VARIANCE ANALYSIS OF MARITIME PASSENGER CLUSTERS TRAVEL SPEEDS WITH REGARDS TO THE HEAD'S WALKING SPEED AND HEAD COUNTS

摘要

The purpose of this study is to determine the movement speed of maritime passenger clusters during evacuations, based on an experimental Korean maritime data set. We primarily focus on how much two factors affect overall group travel time: the walking speed of the head of a small passenger group and the number of people in the group. We use an egress model (maritimeEXODUS) to predict maritime passengers' egress behaviour, with regard to aspects such as travel time, density, flow rate, crowd congestion and clustering pattern. Many researchers have studied and analysed occupants' physical characteristics using full-scale experimental data sets, in some cases addressing occupant groups or crowd behaviours. For example, Zeitz has analysed crowd behaviours in mass gatherings, while Klüpfel studied crowd movement by engaging in egress simulation modelling. Other studies have also dealt with psychological aspects. Crowd movement is usually not based on the individual characteristics of crowd members, such as their free walking speed, gender, or personality. Therefore, it is not easy to study factors that affect group or crowd movement behaviours, and it is impossible to do so in uncontrolled observation studies. The analysis of detailed content is also difficult. As such, differences between factors are generally quantified and then used to develop crowd models for controlled experiments. This can help engineers to develop egress simulation models. In general, it is difficult to observe the members of a crowd and their movements. In evacuation situations, more than ten people might move toward the same destination together along a corridor. The number of members of a crowd is closely related to the crowd's velocity. For this study, we conducted an experiment assuming an egress situation on a vessel. The crew of the vessel acted as evacuation guides for 40 passengers, with one person leading an evacuation group as its head. In order to understand the relationship between the walking speed of the head of the group and the rest of its members, we invited 40 passengers to board a ship. We first analysed the head of the group's impact on its travel speed, and then the relationship between individual free walking speeds and passenger group travel speeds. Under normal-controlled conditions, a number of controlled factors were selected in order to measure their influence on the results. The experimental venue was a U-shaped corridor that was 63 metres long and 1.2 metres width. The participants' free walking speeds were measured under experimental conditions. Before the experiment, cameras were installed on the ceiling to record all behaviours. The experiment consisted of 9 sub-experiments, involving different numbers of passengers in groups. For comparison, the walking speeds of the 40 participants were measured in all cases, and individual free walking and group movement speeds were also compared.