Title :
Algorithms of SVM-AID based on data-level and decision-making-level data fusion methods
Author :
Zhili, Cai ; Guiyan, Jiang ; Yong, Liu
Author_Institution :
Shandong Jiaotong Univ., Jinan
Abstract :
According to the traffic flow characteristic and single vehicle operation characteristic under incidents situation on freeway, two new SVM-AID algorithms with different fusion methods are put forward based on SVM and data fusion techniques. According to a certain rules, the first algorithm fuses data such as volume, speed and occupancy collected by fixed detectors and the data of single vehicle instantaneous speed and average travel time by GPS FCs, and then applies SVM models to realize traffic incidents detection. In the second algorithm, SVM model is firstly applied to detect incidents respectively using data collected by the two detectors above, and then decision-making-class fusion of detection results is realized applying weighting method. The simulation results demonstrated that the two algorithms put forward in this paper can effectively real-time detect traffic incidents occurred on freeway, and they possessed well detection capability.
Keywords :
decision making; sensor fusion; support vector machines; traffic engineering computing; GPS; SVM-AID; decision-making-level data fusion methods; traffic incidents detection; weighting method; Algorithm design and analysis; Detectors; Educational institutions; Fuses; Global Positioning System; Lagrangian functions; Road vehicles; Support vector machines; Traffic control; Vehicle detection; AID; Data Fusion; Freeway; GPS; ITS; SVM;
Conference_Titel :
Control and Decision Conference, 2008. CCDC 2008. Chinese
Conference_Location :
Yantai, Shandong
Print_ISBN :
978-1-4244-1733-9
Electronic_ISBN :
978-1-4244-1734-6
DOI :
10.1109/CCDC.2008.4598053