Lane-Based Traffic Arrival Pattern Estimation Using License Plate Recognition Data
Chengchuan An, Xiaoyu Guo, Rongrong Hong, Zhenbo Lu, Jingxin Xia
Abstract
Understanding the traffic arrival process and its patterns is of vital importance for delay and queue analysis at intersections. The installation of advance loop detectors to sense vehicle arrivals could be costly and biased. Utilizing sampled vehicle trajectory data to reconstruct the traffic arrival flow might suffer from small sample sizes. License plate recognition (LPR) data commonly available at intersections in cities in China are promising for overcoming such limitations. This study aims to estimate a lane-based traffic arrival pattern by using LPR data collected at downstream and upstream intersections. The proposed method develops a probability model with an assumption of a two-stage piecewise arrival process for upstream merge movements. Given observations of vehicle arrivals provided by matched cars and trucks in LPR data, the model estimates second-based mean arrival rates for each lane at the downstream intersection. The proposed method is validated using actual LPR data collected at two adjacent intersections in Kunshan City, China. The results demonstrate that the proposed method can describe the traffic arrival patterns of upstream merge movements with either two-stage or uniform arrival processes in different traffic scenarios. In addition, the proposed method is more robust and reliable than an average-based benchmark method in terms of revealing actual traffic arrival patterns under different match rates.