effect of the pedestrians. To control the macroscopic behavior of the flows, we utilize this self-organized phenomenon. Firstly, we propose the continuum model of the crossing pedestrian flows. In the continuum model, the dynamic change of the congestion in the diagonal stripe pattern is simulated as the density. Secondly, the novel control method to improve average flow velocity is proposed based on the model. The proposed method utilizes the dynamic interaction between the diagonal stripe pattern and guides, who are moving in the flows. The authors derive the control algorithm through an analysis on the temporal and spatial frequencies of the crossing flows. The validity is verified with simulations using the continuum model. Moreover, we apply the proposed method to the particle model, assuming the actual pedestrians.     

采用截面流量统计的高密度人群行人计数

利用摄像头实现行人计数在智能视频监控领域有着重要的价值,但是行人互相遮挡、噪声、摄像机透视效果和图像背景等问题影响了人群计数的准确性.针对高密度人群场景的行人计数准确率的问题,提出了基于截面流量统计的行人计数方法,该方法基于梯度运动历史图像检测前景,并用有效运动图像改进了基于特征提取的行人计数方法,结合运动速度提取方法实现了行人计数.实验结果表明,提出的计数方法在高密度人群场景中具有较高的准确率和实时性,是一种针对高密度人群有效的行人计数方法.     

疏散出口附近的人群踩踏仿真分析

人群聚集场所往往存在着较大的安全隐患,当紧急事件发生时,公共场所出口处易发生踩踏事故。针对该现象,设计了一种行人踩踏模型：将行人在踩踏过程中的跌倒情况分为三种,昏迷倒地、推挤倒地、被行人绊倒;在此基础上,设置了行人跌倒的判定规则和行人跌倒后重新站立的条件。行人跌倒的判定规则融合了改进后的社会力模型。改进的社会力模型考虑了跌倒行人对站立行人的影响。仿真过程中对影响人群踩踏的若干因素进行了分析,包括出口宽度、行人平均运动速率、操场塑胶跑道的摩擦系数、心理作用系数,撞击阻碍角度等。仿真结果显示,出口宽度越小,行人平均速率越快,操场塑胶跑道的摩擦力系数越小,越容易发生踩踏事件。心理作用系数和撞击阻碍角度超过一定阈值后,踩踏现象会显著减弱。     

Collective Crowd Formation Transform with Mutual Information–Based Runtime Feedback

This paper introduces a new crowd formation transform approach to achieve visually pleasing group formation transition and control. Its core idea is to transform crowd formation shapes with a least effort pair assignment using the Kuhn–Munkres algorithm, discover clusters of agent subgroups using affinity propagation and Delaunay triangulation algorithms and apply subgroup‐based social force model (SFM) to the agent subgroups to achieve alignment, cohesion and collision avoidance. Meanwhile, mutual information of the dynamic crowd is used to guide agents' movement at runtime. This approach combines both macroscopic (involving least effort position assignment and clustering) and microscopic (involving SFM) controls of the crowd transformation to maximally maintain subgroups' local stability and dynamic collective behaviour, while minimizing the overall effort (i.e. travelling distance) of the agents during the transformation. Through simulation experiments and comparisons, we demonstrate that this approach is efficient and effective to generate visually pleasing and smooth transformations and outperform several existing crowd simulation approaches including reciprocal velocity avoidances, optimal reciprocal collision avoidance and OpenSteer.