Robust Tolling

We revisited the second-best toll pricing (SBTP) problem, which has been traditionally formulated as bi-level or MPEC problems. The upper level is to optimize certain objective from the transportation system's point of view (such as the total network travel times) and the lower level is a user equilibrium (UE) problem to account for the choice behavior of individual motorists. Most existing SBTP models assume (explicitly or implicitly) that under a certain toll the lower level UE has a unique solution. Such a uniqueness assumption requires strong assumption on travel time or cost function of the traffic, which may not hold especially for metro areas if strong interactions exist among different links. The non-uniqueness of motorist’s response to pricing, therefore, represents uncertainty in the SBTP design, which has not been well recognized in the literature. The question is then how tolls should be designed to account for this uncertainty. We recently discovered that from a toll designer's perspective, the traditional bi-level SBTP scheme is risk-prone if lower level UE does not have a unique solution. Such a toll design scheme is not reliable as one aims to optimize only for the best case scenarios. We are now working on developing models and algorithms for risk-averse and risk-neutral SBTP schemes. The former scheme is designed to be optimal for the worst case scenario, while the latter scheme is optimal for the average case scenario.

Publications

  1. Ban, X., and Ma, R., 2010. Realization probability of user equilibrium solutions under nonunique equilibria. Presented at the 90th Transportation Research Board Annual Meeting.
  2. Ban, X., Ferris, M., Liu, H., 2010. Numerical studies on reformulation techniques for continuous network design problems with asymmetric user equilibrium. International Journal of Operations Research and Information Systems, 1(1), 52-72.
  3. Ban, X.,Ferris, M.C., and Tang, L., 2009. Risk netural second best toll pricing. Submitted for publication
  4. Ban, X.,Lu, S., Ferris, M.C., and Liu, H., 2009. Risk averse second best toll pricing. In Transportation and Traffic Theory, Chpt. 10 (W.H.K. Lam, S.C. Wong, H.K. Lo eds.), Springer, 197-218..
  5. Ban, X., and Lu, S., 2007. Risk taking behaviors in second best toll pricing. Presented at the 86th Transportation Research Board Annual Meeting.
  6. Ban, X., Liu, H., Ferris, M., and Ran, B., 2006. A general MPCC model and its solution algorithm for continuous network design problem. Mathematical and Computer Modeling 43, 493-505.
  7. Ban, X., Liu, H, and Ferris, M.C., 2006. Decomposition scheme for continuous network design problem with asymmetric user equilibria. Transportation Research Record 1964, 185-192.
  8. Ban, X., Liu, H, and Ferris, M.C., 2006. A link-node based complementarity model and its solution algorithm for asymmetric user equilibria. In Proceedings of the 85th Transportation Research Board Annual Meeting (CD-ROM).

Optimal Traffic Sensor Deployment

A large portion of traffic sensors are currently deployed on a case by case basis by practitioners without a systematic study of the quantity and locations of sensors needed. Since traffic sensors are limited resources, determining optimal placement strategies maximizes the value of this resource. Sponsored by California Department of Transportation (Caltrans), we developed a Dynamic Programming (DP) model to determine the optimal sensor locations for various traffic applications such as freeway travel time estimation, ramp metering (density or occupancy estimation). The developed algorithm is polynomial and can be solved as a shortest path problem in an acyclic graph. The DP model can incorporate sensors that have already been deployed in the field and the results match well with the bottleneck areas of the freeway. See here for an illustration of the optimal locations of 6 sensors for a 5.5-mile segment of Interstate 880 in the Bay Area, and here for how the optimal sensor locations evolve if more sensors are added in. Below is a list of our recent publications in this topic:

Publications

  1. Ban, X., Chu, L., Herring, R., and Margulici, J.D. (2010) A sequential modeling framework for optimal sensor placement for multiple ITS applications. ASCE Journal of Transportation Engineering, in press.
  2. Ban, X., Chu, L., Herring, R., and Margulici, J.D. (2009) Optimal sensor placement for both traffic control and traveler information applications. In Proceedings of the 88th Transportation Research Board Annual Meeting.
  3. Ban, X., Herring, R., Margulici, J.D., and Alex Bayen, 2009. Optimal sensor placement for freeway travel time estimation. Transportation and Traffic Theory, Chpt. 34 (W.H.K. Lam, S.C. Wong, H.K. Lo eds.), Springer, 697-721. .

