Optimizing Work Zone Schedules Considering Traffic Diversion With Artificial Bee Colony Algorithm

Highway maintenance activities often decrease roadway capacity and intrude traffic movements. The need to finish the project on time and under a specific budget while minimizing the traffic congestion and complying with the emission standards requires an appropriate work zone schedule optimization. The objective of this research is to improve the efficiency of work zone activities and minimize the total project cost including maintenance, user, and emission cost.

While previous studies investigated the work zone optimization problem, they did not consider the implementation of emission standards nor applied a green diversion strategy. This dissertation analyzes the optimization of work zone schedule considering a discrete time-cost relation and a time-dependent traffic flow. The objective function is to minimize the total cost including the emission cost under various realistic constraints. Moreover, the effect of traffic diversion as a congestion mitigation strategy is evaluated under the User Equilibrium (UE) and System Optimum (SO) strategies.

In the developed model many variables interrelate to create a combinatorial optimization problem that is difficult to solve analytically (e.g., the length and duration of work zones, the productivity of the crew, etc.). Consequently, an Artificial Bee Colony (ABC) algorithm is used as a tool to optimize the work zone schedule and minimize the total cost under multiple constraints such as maximum project duration and budget constraint. Traffic diversion is optimized by finding the best diversion rate into the alternative route while considering the delay cost and emission cost on the mainline and alternative route. The emission rates caused by work zone activities and diverted traffic are estimated using the state-of-the-art Motor Vehicle Emission Simulator (MOVES3) developed by the Environmental Protection Agency (EPA). Consequently, two projects are created that illustrate the conditions of the roadways without and during the work zone activities.

Two case studies are presented in this research: Case A and Case B. The purpose of Case A is to validate the applicability of the model whereas Case B proves the ability to optimize real-life work zone projects using different databases under Tier 3 federal emission standards. Sensitivity analyses are conducted to explore the relationships between the model parameters and the decision variables. The results prove the efficiency of ABC in solving the work zone optimization problem and the importance of considering the vehicle emissions during work zone activities.

The developed model can assist transportation agencies in alleviating the congestion and minimizing the total cost considering vehicle emissions and two different traffic diversion strategies. Additionally, the model offers flexibility to investigate the effect of various strategies on the optimal work zone schedule. Hence, the developed model can be applied to evaluate and optimize the work zone schedule in case of a tight schedule or the need to finish the work before a certain event. The model can also suggest a work zone schedule that complies with the federal emission standards.