Traffic Incident Detection Algorithm for Urban Expressways Based on Probe Vehicle Data

Abstract

Traffic incidents are major contributors to traffic delay. Expedient and reliable incident detection is an important part of expressway traffic management and control in urban areas. This paper proposes an automated incident detection algorithm for urban expressways based on probe vehicle data by the CUSUM algorithm based approach. The presented algorithm is validated using the actual incident information and the probe vehicle data collected in Beijing at different time intervals. Its performance is evaluated and further compared with that of an existing probe-vehicle-based incident detection algorithm, the UCB algorithm. The results show that the performance of the proposed algorithm under 1-minute time interval is better than that under 5-minute time interval. It is also observed that the performance of the proposed algorithm is better than that of the UCB algorithm. Under the same false alarm rate, the detection rate of the proposed algorithm is improved by 10%.