| 1. |
|
|
| 2. |
- Allström, Andreas, 1978-, et al.
(författare)
-
Mobile Millennium Stockholm
- 2011
-
Ingår i: 2nd International Conference on Models and Technologies for Intelligent Transportation Systems.
-
Konferensbidrag (övrigt vetenskapligt/konstnärligt)
|
|
| 3. |
- Antoniou, C, et al.
(författare)
-
A Comparison of Machine Learning Models for Speed Estimation
- 2006
-
Ingår i: IFAC Proceedings Volumes. - Delft, The Netherlands. - 9783902661135 ; , s. 55-60
-
Konferensbidrag (refereegranskat)abstract
- Speed-density relationships are a classic way of modeling stationary traffic relationships. Besides offering valuable insight in traffic stream flows, such relationships are widely used in simulation-based Dynamic Traffic Assignment (DTA) systems. In this paper, alternative approaches for modeling traffic dynamics, appropriate for traffic simulation, are proposed. Their basic premise is the wide availability of sensor data. The approaches are based on machine learning methods such as locally weighted regression and support vector regression. Neural networks are also considered, as they are a well-established approach, successful in many applications. While such models may not provide as much insight into traffic flow theory, they allow for easy incorporation of additional information tospeed estimation, and hence, may be more appropriate for use in DTA models, especially simulation based. In particular, in this paper, it is demonstrated (using data from a network in Irvine, CA) that the use of such machine learning methods can improve the accuracy of speed estimation.
|
|
| 4. |
- Antoniou, C, et al.
(författare)
-
A synthesis of emerging data collection technologies and their impact on traffic management applications
- 2011
-
Ingår i: European Transport Research Review. - : Springer Science and Business Media LLC. - 1867-0717 .- 1866-8887. ; 3:3, s. 139-148
-
Tidskriftsartikel (refereegranskat)abstract
- act Purpose: The objective of this research is to provide an overview of emerging datacollection technologies and their impact on traffic management applications. Methods: Several existing and emerging surveillance technologies are being used for traffic datacollection. Each of these technologies has different technical characteristics and operating principles, which determine the types of data collected, accuracy of the measurements, levels of maturity, feasibility and cost, and network coverage. This paper reviews the different sources of traffic surveillance data currently employed, and the types of traffic management applications they may support. Results: Automated Vehicle Identification data have several applications in traffic management and many more are certain to emerge as these data become more widely available, reliable, and accessible. Representative examples in this field are presented. Furthermore, the fusion of condition information with traffic data can result in better and more responsive dynamic trafficmanagement applications with a richer data background. Conclusions: The current state-of-the-art of traffic modeling is discussed, in the context of using emerging data sources for better planning, operations and dynamic management of road networks.
|
|
| 5. |
- Antoniou, C, et al.
(författare)
-
An Efficient Non-linear Kalman Filtering Algorithm Using Simultaneous Perturbation and Applications in Traffic Estimation and Prediction
- 2007
-
Ingår i: IEEE Conference on Intelligent Transportation Systems, Proceedings, ITSC. - 9781424413966 ; , s. 217-222
-
Konferensbidrag (refereegranskat)abstract
- The Extended Kalman Filter, a well-established and straightforward extension of theKalman filter, requires a computationally intensive linearization step. In this paper, the use of the simultaneous perturbation is proposed for the computation of the gradient in a far more efficient way than the usual numerical derivatives. The resulting algorithm is applied to the problem of on-line calibration of traffic dynamics models and empirical results are presented. The use of the simultaneous perturbation gradient approximation provides significant improvement over the base case, and comparable results to those obtained by the more computationally intensive finite difference gradient approximation.
|
|
| 6. |
- Antoniou, C, et al.
(författare)
-
Calibration Methods for Simulation-Based Dynamic Traffic Assignment Systems
- 2011
-
Ingår i: International Journal of Modelling and Simulation. - 0228-6203. ; 31:3, s. 227-233
-
Tidskriftsartikel (refereegranskat)abstract
- Dynamic Traffic Assignment (DTA) integrates complex transportation demand and network supply simulation models to estimate prevailing traffic conditions, predict future network performance and generate consistent, anticipatory route guidance. Prior to deployment, the DTA's parameters and inputs must be calibrated to accurately reflect travel behaviour and traffic dynamics. This paper presents a unified framework for off-line and on-line DTA calibration. Off-line calibration simultaneously estimates demand and supply model parameters. On-line calibration jointly updates - in real-time - the off-line estimates in order to more accurately capture current conditions. The developed methodsare flexible and can be applied to any simulation model and may utilize any availabletraffic surveillance information (including Automated Vehicle Identification systems, probe vehicles and other emerging data sources). The off-line and on-line components complement each other to efficiently combine historical and real-time information. Thecalibration approaches are demonstrated with DynaMIT (Dynamic network assignmentfor the Management of Information to Travelers), using time-varying count, speed and density data from conventional traffic sensors.
|
|
| 7. |
|
|
| 8. |
- Antoniou, Constantinos, et al.
(författare)
-
Dynamic Traffic Demand Prediction Using Conventional and Emerging Data Sources
- 2006
-
Ingår i: IEE Proceedings Intelligent Transport Systems. - : Institution of Engineering and Technology (IET). - 1748-0248. ; 153:1, s. 97-104
-
Tidskriftsartikel (refereegranskat)abstract
- Origin-destination (OD) flow estimation and prediction is an important problem with applications in Dynamic Traffic Management, and traffic estimation and prediction systems. Recent developments in traffic data collection technologies provide data that have not yet been used in OD estimation and prediction. In this paper, a new, flexible, and general methodology for OD estimation and prediction is presented. The methodology can incorporate any available information from conventional and emerging traffic data collection technologies (such as automatic vehicle identification systems and probe vehicles). The application of the methodology is presented through a case study. The results support the importance of incorporating additional data in the OD estimation problem. An overall improvement for estimation and one-step prediction exceeds 45 when point-to-point information is added to the model (over the base case when only point link flows are available), while an improvement of more than 35 is maintained even for four-step prediction (i.e. 1 h into the future).
|
|
| 9. |
|
|
| 10. |
- Antoniou, Constantinos, et al.
(författare)
-
Estimation of Traffic Dynamics Models with Machine Learning Methods
- 2006
-
Ingår i: Transportation Research Record. - 0361-1981 .- 2169-4052. ; 1965, s. 103-111
-
Tidskriftsartikel (refereegranskat)abstract
- Speed-density relationships are a classic way of modeling stationary traffic relationships. Besides offering valuable insight into traffic stream flows, such relationships are widely used in dynamic traffic assignment (DTA) systems. In this research, an alternative paradigm for traffic dynamics models, appropriate for traffic simulation models and based on machine-learning approaches such as k-means clustering, k-nearest-neighborhood classification, and locally weighted regression is proposed. Although these models may not provide as much insight into traffic flow theory as speed-density relationships do, they allow for easy incorporation of additional information to speed estimation and hence may be more appropriate for use in DTA models, especially simulation-based models. This paper (with data from a network in Irvine, California) demonstrates that such machine-learning methods can considerably improve the accuracy of speed estimation.
|
|
