Collective Dynamics

Towards a Reference Database for Pedestrian Destination Choice Model Development

Christopher King, Nikolai Bode — Thu, 12 Jan 2023 00:00:00 +0100

The move towards publishing research data openly has led to the formation of reference databases in many fields. The benefits of such resources are numerous, particularly in the development of models. While these exist in research on other aspects of pedestrian behaviour, no reference database is available for modelling pedestrian destination choice, the process by which pedestrians choose where they wish to visit next. This work seeks to construct such a database from the literature. The resulting data obtained are described and potential ways in which they could be used to calibrate a simple pedestrian destination choice model are presented. It contains four datasets that include destination choices for hundreds of pedestrians in settings ranging from university campuses and music festivals to highly structured stated preference surveys. A case study using one of these datasets to calibrate a simple pedestrian destination choice model is provided. These efforts highlight some general issues from creating and using reference data openly. Discussing these issues will hopefully guide the development of reference data and accelerate the development of accurate pedestrian destination choice models that can be applied generally.

Microscopic insights into pedestrian motion through a bottleneck, resolving spatial and temporal variations

Wed, 10 Aug 2022 00:00:00 +0200

The motion of pedestrians is subject to a wide range of influences and exhibits a rich phenomenology. To enable precise measurement of the density and velocity we use an alternative definition using Voronoi diagrams which exhibits smaller fluctuations than the standard definitions. This method permits examination on scales smaller than the pedestrians. We use this method to investigate the spatial and temporal variation of the observables at bottlenecks. Experiments were performed with 180 test subjects and a wide range of bottleneck parameters. The anomalous flow through short bottlenecks and non-stationary states present with narrow bottlenecks are analysed.

Pushing and Non-pushing Forward Motion in Crowds: A Systematic Psychological Observation Method for Rating Individual Behavior in Pedestrian Dynamics

Ezel Üsten, Helena Lügering, Anna Sieben — Fri, 05 Aug 2022 00:00:00 +0200

Pushing behavior impairs people’s sense of well-being in a crowd and represents a significant safety risk. There are nevertheless still a lot of unanswered questions about who behaves how in a crowded situation, and when, where, and why pushing behavior occurs. Beginning from the supposition that a crowd is not thoroughly homogenous and that behavior can change over time, we developed a method to observe and rate forward motion. Based on the guidelines of quantitative content analysis, we came up with four categories: (1) falling behind, (2) just walking, (3) mild pushing, and (4) strong pushing. These categories allow for the classification of the behavior of any person at any time in a video, and thereby the method allows for a comprehensive systematization of individuals’ actions alongside temporal crowd dynamics. The application of this method involves videos of moving crowds including trajectories. The initial results show a very good inter-coder reliability between two trained raters demonstrating the general suitability of the system to describe forward motion in crowds systematically and quantify it for further analysis. In this way, pushing behavior can be better understood and, prospectively, risks better identified. This article offers a comprehensive presentation of this method of observation.

Effects of Driving Style on Energy Consumption and CO2 Emissions

Susana Carreón-Sierra, Alejandro Salcido — Wed, 11 May 2022 00:00:00 +0200

The tractive force developed by energy consumption (EC) in a car engine produces its acceleration and sustains the motion against velocity dependent resistance forces. In internal combustion engines, fuel burning entails pollutant emissions (PE) released into the atmosphere. In vehicular traffic, EC and PE depend on the driving style. This paper assumed that the transition rules in a traffic cellular automata (TCA) represent a driving style, and its effect on EC and PE in TCA is studied. Extending empirical relationships, we proposed models to estimate EC and PE in TCA from the velocity and acceleration distributions, which we obtained by computer simulations for three well-known TCA. The Nagel-Schreckenberg (NS) and Fukui-Ishibashi (FI) models, and a variant (NS+FI) defined by combining the NS and FI rules, were considered. The FI driving style revealed EC and CO2 emission rates dependent on the stochastic delay (p) only for low vehicular densities. We also detected that the larger EC and CO2 emission rates were 45.4 kW and 26.7 g/s with no dependence on p. With NS and NS+FI driving styles, the larger energy consumption and CO2 emission rates occurred for small stochastic delays, 18.4 kW and 6.6 g/s and 61.1kW and 30.2 g/s for p = 0.2. On average, for NS, FI, and NS+FI models (p = 0.2), we obtained energy consumptions of 1.88, 2.60, and 2.76 MJ/km, fuel consumptions of 0.08, 0.12, and 0.13 L/km, and CO2 emissions of 0.158, 0.460, and 0.562 kgCO2/km. Our results agree with those (3.37 MJ/km and 0.235 kgCO2/km) of petrol combustion car engines at 10 km/L. This work may help in designing flow and driving style scenarios to optimize vehicular traffic EC and reduce PE.