Robust PCA-Based Abnormal Traffic Flow Pattern Isolation and Loop Detector Fault Detection

• Published in: Tsinghua science and technology Vol. 13; no. 6; pp. 829 - 835
• Main Authors: Jin, Xuexiang, Zhang, Yi, Li, Li, Hu, Jianming
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.12.2008; Department of Automation, Tsinghua University, Beijing 100084, China
• Subjects: loop detector faults; robust principal components analysis (RPCA); traffic flow pattern; robust principal components analysis (RPCA); loop detector faults; traffic flow pattern
• ISSN: 1007-0214; 1878-7606; 1007-0214
• DOI: 10.1016/S1007-0214(08)72208-9

One key function of intelligent transportation systems is to automatically detect abnormal traffic phenomena and to help further investigations of the cause of the abnormality. This paper describes a robust principal components analysis (RPCA)-based abnormal traffic flow pattern isolation and loop detector fault detection method. The results show that RPCA is a useful tool to distinguish regular traffic flow from abnormal traffic flow patterns caused by accidents and loop detector faults. This approach gives an effective traffic flow data pre-processing method to reduce the human effort in finding potential loop detector faults. The method can also be used to further investigate the causes of the abnormality.