Weiguo Song

New insights into turbulent pedestrian movement pattern in crowd-quakes

Video recordings right before the Love Parade disaster have been quantitatively analyzed to explore the bursts of unusual crowd movement patterns, crowd-quakes. The pedestrian movement pattern in this incident was special for the reason that it happened in a congested counter flow scenario, where stopped pedestrians were involved. No one was believed to have pushed others intentionally at the beginning, however, under this situation, the body contacts among the pedestrians still induced a force spread, which then led to velocity fluctuation. As indicated by the individual velocity-related features, the densely crowded pedestrian movement displayed turbulent flow features. Further analyzing the overall flow field, we also found that the pedestrian flow field shared typical patterns with turbulent fluid flow. As a result of the turbulent state, different clusters of pedestrians displayed different velocity features. Thus crowd pressure which took into account the velocity and density information was proved to be a good indicator of crowd disasters. Based on these essential features of pedestrian crowd-quakes, a minimal model, i.e., a pedestrian crowd-quake model, was established. Effects including pedestrian gait, stress conservation level and personal intention to escape were explored.

Defining static floor field of evacuation model in large exit scenario

Abstract The floor field model has been widely used to study pedestrian movement. But the traditional methods of setting the static floor field would lead to highly insufficient utilization of the exit region when the exit width is very large. In order to solve the problem and to study the utilization of wide exit, in this paper we put forward an idea of “virtual reference point” and propose a new method of building static floor field. A virtual reference point can be regarded as a point sink of the floor field, it has minimum field value. The position of the virtual reference point decides the distribution of static floor field values in the model. We further explore the relationship between the virtual reference point position and the exit width using regression analysis. It seems that a proper position of virtual reference point will make the exit be fully used and get high evacuation efficiency. We then analyze how to lead people to fully utilize the exit by changing the configuration of exit guidance, and give a primary scheme.

Artificial neural network approach for modeling the impact of population density and weather parameters on forest fire risk.

The risk of forest fire occurrence is affected by the interactions among forest fuels, weather, human activities, etc. In the present paper, we try to build a method to model and forecast forest fire risk based on artificial neural networks. The data considered include population density and several weather parameters, i.e. average relative humidity, wind velocity and daily sunshine hours. With an interpolation method, these data have been expanded into 1 by 1 km meshes that are calculated according to the standard mesh code system in Japan, where the Japanese territory is divided into a lattice by latitude and longitude. Different parameter combinations and corresponding fire probabilities are computed. The correlations between forest fire probability and population density, and sequentially that between forest fire probability and combinations of population density together with one or several weather parameters are analyzed with three back-propagation neural networks in comparison with polynomial regression investigations. The results indicate that non-linear relationships exist among the influential factors and forest fire probability; artificial neural networks could better capture the non-linearity and give closer results to the test set compared with polynomial regression. The proposed method may be used to investigate and forecast forest fire risk providing there are enough data.

A Two-Dimensional Optimal Velocity Model for Unidirectional Pedestrian Flow Based on Pedestrian's Visual Hindrance Field

Modeling and simulation of pedestrian movement is a feasible and effective way to evaluate evacuation facilities and risk. Inspired by the visual field and movement characteristic of pedestrians, we developed a 2-D continuous model that integrates a self-slowing, local direction-changing mechanism, and visual hindrance information. The model allows for the movement in continuous space and time, only controlled by simple kinematic equations and visual hindrance distribution. In order to get the parameters of the kinematic equations, we conducted controlled experiments, collected empirical data, and obtained velocity-changing and direction-changing relations. We then validate the model by simulating three experimental scenarios, i.e., passage, bottleneck, and classroom evacuation. It is found that some typical phenomena such as the stop-and-go waves in the passage and lane formation in the bottleneck can be reproduced. The obtained fundamental diagram and specific flow agree with classic conclusions and experimental measures very well. It is hoped that the idea of this study may be helpful in promoting the modeling and simulation study of pedestrian flow.

Analyzing pedestrian merging flow on a floor-stair interface using an extended lattice gas model

The floor–stair interface is a critical component of the escape stairs in high-rise buildings, and the characteristics of the merging flow in the floor–stair interface is also significant for investigating the evacuation of high-rise buildings. In this paper, the original lattice gas model is extended by considering inner-side walking preferences, turning behavior and different desired speeds. Then the merging flow in the floor–stair interface is simulated with the extended model. The fundamental diagram of the simulation results has the same tendency as the empirical data and it is found that the merging at the floor–stair interface could retard the motion of pedestrians on the incoming stairs and reduce the motion speed in the case when the corridors are adjacent to the incoming stairs. The effect of two kinds of typical structures of the floor–stair interface on pedestrian merging behavior is investigated. The results show that the evacuation process will be improved when the corridor is connected to the landing opposite to the incoming stairs. The results given in this paper may be helpful in understanding the characteristics of pedestrian merging flow in floor–stair interfaces and in designing building structures.

