Rui Jiang

A behaviour based cellular automaton model for pedestrian counter flow

This paper presents a behaviour based cellular automaton (CA) model for pedestrian counter flow. The behaviours of active slowing down and lane changing are considered to simulate the evolution of the pedestrian counter flow in a corridor with open boundaries. A concept of a dominant row is introduced to depict the pedestrians lane changing behaviour. The velocity profile has been investigated and two separation times have been studied, which quantify the evolution process of the lane formation. The simulation shows that steady separate lanes can form even in dense conditions. Therefore, the gridlock formation has been remarkably suppressed.

Pedestrian counter flow in discrete space and time: experiment and its implication for CA modelling

ABSTRACT Cellular automaton (CA) approach has been widely used in pedestrian flow studies. However, one problem of CA approach is that gridlock usually occurs in counter flow. One possible reason is that CA models still do not represent correctly cognitive processes of real people. To study whether the gridlock would occur when real people were required to walk in discrete space and time, we performed an experimental study on pedestrian counter flow. Remarkable collaboration behaviours of pedestrians have been observed, which enable the formation of exit rows and prevent the formation of gridlock. Furthermore, we performed a comparative experiment in which pedestrians walked under normal condition. Our studies indicate that to fully describe pedestrian counter flow with CA approach, the remarkable collaboration behaviours of pedestrians need to be considered. Moreover, one might also need to consider other factors such as flexible characteristics of pedestrians, small cell size, and various walk speeds.

Impact of holding umbrella on uni- and bi-directional pedestrian flow: experiments and modeling

ABSTRACT In this paper, the impact of holding an umbrella on the uni- and bi-directional flow has been investigated via experiment and modeling. In experiments, pedestrians walk clockwise/anti-clockwise in a ring-shaped corridor under the normal situation and holding umbrella situation. In the unidirectional flow, the flow rate under the holding umbrella situation decreases from 0.1 to 0.25 ped/s/m when compared to the normal situation. In the holding umbrella situation, the bidirectional flow rate even reduces to 0.2 ped/s/m in contrast to the unidirectional flow. In the bidirectional flow, pedestrians segregate into two opposite moving streams very quickly under the normal situation and have right-walking preference. Under the holding umbrella situation, spontaneous lane formation also occurs. However, pedestrians can easily separate into three or four lanes. Moreover, the merge of lanes is observed, and clockwise/anti-clockwise pedestrians are not always in the inner/outer lane. The simulation results by an improved social force model considering the umbrella size are in agreement with the experimental ones.

Flowrate behavior and clustering of self-driven robots in a channel*

In this paper, the collective motion of self-driven robots is studied experimentally and theoretically. In the channel, the flowrate of robots increases with the density linearly, even if the density of the robots tends to 1.0. There is no abrupt drop in the flowrate, similar to the collective motion of ants. We find that the robots will adjust their velocities by a serial of tiny collisions. The speed-adjustment will affect both robots involved in the collision, and will help to maintain a nearly uniform velocity for the robots. As a result, the flowrate drop will disappear. In the motion, the robots neither gather together nor scatter completely. Instead, they form some clusters to move together. These clusters are not stable during the moving process, but their sizes follow a power-law-alike distribution. We propose a theoretical model to simulate this collective motion process, which can reproduce these behaviors well. Analytic results about the flowrate behavior are also consistent with experiments.

A cellular automaton model reproducing realistic propagation speed of downstream front of the moving synchronized pattern

ABSTRACT The moving synchronized pattern (MSP) is an important traffic pattern that can emerge when traffic breakdown occurs. However, up to now most models cannot reproduce a realistic propagation speed of downstream front of the MSP, which significantly weakens their applications in traffic breakdown prediction and control. In this paper, a new brake light cellular automaton model is proposed, which assumes that: (i) the drivers would be sensitive to the brake light only when their speeds are larger than a critical speed; (ii) the anticipated deceleration of a preceding vehicle increases with the increase of the speed of the following vehicle. Simulation analysis shows that the new model can depict traffic breakdown and related synchronized traffic flow patterns. Importantly, it can realistically reproduce the propagation speed of downstream front of the MSP. Finally, the new model is calibrated and validated by NGSIM detector data.

Traffic Dynamics Based on Local Routing Strategy in a Weighted Scale-Free Network

In this paper, the traffic flow on weighted scale-free networks is investigated based on local routing strategy using link weights: (P_{lrightarrow i}=frac{w_{li}^{alpha}}{sum_{j}w_{lj}^{alpha}}) . The capacity of links is controlled by max(β w lj ,1). It is shown by simulations that two critical threshold β c1 and β c2 exist. When β>β c1, both the network capacity and the corresponding α c value remain unchanged. When β c1>β>β c2, the network capacity decreases and the critical value of α c increases with the decrease of β. When β<β c2, α c decreases with the decrease of β. The behaviour can be explained by investigating the average number of packets on nodes and delivered through links.