Urban Traffic Jam Simulation and Prediction

Abstract

the rapid increase of traffic flow around the world, the development of the city is increasingly becoming saturated. The traffic flow theory that can guide and schedule traffic is needed. Traffic flow model based on cellular automata which is used to describe the freeway traffic flow model has been reaching maturity and applied to the actual traffic simulation system. And the traffic flow model that used to describe the city road network, referred as BML model) still remains in the stage of theoretical research. Many scholars analyze the BML model by analyzing phase-change principle and making theoretical proof. Yet the scholars have not applied the BML model into the actual urban traffic network, and intelligent scheduling and guidance. The basic BML model is improved to predict traffic congestion of the actual city road network. This study focuses on the amelioration of BML model and the numerical simulation based on this model. The main idea of this paper is: (1) the real urban traffic road network mapped to the D-BML model in accordance with a certain strategy. And the proportion of traffic volume density should be mapped to the D-BML model in accordance with of the road network in a time period. (2)D-BML model evolves in accordance with rule 184 of the cellular automaton evolution. The jam coordinate values of model grid points can be got when the model eventually enter the blocked state. (3)The severe congestion at the intersection also can be got in the next period of time. The experimental results show that the proposed D-BML model can accurately predict the blocked intersection in the road network. The first analysis of the D-BML model density of traffic is on the uniform distribution and Poisson distributions. The blockage values on the model grid are analyzed. The experiments show that as long as the initialization traffic flow density distribution determined, the clogging grid points will not change. Then engineering experiments are made. The simulation experiments of this article are about England's Birmingham City traffic network. The final experimental results are very consistent with the actual traffic clogged intersections, which strongly proved correctness and effectiveness of M-BML model.

Original document

The different versions of the original document can be found in:

• http://dx.doi.org/10.2991/iccsae-15.2016.186 under the license https://creativecommons.org/licenses/by-nc

• https://download.atlantis-press.com/article/25848298.pdf

• https://www.atlantis-press.com/proceedings/iccsae-15/25848298,

https://academic.microsoft.com/#/detail/2295485565 under the license cc-by-nc