Multi-grid simulation of counter flow pedestrian dynamics with emotion propagation

Abstract Much insight into pedestrian flow dynamics has been achieved by conducting simulations and experiments. However, the combination of pedestrian movement and psychological influence in discrete models has not been well investigated. Here we employ a multi-grid model to understand pedestrian dynamics in counter flow, coupled with the effect of emotion propagation in a crowd stampede induced by panic. The time evolution of average speed under conditions with and without emotion propagation is obtained. Though emotion propagation increases the desired speed, it easily leads to congestion because of an increase in competition in counter flow. According to the speed–density relationship, a function of the crowd density and percentage of pedestrians in counter flow is employed to identify the flowability of pedestrians in a long channel. Thus, three regimes, i.e., lane formation, the transition stage and clogging, are presented. This is useful in the organization of safe mass events with respect to crowd density. The force distribution and cumulative distribution of the largest force that pedestrians bear are further analyzed to understand a tragic incident related to crushing such as the Cambodian stampede. These results provide an estimation of the number of injuries from the perspective of force. It is hoped that this research will be helpful in crowd management.

An experimental study on four-directional intersecting pedestrian flows

Intersecting pedestrian flows especially multi-directional ones are complicated in dynamics. People will face unavoidable head-on conflicts and obstruct each other. In this paper, controlled experiments of a four-directional intersecting pedestrian flow were conducted. Up to 364 university students took part in the experiments and their trajectories were extracted by a mean-shift algorithm. The global density–velocity relations in the cross area in different scenarios are compared. Moreover, local density–velocity and local density-flow relations in the cross area are investigated. In order to adapt the study of a fundamental diagram for four directional intersecting flows, a new coordinate system based on pedestrian motion is built. The results indicate that the coordinate system is suitable for the analysis of multi-directional flows. The local density–velocity relation seems consistent with previous results obtained from an actual high-density pedestrian flow. At high densities, the average local velocity in the cross area is a bit larger than a previous study. The reason may be due to the density difference between the cross area and the corridors, which can be observed in real life.

Power-law distribution of city fires

This paper presents results of fire distributions in China and Switzerland. The frequency-size (loss) distribution and the rank-size distribution (or Zipf plot) of city fires are plotted to check their fit with a power-law relationship. The results indicate that fire data in both countries have good power-law distributions and the power-law relation is invariant with scale and time, i.e. fires in different places or in different periods will have the same distribution characteristics. Steady power-law distribution corresponds with self-organized criticality (SOC) and is expected to provide practical implications for fire protection.

Behavior of Ants Escaping from a Single-Exit Room

To study the rules of ant behavior and group-formation phenomena, we examined the behaviors of Camponotus japonicus, a species of large ant, in a range of situations. For these experiments, ants were placed inside a rectangular chamber with a single exit that also contained a filter paper soaked in citronella oil, a powerful repellent. The ants formed several groups as they moved toward the exit to escape. We measured the time intervals between individual escapes in six versions of the experiment, each containing an exit of a different width, to quantify the movement of the groups. As the ants exited the chamber, the time intervals between individual escapes changed and the frequency distribution of the time intervals exhibited exponential decay. We also investigated the relationship between the number of ants in a group and the group flow rate.

Forest Fire Smoke Detection Using Back-Propagation Neural Network Based on MODIS Data

Satellite remote sensing provides global observations of the Earth’s surface and provides useful information for monitoring smoke plumes emitted from forest fires. The aim of this study is to automatically separate smoke plumes from the background by analyzing the MODIS data. An identification algorithm was improved based on the spectral analysis among the smoke, cloud and underlying surface. In order to get satisfactory results, a multi-threshold method is used for extracting training sample sets to train back-propagation neural network (BPNN) classification for merging the smoke detection algorithm. The MODIS data from three forest fires were used to develop the algorithm and get parameter values. These fires occurred in (i) China on 16 October 2004, (ii) Northeast Asia on 29 April 2009 and (iii) Russia on 29 July 2010 in different seasons. Then, the data from four other fires were used to validate the algorithm. Results indicated that the algorithm captured both thick smoke and thin dispersed smoke over land, as well as the mixed pixels of smoke over the ocean. These results could provide valuable information concerning forest fire location, fire spreading and so on.