Bruce X Wang

An Exact Markov Process for Multihop Connectivity via Intervehicle Communication on Parallel Roads

This paper identifies a Markov process for the multihop connectivity along two parallel roads through inter-vehicle communication. Vehicles are assumed to follow two Poisson processes on both roads. Exact mean, variance and probability distribution of the instantaneous propagation distance are derived. A closed form approximation for the expected distance is also proposed.

Modeling Delay During Heavy Traffic for Signalized Intersections with Short Left-Turn Bays

This paper develops theoretical delay models for protected left-turn operations at a pretimed signalized intersection during heavy traffic. When the through traffic demand is heavy, residual queues from the previous cycle may occur and elevate the probability of blockage to the left-turn bay, leading to increased delay for left turns. A probabilistic left-turn delay model based on the queuing diagram is proposed for a leading left-turn operation; the influence of residual queues and blockage by the through traffic are taken into account. When the left-turn demand becomes heavy, the left turns may spill back to block the through traffic, resulting in through traffic delays. Through traffic delay is modeled probabilistically on the basis of the analysis of left-turn bay spillback for a lagging protected left-turn operation. The left-turn delay models are validated through carefully designed simulation studies using VISSIM and the results are compared with those from the Highway Capacity Manual (HCM) delay model, which does not consider blockage or spillback situations. The proposed delay models can be used to replace the uniform delay term in the HCM model for high-demand situations when the left-turn operation is affected by spillback and blockage.

Analytical Models for Protected plus Permitted Left-Turn Capacity at Signalized Intersection with Heavy Traffic

This paper presents theoretical analysis of left-turn operations under heavy traffic. When traffic demand is heavy, residual queues from previous cycles may occur and contribute to blockage of the left-turn bay by adjacent through traffic. Such blockage reduces the left-turn capacity and is a random phenomenon. An improved model for protected left-turn capacity is first proposed, taking into account the influence of residual queues and blockage by through traffic. In addition, the classic model always overestimates the potential permitted left-turn capacity when opposing traffic is heavy. The potential permitted left-turn capacity is then modeled in a probabilistic way, and a simple estimation is suggested by investigating its relation to the classic model. Furthermore, because left-turn capacity may vary in response to a blockage, the coefficients of variance for protected left-turn capacity and for queue clearance time are introduced to evaluate the protected left-turn timing plan. The developed analytical models are validated with CORSIM simulations.

Dispatch Problem of Automated Guided Vehicles for Serving Tandem Lift Quay Crane

New quay cranes (QCs) have been designed to increase terminal productivity by lifting more containers simultaneously. But QC productivity relies on efficient cooperation with the vehicles carrying the containers. This paper investigates the synchronization scheduling problem between the automated guided vehicles and these new QCs. The problem is formulated as a mixed integer linear programming model. Because of the problems complexity, a heuristic dispatch rule is proposed for practical purposes. Then, to balance the computation time and the quality of the solution, a neighborhood search method is designed by investigating the working sequences of automated guided vehicles. Numerical experiments show that both heuristics obtain good solutions within extremely short times and that the neighborhood search method generally performs better in relation to the objective value.

Selection of Dredging Projects for Maximizing Waterway System Performance

Budget allocation to annual maintenance dredging projects conducted by the U.S. Army Corps of Engineers is addressed. Of the many projects requested, only a subset may be selected for funding because of budgetary constraints. The benefits of conducting maintenance dredging at one dredging project depend on other projects in the system that also are being dredged to take advantage of transportation efficiency gains derived from deeper navigable waterways. The proposed integer programming models and heuristic solution algorithms selected dredging projects for funding, while taking into account interdependent benefits. The models considered the waterway network topology, which was derived from flow data of historic waterborne cargo. The solution algorithms ensured efficient computational performance and quality of solutions. The models and algorithms were tested on example port systems from the Great Lakes region, as well as from mainline coastal ports and inland waterways. The models were shown to increase the amounts of system cargo throughput directly supported by dredging relative to traditional rank-order approaches that did not consider project interdependencies. This result was true especially for scenarios in which the overall budget was small relative to the total of all funding requests in question.

