Detecting Anomalous Bus-Driving Behaviors from Trajectories
- PDF / 5,832,030 Bytes
- 17 Pages / 595 x 842 pts (A4) Page_size
- 61 Downloads / 190 Views
Detecting Anomalous Bus-Driving Behaviors from Trajectories Zhao-Yang Wang1,2 , Student Member, CCF, Bei-Hong Jin1,2,∗ , Senior Member, CCF, Member, ACM, IEEE Tingjian Ge3 , and Tao-Feng Xue1,2 , Student Member, CCF 1
State Key Laboratory of Computer Science, Institute of Software, Chinese Academy of Sciences, Beijing 100190, China
2
University of Chinese Academy of Sciences, Beijing 100049, China
3
Department of Computer Science, University of Massachusetts Lowell, Lowell, MA 01854, U.S.A.
E-mail: [email protected]; [email protected]; [email protected] E-mail: [email protected] Received August 14, 2019; revised October 9, 2019. Abstract In urban transit systems, discovering anomalous bus-driving behaviors in time is an important technique for monitoring the safety risk of public transportation and improving the satisfaction of passengers. This paper proposes a twophase approach named Cygnus to detect anomalous driving behaviors from bus trajectories, which utilizes collected sensor data of smart phones as well as subjective assessments from bus passengers by crowd sensing. By optimizing support vector machines, Cygnus discovers the anomalous bus trajectory candidates in the first phase, and distinguishes real anomalies from the candidates, as well as identifies the types of driving anomalies in the second phase. To improve the anomaly detection performance and robustness, Cygnus introduces virtual labels of trajectories and proposes a correntropy-based policy to improve the robustness to noise, combines the unsupervised anomaly detection and supervised classification, and further refines the classification procedure, thus forming an integrated and practical solution. Extensive experiments are conducted on real-world bus trajectories. The experimental results demonstrate that Cygnus detects anomalous bus-driving behaviors in an effective, robust, and timely manner. Keywords
1
anomaly detection, bus trajectory, crowd sensing, bus-driving safety
Introduction
In most areas of the world, buses play an important role in urban transit systems. Taking Beijing as an example, there are nearly 60 million passenger bus-riding trips per day. As to the operation of public transportation, safe driving of transit vehicles is an even higherpriority concern than other issues such as bus line coverage, distance between bus stops, and uniform distribution of arrival times at a stop. Due to the limited number of seats on a bus, most passengers are likely in the standing position. Bad driving behaviors, such as sudden braking, abrupt starts and swerves, may reduce the comfort of bus rides, and more severely, cause tumbles or bruises of passengers, especially the elderly. Moreover, anomalous bus-driving behaviors also disturb the
operation of other vehicles, causing traffic accidents and compromising transportation safety. Timely discovery of anomalous bus-driving behaviors would be a critical technical means to monitor the safety risks of public transportation and to improve the bus-riding s
Data Loading...
