Yingyan Lou

Robust Signal Timing for Arterials Under Day-to-Day Demand Variations

This paper formulates a scenario-based stochastic programming model to optimize the timing of pretimed signals along arterials under day-to-day demand variations or future uncertain traffic growth. Demand scenarios and their corresponding probabilities of occurrence are introduced to represent the demand uncertainty. On the basis of a cell-transmission representation of traffic dynamics, cycle length, green splits, phase sequences, and offsets are determined to minimize the expected delay incurred by high-consequence demand scenarios. A simulation-based genetic algorithm is proposed to solve the model, and a numerical example is presented to verify and validate the model.

Robust approach to discrete network designs with demand uncertainty

A discrete network design problem under demand uncertainty is considered. It is assumed that the future travel demand for each origin-destination (O-D) pair can take on nominal or one of several other values, where the probabilities of occurrence of these values are unknown. However, the vector of O-D demands realized for the network must belong to an uncertainty set, a set that allows demands for a group of O-D pairs to deviate from their nominal values simultaneously. A robust counterpart of deterministic discrete network design is formulated as a mathematical program with complementarity constraints under the Wardropian user equilibrium conditions. The algorithm proposed for the problem terminates in a finite number of iterations and converges to a global optimum solution under certain conditions. Numerical results that use two networks from the literature empirically demonstrate that the algorithm is effective and has the potential to solve realistic problems.

Calibration and Development of Safety Performance Functions for Alabama

One critical component of the Highway Safety Manual (HSM) statistical methods is the safety performance function (SPF). SPFs are essentially regression models that correlate quantitatively the expected number of crashes with traffic exposure and geometric characteristics of the road. As part of a project performed by the University of Alabama to facilitate implementation of the new HSM procedures in the state, this study aims to evaluate the applicability of HSM predictive methods to Alabama data and to develop state-specific statistical models for two facility types: two-lane, two-way rural roads and four-lane divided highways. This study first calibrates HSM base SPFs by using two approaches: the method recommended by the HSM and a newly proposed approach that treats the estimation of calibration factors as a special case of a negative binomial regression. In addition, new forms of state-specific SPFs are further investigated by using Poisson-gamma regression techniques. Four new functional forms are studied in this project. The prediction capabilities of the two calibrated models and the four newly developed state-specific SPFs are evaluated with a validation data set. Five performance measures are considered for model evaluation. The study is able to identify a particular state-specific SPF that fits the Alabama data well and outperforms other models, including the calibrated SPFs. The best model describes the mean crash frequency as a function of annual average daily traffic, segment length, lane width, year, and speed limit. The study finds that the HSM-recommended method for calibration factor estimation also performs well.

Defending transportation networks against random and targeted attacks

This paper explores several reliability and vulnerability measures for transportation networks and proposes three models for optimal resource allocation for transportation network design or defense to minimize the disruption caused by both random and targeted attacks. The common day-to-day disturbances with less severe consequences are referred to as random attacks, but targeted attacks include both coordinated terrorist strikes and large-scale natural disasters. For random attacks, the major concern would be the reliability of the total system travel time. A robust discrete network design problem is formulated to take into account random attacks in the planning stage. The transport capacity or the unsatisfied demand would be critical in case of emergency evacuation, and law enforcement forces could be deployed to prevent malicious attacks in the first place or to ensure a smooth evacuation operation. The proposed models feature an intrinsic trilevel game structure of the network users, the attacker, and the defender (planner). By exploring the unique properties of the proposed measures and reformulating the problems, the trilevel structure models are reduced to mixed-integer semi-infinite optimization programs. This paper further applies an active-set algorithm, combined with a cutting-plane scheme to solve the proposed models. Numerical examples indicate that the proposed formulations are valid and that the solution algorithm can solve the problems effectively and efficiently. The models for targeted attacks provide practical implications on identifying critical infrastructures for evacuation.

Safety performance function development for analysis of bridges

Bridges are an integral infrastructure component and, as such, have been the subject of extensive research efforts related to structural performance. However, there has been little study on the traffic safety performance of bridges, which have very different physical and operational characteristics compared with regular roadway facilities. This study develops safety performance functions (SPFs) for overall vehicle crashes and single-vehicle crashes occurring on major highway bridges in Alabama. The bridge characteristic data and crash information are obtained from three different databases. Geographic information systems (GIS) are used to spatially represent bridges as vectors and associate crashes to the bridges based on location attributes from the crash data. SPFs of several functional forms are developed and investigated for identifying the best model using negative binomial regression. The models are validated by comparing their relative predictive capabilities. This paper recommends models that fit the Alabama data well. These models can be used for estimating the expected number of crashes on bridges along major highways in Alabama.