EconPapers    
Economics at your fingertips  
 

Identification of Traffic Prediction Parameters

Anuchit Ratanaparadorn, Sasivimol Meeampol, Thaneerat Siripachana and Pornthep Anussornnitisarn
Additional contact information
Anuchit Ratanaparadorn: Kasetsart University, Thailand
Sasivimol Meeampol: Kasetsart University, Thailand
Thaneerat Siripachana: Office of the National Boardcasting and Telecommunication Commission, Thailand
Pornthep Anussornnitisarn: Kasetsart University, Thailand

from ToKnowPress

Abstract: Most traffic information we have are real-time, but it is not so useful because we use this information when we decide which way we should go. Imagine we are about to go to work, we look at traffic information and choose the least traffic route. Eventually, traffic might be jam already when we arrive at the road. The problem is real-time information becomes outdated very fast. So how could we fix this? Using Forecasting? It might be good but what factors affects the traffic condition? We should identify it first. There are many factors that could affect the traffic condition. The current and historical traffic condition could be the factors that could predict traffic condition. The ideas are very simple, if the traffic is bad right now, it is more likely that the next five, ten or fifteen minutes later the traffic condition might be the same. Historical traffic condition, we mention is the traffic condition at the same time and same day, for example Traffic condition on this Monday at 8 o’clock might look like the last Monday at the 8 o’clock. Different day such as weekday, weekend or holiday might have different traffic volume, this could also affects the traffic condition. Time period in the day, for example, rush hours or congestion period might be easier to forecast because the driving speed is limited by the vehicle in front of us. And uncongested period might be harder to forecast because it depends on the drivers to decide how fast they want. Many researchers found that the driving speed distribution is not only influenced by weather conditions but also by the degree of the roads and the time of the day. However, Bangkok does not have the weather data available on the road; there are only fifty-two stations (plus two airport weather condition) around it. No thorough study has been conducted to determine if the weather data near the road and time of the day combining with traffic volume and traffic flow speed in multiple periods: weekday, weekend, rush hours could increase prediction result. Even though, the prediction accuracy result is not high as expected but it is possible that current and history speed can be used to predict the future traffic speed. We also proved that other parameters might be time period. When we include time period into the model, the first two links and the last link have better correlation but the rest are the same. It might depend on the link, some link may have different traffic condition in different time and some link may have almost the same traffic condition all day. The weekday, weekend, and holiday could help prediction accuracy higher, but their significant statistical correlations are minimal.

Keywords: traffic prediction; traffic condition; traffic prediction accuracy (search for similar items in EconPapers)
Date: 2013
References: Add references at CitEc
Citations: Track citations by RSS feed

Downloads: (external link)
http://www.toknowpress.net/ISBN/978-961-6914-02-4/papers/ML13-999.pdf full text (application/pdf)
http://www.toknowpress.net/ISBN/978-961-6914-02-4/MakeLearn2013.pdf Conference Programme (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:tkp:mklp13:1479-1486

Access Statistics for this chapter

More chapters in Active Citizenship by Knowledge Management & Innovation: Proceedings of the Management, Knowledge and Learning International Conference 2013 from ToKnowPress
Bibliographic data for series maintained by Alen Jezovnik ().

 
Page updated 2020-06-23
Handle: RePEc:tkp:mklp13:1479-1486