Analysis of Taxiway Aircraft Traffic at George Bush Intercontinental Airport, Houston, Texas

Serving one of the largest metropolitan areas in the United States, the George Bush Intercontinental Airport (IAH) in Houston, Texas, is one of the 10 airports with the longest average taxi-out and taxi-in times. This paper assesses the congestion at IAH by analyzing taxi times and flight data during different hours of the day. The capacity of IAH is investigated by examining the number of departing flights on the ground. IAH is operating close to capacity most of the time. Because increasing airport capacity can mitigate congestion, this report develops a surface operation model based on analyzed results to achieve this aim. A mixed-integer programming formulation is proposed to optimize total taxi times by finding optimal taxi routes and the related schedules. The model is applied to a sample from real data.

Column Generation Method for U.S. Army Logistics Air Fleet Scheduling

In a U.S. Army problem, an air fleet is scheduled to pick up and deliver military personnel with time window and aircraft capacity constraints. Demand for personnel flights between military bases has a number of unique features, such as priority demand and flexible delivery locations. A column generation method was adopted in which the resource-constrained shortest-path subproblem is solved as a schedule for a fleet vehicle. The master problem solves for an optimal set of schedules for the entire fleet. This column generation method is embedded in an automatic scheduling program (ASP) at the U.S. Joint Operational Support Airlift Center. A comparison of the ASP prototype and the current manual scheduling shows significant improvement potential to the operations for demand satisfaction and resource utilization.

U.S. Postal Airmail Routing Optimization

An airmail routing problem at the United States Postal Service (USPS) is studied in this paper. Each day, millions of airmails of different priorities are collected and shipped between local USPS mail processing and distribution centers (P&DCs) by commercial air carriers. The carriers provide flight capacities and services according to contracts. Each available flight has a finite, maximum capacity for the airmails. In addition, the shipping between P&DCs incurs a cost that is determined by the type of mail, the volume, and the shipping distance. The research objective is to decide airmail allocation onto flights subject to the agreed-on air capacity to minimize system shipping costs while maintaining service quality. An integrated mathematical model is developed to design air routes and to allocate airmails optimally to routes for an improvement on the practice of isolated scheduling at each P&DC. A column generation method is proposed for the model. A numerical test is conducted with operational data. The test not only shows a significant increase in the total volume that can be shipped but also indicates better service in terms of on-time and early deliveries.

Using a Prospect Theory Approach to Studying the Car-Following Model

Car following is a fundamental traffic feature that has been widely studied in literature using vehicles’ speed and acceleration. This study investigates car following from an entirely different perspective, a psychological approach based on the Prospect Theory (PT). PT is a behavioral economic theory that explains human reaction under risk situations. Since car following can be regarded as a risk containing task that addresses the need for balancing safety with travel time reduction, PT is an ideal approach to model car following. Employing PT can provide a spectrum of probabilistic locations of following vehicles in contradiction with traditional methods that define the exact position. This study presents a sensitivity analysis in order to validate the results and calibrate PT’s parameters. The results reveal that PT generates similar probability distributions to the simulation scenarios that proxy the space headway in real situations.

Integrating MTS Commerce Data with Multimodal Freight Transportation Performance Measures to Support MTS Maintenance Investment Decision Making

This report investigates the feasibility of evaluating potential navigation operation and maintenance projects on the Marine Transportation System (MTS) not only as they relate to waterborne commerce, but also in light of the landside freight connections as well. A network optimization model is described that maximizes the multimodal system capacity by choosing the navigation maintenance projects that will either fully accommodate expected demand or provide the greatest potential throughput within overall budget constraints, taking into account the origins and destinations of the commodities that move through a regional multimodal network. Five ports are selected for analysis: Duluth, Minnesota; Hampton Roads, Virginia; Huntington, West Virginia; Plaquemines, Louisiana; and Portland, Oregon.