Modeling Traffic Congestion with Spatiotemporal Big Data for An Intelligent Freeway Monitoring System
Karisma Trinanda Putra (a*)(b), Jing-Doo Wang (b), Eko Prasetyo (a)(b), Prayitno (b)(c)
a) Universitas Muhammadiyah Yogyakarta,
Jl. Brawijaya, Kasihan, Bantul, Yogyakarta, Indonesia
*karisma[at]ft.umy.ac.id
b) Asia University,
No. 500, Liufeng Road, Wufeng District, Taichung City, Guangdong Province, Republic of China
c) Politeknik Negeri Semarang,
Jl. Prof. H. Soedarto S.H., Tembalang, Semarang, Indonesia
Abstract
Traffic congestion is a complex, nonlinear spatiotemporal modeling problem. By collecting and analyzing a vast quantity and different categories of information, traffic flow and road congestion can be predicted and avoided in intelligent transportation system. This report provides an analysis about traveling time across Taiwan from North to South, vice versa. We analyze traffic in a national freeway section between Tainan and Kaohsiung, which represents the common trip of the population. The case study is recorded especially between Tainan and Kaohsiung using TDCS database provided by Ministry of Transportation in Taiwan. We use MapReduce framework to process data into smaller task which can be distributed on several computer cluster to speed up the process. The results show that traffic flow spatiotemporal model is strongly influenced by holidays, direction and working hour with a recurring pattern for each week.
Keywords: Traffic congestion, spatiotemporal, TDCS, MapReduce
Topic: International Symposium of Engineering, Technology, and Health Sciences