Traffic Information Detection Based on Scattered Sensor Data: Model and Algorithms
Wei Zhang, Guo-Zhen Tan and Nan Ding
This paper presents the mathematical model and algorithms for traffic flow information detection based on proximity sensor networks. Take into account the intrinsic properties of traffic flow and the principle of traffic congestion formation to build an observation model in the intersection and near segments. Based on the analytical model, this paper developed the method and algorithms to estimate traffic parameter with the scattered sensor data, and reconstruct the traffic surface using numerical interpolation and finite elements method. The result is expected to support the optimal global timing for the purpose of traffic light control, real-time traffic state monitoring and evaluation, and try to avoid the traffic congestion before it formation. The performance is analyzed based on the Mobile Century dataset. The simulation result shows that this method can improve the spatial-temporal resolution of traffic detection, and it is helpful to make quantitative analysis of traffic congestion.
Keywords: Intelligent transportation system (ITS); wireless sensor networks (WSN); traffic surveillance; traffic flow theory; traffic congestion model; nonlinear surface reconstruction; numerical interpolation; scattered data fitting; finite elements method.