安全共乘服務之近似最佳配對及異常駕駛偵測

Sub-optimal Matching and Anomalous Driver Detection for Safe Carpool Services

張家裕(Jia-Yu Zhang)

指導教授 ： 吳宜鴻

中原大學/電機資訊學院/通訊工程碩士學位學程/碩士(2013年)

https://doi.org/10.6840/cycu201300829

摘要

世界各國推廣共乘主要理由是減緩交通堵塞和空氣汙染，駕駛共乘的目的是節省旅行車資，乘客則是要找到一台可以搭乘的汽車。近年來智慧型手機以及通訊技術的發展帶來了更多參與共乘的機會，但並不是在每一個國家推廣共乘都是順利的；乘客常因為陌生的駕駛且缺乏安全的保障而排斥共乘。大多數先前的共乘系統推薦給使用者符合條件的所有候選者，使用者必須付出許多時間挑選對象，而且不確定哪些是比較安全的對象。本論文提出一個安全共乘服務的架構以減輕挑選對象的負擔，以評價機制決定相對比較安全的共乘配對，接著分析智慧型手機的定位資訊，並監控駕駛的移動軌跡以便偵測異常狀況，藉此確保參與共乘者的安全。

關鍵字

合用共乘 ； 配對 ； 代表軌跡 ； 異常偵測

並列摘要

The reason why countries all over the world promote the carpool services is to reduce the traffic congestion and air pollution. Drivers aim at saving the traveling expenses while passengers attempt to find a car they can take a ride. In recent years, the development of smart phones and communication technology led to more opportunities of sharing rides, but the promotion of carpooling was not always successful in every country. Passengers often reject a carpool service because of unfamiliar drivers and no safety guarantee. Most of the previous carpool systems recommended users all the candidates qualified with the user-specified conditions. Users had to spend a lot of time on candidate selection and were uncertain which ones would be safer. This thesis proposes a framework of the safe carpool service to reduce the overhead on candidate selection. With an evaluation mechanism, relatively safer pairs of ridesharing can be decided. After that, we analyze the GPS information from smart phones and monitor the moving trajectory of the driver to detect abnormal situations. Accordingly, the safety of ridesharing participants can be ensured.

並列關鍵字

anomaly detection ； ridesharing ； matching ； representative trajectory

參考文獻

[1] A. Volgenant, “Linear and semi-assignment problems: A core oriented approach,” Computers & OR, 23(10): 917-932, 1996.

[2] Chao Chen, Daqing Zhang, Pablo Samuel Castro, Nan Li, Lin Sun, and Shijian Li, “Real-Time Detection of Anomalous Taxi Trajectories from GPS Traces,” MobiQuitous, pp. 63-74, 2011.

[4] Daqing Zhang, Nan Li, Zhi-Hua Zhou, Chao Chen, Lin Sun, and Shijian Li, “iBAT: Detecting Anomalous Taxi Trajectories from GPS Traces,” UbiComp, pp. 99-108, 2011.

[6] Harold N. Gabow and Robert Endre Tarjan, “Faster Scaling Algorithms for Network Problems,” SIAM J. Comput., 18(5): 1013-1036, 1989.

[7] Harold N. Gabow, “Scaling Algorithms for Network Problems,” FOCS, pp. 248-257, 1983.