|
Siamtraffic2.0: Traffic pattern search for travel time prediction in Bangkok road network |
|---|---|
| รหัสดีโอไอ | |
| Creator | เกรียงศักดิ์ วณิชชากรพงศ์ |
| Title | Siamtraffic2.0: Traffic pattern search for travel time prediction in Bangkok road network |
| Contributor | ณกร อินทร์พยุง, เอกชัย สุมาลี |
| Publisher | Faculty of Information Science and Technology, Mahanakorn University of Technology |
| Publication Year | 2556 |
| Journal Title | Journal of Information Science and Technology |
| Journal Vol. | 4 |
| Journal No. | 1 |
| Page no. | 1-10 |
| Keyword | Travel time prediction, Traffic pattern search, Non-parametric, forecasting |
| URL Website | https://tci-thaijo.org/index.php/JIST |
| Website title | Journal of Information Science and Technology |
| ISSN | 2651-1053 |
| Abstract | This paper presents the travel time prediction system to support the trip planning in Bangkok road network. The system relies on pattern matching taking spatial and temporal traffic speed correlations into account. The road network is divided into traffic link classified independently by type of day for every 5-min time interval. Traffic pattern is then defined by the speed interval. The objective function of the problem is to find the closest match of the traffic pattern in the historical database, in order to estimate the travel time for a specific traffic link. We evaluate the accuracy of forecast by comparing the predicted travel time with the observed data. Four types of the observed links indicate different behaviors of traffic patterns that affect to the applicability of the prediction system. The results have shown that the free-flow traffic link is greater than 85 percent correctly predicted, and the link types of the interrupted traffic flow indicate 60 80 percent in the accuracy of forecast. The efficiency of the algorithm in a computational run-time is less than one second per link in 5-min time interval. |
