Siuming Lo

Experimental Study on the Influence of Step Phase in Pedestrian Movement

There is a fundamental relationship between the headway and velocity in pedestrian dynamics; this is of great importance in building a pedestrian movement model. However, it is found that even if experiment conditions are strictly controlled, the velocity of pedestrians with the same headway varies considerably. This implies that states of pedestrians should be considered thoroughly, and its influence to macroscopic parameters needs to be quantified. A single-file pedestrian movement experiment is carried out to analyse the microscopic moving characteristics of pedestrians. The trajectories of participants in a straight passageway are extracted through an image processing method based on a mean-shift algorithm. The transverse swing of trajectories is correlated with the phase of pedestrians’ steps period. The stride length and frequency of pedestrians are collected from the trajectories. Besides, the synchronisation of the lock-step group varies during the movement. The findings can give the stop-and-go phenomenon a new microscopic explanation. This study can be used to evaluate pedestrian microscopic movement and the development of evacuation models.

Collection of passenger flow data and development of passenger flow maps in metro stations

Safety and efficiency have become two of the most important problems for metro system operations because of the ever growing patronage size. Collecting passenger flow data about flow, route choice and various other behaviors is a fundamental task towards understanding the crowd and controlling it. Static or location based passenger flow information, such as the flow through a particular escalator, can easily be counted. However, some other regional information which can be used to develop the “passenger flow map” of the whole station, such as passengers walking path and the distribution of passengers on each path, cannot be easily acquired due to the difficulty in tracking individual passenger. This paper presents and compares several basic methods for collecting crowd flow data under station normal service condition and emergency evacuation drill exercise scenario. For normal service condition, the method aims at collecting passenger flow data of a whole metro station, analyzing the passenger travel patterns and assessing the crowding levels. The method is featured as a data synthetic procedure where the most reliable ticket gates data is used as the baseline and then distributed to each route forward and backward according to the proportions of passengers choosing each route. For the emergency evacuation drill scenario, two types of data sources including video camera records and RFID system are adopted. The different features of these two types of data sources, their advantages and disadvantages, the possible problems that may be encountered as well as the tricks that have been proved useful for solving those problems are introduced and discussed.

A cognitive pedestrian behavior model for exploratory navigation: Visibility graph based heuristics approach

Abstract Navigation is a complex issue in simulating pedestrian dynamics. Many existing studies have investigated how pedestrians navigate from an origin to a specific destination, i.e., traveling to familiar or novel destinations. However, pedestrians’ navigation through built environment without specific destinations still remains as an open issue. This exploratory navigation usually involves more spatial cognitive behaviors and perceptual considerations. To represent realistic route choice and natural movement of theses pedestrians, this paper presents a cognitive pedestrian behavior model with a focus on the space visibility and individual characteristics. Then the sensitivity of model parameters is analyzed, indicating that regional target, memory of visiting records and individual crowd tendency can have varied influences on pedestrians’ navigation decisions and movement patterns. As a case study, the proposed model is implemented using the pedestrian movement data collected from an observational study of Tate Gallery Museum. The simulation results show good consistency with the actual data at the aggregate level and those obtained individually, indicating that our proposed model is credible and can benefit real applications such as master planning of a museum.