Travel time prediction for urban network is a hard and complicated task so that it is regarded as theoretically feasible but difficult to accomplish using traditional models.
Knowledge based TTP model proposed in this thesis has demonstrated that TTP for urban network could be achieved by utilizing the LBS application data. The data mining technique is used to discover some traffic patterns for building some knowledge classes and expertise with human expert, and then formalize a TTP equation in complex Taipei urban traffic network. Real-time external traffic data sources and meta-rules mechanism dynamically tunes the combination ratio of real-time and historical travel time predictors promote the precision one step further. The result of experiments shows that TTP in urban network can be achieved in tolerable range.
In the experiment, we found that link granularity might be a problem in system implementation. We choose section in a road as our link granularity, which may contain several crosses with other streets. Greater link choice reduces the complexity of road network but loses the precision comparing with small link choice.
In the implemented prototype system, we categorized the temporal domain as
“holiday”, “weekday”, “rush hour”, and “normal hour”. The granularity of this categorization seems too simple to make advance traffic status analyzing. In the near future work, we may focus on finding more historical traffic patterns to make the historical TTP more precisely, such as after a long vocation days or Monday after Sunday, etc. [1, 2] Besides, in order to make our historical TTP more robust, the Kalman Filter technique can help to regress the TTP with previous travel time for promoting our
historical TTP part more smoothly.
In real-time TTP issue, the consideration of the real-time events is important for our TTP results, but how can we estimate the impact on real-time TTP? For example, how long the vehicles will be delayed when there is a traffic accident on the candidate path? Or, how a heavy rainfall affects our real-time TTP? The Case Base Reasoning (CBR) might solve this problem. We can analyze historical traffic events in the past to realize the average traffic speed and the recovery time slot, then record it to our CBR database for measuring our real-time TTP in advanced.
In further research of this thesis, CSTR and CSSTP can be utilized for other topics, such as routing problem. This thesis deals the routing problem with statistic method to find the taxi driver’s candidate paths, but sometime these paths may not be the best solutions according to the current traffic status (The candidate paths were calculated in past traffic time or there maybe exist some disguisedly optimal paths).The CSTR, CSSTP or other designed traffic patterns can be used to solve the routing problem in order to generate suggested best paths by considering real-time events and traffic status in the future.
Reference
1. S.-I. Bajwa, E. Chung, and M. Kuwahara, “An Adaptive Travel Time Prediction Model Based On Pattern Matching”, Proc. of 11th World Congress on ITS, 2004.10.
2. E. Chung, ”Classification of Traffic Pattern”, Proc. of the 11th World Congress on ITS, 2003.
3. E. Chung, et al., ”Cleansing of Probe Car Data to Determine Trip OD”, 21st ARRB and 11th REAAA Conference, Cairns, Australia, 2003
4. E. Chung, et al., “Travel Time Prediction: Issues and Benefits”,Proc. of the 10th World Conference on Transport Research, 2003.
5. DRAMA, Coretech Inc, http://www.ctknow.com.tw/rule.htm, 2003.
6. e-IOT, Institute of Transportation MOTC, http://e-iot.iot.gov.tw/
7. T. Iroy and M. Kuwahara, “Simulation Analyses of Traffic Congestion Alleviation by Demand Spreading Over Time”, Journal of the Eastern Asia Society for Transportation Studies, Vol.4, No.4, pp.161 - 174, October 2001.
8. W.-H. Lee, et al., “Electronic Toll Collection Based on Vehicle-Positioning System Techniques”, Proc. 2004 IEEE International Conf. on Networking Sensing and Control.
9. B. Lent, A. Swami, and J. Widom, “Clustering Association Rules”, ACM Proc. of 13th International Conference on Data Engineering, 1996.
10. H.-E. Lin and R. Zito, “A Review of Travel-Time Prediction in Transport and Logistics”, Proc. of the Eastern Asia Society for Transportation Studies, Vol. 5, pp.
1433 - 1448, 2005.
11. W.-H. Lin, A. Kulkarni, and P. Mirchandani, “Short-Term Arterial Travel Time Prediction for Advanced Traveler Information Systems”, Journal of Intelligent
Transportation Systems, 8:143 - 154, 2004
12. H.-Y. Liu, et al., “A Intelligent Taxi Dispatching Management System”, Proc. of Cross Strait Conference on ITS, pp. 111 - 117, 2004.
13. Y.-T. Lin, et al., “Design and Implement of New Object-Oriented Rule Base Management System”, Expert Systems with Applications, 25, 369 - 385, 2005.
14. C.-D. Mark, A. W. Sadek, and D. Rizzo, ”Predicting Experienced Travel Time with Neural Networks: a PARAMICS Simulation Study”, Proc. of IEEE Conference on Intelligent Transportation Systems, pp.906 - 911, October 2004.
15. K. Nagaoka, “Travel Time System by Using Vehicle Information and Communication System (VICS)”, Proc. of IEEE on ITS conference, 1999.
16. T. Nakata and J. Takeuchi, “Mining Traffic Data form Probe-Car System for Travel Time Prediction,” Proc. of the 2004 ACM SIGKDD international conference on Knowledge Discovery and Data Mining, Seattle, WA, USA, pp 817 - 822, August 2004.
17. I. Safa, et al., “Spatiotemporal Aggregate Computation: A Survey”, IEEE Trans. on KDE, Vol. 17, No. 2, pp 271 - 286, February 2005.
18. M. Sarvi, et al., “A Methodology to Identify Traffic Condition Using Intelligent Probe Vehicles”, Proc. of 10th World Congress on ITS, Madrid, 2003.
19. H. Takahashi, S. Takahashi, and T. Izumi, “Travel Time measurement using Genetic Algorithm in Vehicle Sequence Matching Processing”, 2nd ITS Symposium, Osaka, 2003
20. C.-W. Wang, et al., “A Geocoding Application on GIS Using Address Data: Case Study of Taiwan Address Database”, Proc. of 2003 Annual Conference on TGIS, Taipei, Taiwan, pp.25-6.
21. Y.-H. Wen, T.-T. Lee, and H.-J. Cho, “Missing Data Treatment and Data Fusion Toward Travel Time Estimation for ATIS”, Journal of the Eastern Asia Society for
Transportation Studies, Vol. 6, pp. 2546 - 2560, 2005
22. C.-H. Wu, J.-M. Ho, and D.T. Lee, “Travel Time Prediction with Support Vector Regression”, Proc. of IEEE Intelligent Transportation Systems Conference, October, 2003.
23. J.-S. Yang “Travel Time Prediction Using the GPS Test Vehicle and Kalman Filtering Techniques”, Proc. of the 2005 American Control Conference, pp.
2128-2133, June 8 - 10, 2005.
24. B.-S. Yoo, S.-P. Kang, and C.-H. Park, “Travel Time Estimation Using Mobile Data”, Proc. of the Eastern Asia Society for Transportation Studies, Vol. 5, pp.
1533 - 1547, 2005.
25. H.-S. Zhang, et al., “Spatial-Temporal Traffic Data Analysis Based on Global Data Management Using MAS”, IEEE Trans. on ITS, Vol. 5, No. 4, pp 267 - 275, December 2004.