Corridor-Based Transportation Management and Simulation

Corridor-based Transportation Management (CTM) aims to develop systematic, multimodal, and sustainable improvement strategies for heavy traveled transportation corridors, rather than focusing on individual problems at isolated locations or modes. Examples include the Integrated Corridor Management (ICM) initiative by the Federal Highway Administrations (FHWA), the state-wide Corridor System Management Planning (CSMP) in California, and the Corridor-Based Transportation Management in New York State (the 4th Theme of the NYSDOT Transportation Master Plan for 2030). Sponsored by Caltrans and FHWA, we investigated (a) data needs and requirements for corridor-wide performance assessment and (b) developed a simulation-based framework to evaluate various improvement strategies. We especially developed percentile-speed-based methods for corridor-wide bottleneck identification and calibration of bottlenecks in micro-simulation. The work can help practitioners make informed decisions to select the most cost-effective improvement strategies for a given transportation corridor. Below is a list of our recent publications in this topic:

Publications

  1. Ban, X., Sun, Z.*, 2013. Simulation-based Decision-making Tool for Adaptive Traffic Signal Control on Tarrytown Road in the City of White Plains. Fine Report Submitted to NYSERDA and NYSDOT.
  2. Alm, E., Lingham, V., Benouar, H., Ban, X., and Chu, L. (2008) An integrated methodology for corridor management planning. In Proceedings of the 87th Transportation Research Board Annual Meeting (CD-ROM).
  3. Ban, X., Chu, L., and Benouar, H. (2007) Bottleneck Identification and Calibration for Corridor Management Planning. Transportation Research Record 1999, 40-53.
  4. Ban, X., Chu, L., and Benouar, H. (2007) Bottleneck calibration in micro-simulation for corridor management using data from single loop detectors. In Proceedings of the 14th ITS World Congress (CD-ROM) .
  5. Liu, H., Ding, L., Ban, X., Chen, A., and Chootinan, P. (2006) A Streamlined Network Calibration Procedure for California SR41 Corridor Traffic Simulation Study. In Proceedings of the 85th Transportation Research Board Annual Meeting (CD-ROM).
  6. Liu, H., Ma., W., Ban, X., and Mirchandani, P. (2005) Dynamic equilibrium assignment with microscopic traffic simulation. In Proceedings of the 8th IEEE International Conference on Intelligent Transportation Systems (CD-ROM) .

Posting Travel Times on Variable Message Signs (VMS)

Travel time is a crucial measure of traffic conditions and system performance, which can also enable motorists for more informed decisions regarding their mode route choices. Posting travel times has now become a common practice by all state DOTs around the US, as recommended by FHWA that “No new DMS should be installed in a major metropolitan area or along a heavily traveled route unless the operating agency and the jurisdiction have the capability to display travel time messages.” As indicated in this figure, a typical VMS travel time system consists of three major components: the Signs, travel time estimation, and reactions and compliance of motorists. Sponsored by Caltrans, we defined quality measures and conducted benchmarking studied for travel times estimated from various sources and via commonly used algorithms. Below is a list of our recent publications in this topic:

Publications

  1. Margulici, J.D., and Ban, X. (2008) Benchmarking travel time estimates. IET Journal of Intelligent Transportation Systems, in press.
  2. Ban, X., Li, Y., and Skabardonis, A. (2007) Local MAD Method for Probe Vehicle Data Processing. In Proceedings of the 14th ITS World Congress.
  3. Ban, X., Li, Y., Skabardonis, A., and Margulici, J.D. (2007) Performance evaluation of travel time estimation methods for real time traffic applications.In Proceedings of the 11th World Congress on Transport Research, Berkeley, CA.
  4. Lu, J.G., Yang, F., Ban, X., and Ran, B. (2006) Moments analysis for improving decision reliability based on travel time. Transportation Research Record 1968, 109-116.
  5. Lu, J.G., Ban, X., Qiu, Z.J., Yang, F., and Ran, B. (2005) Robust route guidance model based on advanced traveler information systems. Transportation Research Record 1935, 1-7.

Real-Time Emergency Evacuation Modeling

We proposed an adaptive control based framework to determine optimal real time control strategies under emergency evacuation. The modeling framework will help reduce congestion, and further life losses and property damages during evacuation, especially unnoticed evacuation such as terrorist attacks. Below is a list of our recent publications in this topic:

Publications

  1. Liu, H., Ban, X., Ma, W.T., and Mirchandani, P. (2007) Model Reference Adaptive Control Framework for Real Time Traffic Management under Emergency Evacuation. Journal of Urban Planning and Development 133(1), 43-50.
  2. Liu, H., He, X., Ban, X. (2007) A Cell-Based Many-to-One Dynamic System Optimal Model and Its Heuristic Solution Method for Emergency Evacuation, In Proceedings of the 86th Transportation Research Board Annual Meeting (CD-ROM).