This dissertation models and simulates optimization problems to find the optimal set of strategies to allocate network space (lanes or zones) to autonomous vehicles (AVs) in partially automated dynamic networks. This dissertation aims to develop a set of methods to determine the appropriate AV managed space strategies. Managed space strategies discussed in this study are allocating lanes to AVs throughout the highway network and determining areas of the network that AVs are allowed to operate/serve. The proposed methodologies determine the layout of dedicated network resources for AVs and other AV-related policies (e.g., a higher speed limit on AV exclusive lanes) to maximize the overall performance of the transportation system. In this dissertation, two main problems are solved that guide how to deploy AV managed space strategies and other AV-related strategies to amplify AVs' benefits and improve the overall system's performance. The first problem is a network lane allocation to AVs problem in a partially automated network. The second problem is a dynamic mixed fleet assignment problem for a shared-use mobility service (SMS) with zone-level restrictions defined based on the vehicle type. Bi-level mathematical programs are formulated for both problems, and the genetic algorithm (GA) as a heuristic approach is employed to solve the problems. Systematic approaches are designed to reduce the solution space of the optimization problem to achieve faster convergence of the GA. The fundamental idea for the GA-based approach in both problems is to define the decision variables in the upper-level problem as chromosomes. The solution for each problem is obtained through an iterative process that involves reproduction, crossover, and mutation operations. The results generated throughout the GA-based solutions for both problems emphasize the importance of choosing the appropriate allocation of network resources to AVs to improve the system performance. High-quality solutions attained by the GA for the network resource allocation problems can unlock the full potential of AVs.

Creator

• Hosseini, Moein

DOI

• https://doi.org/10.21985/n2-6rkm-4n48

Subject

• Transportation

Language

• en

Alternate Identifier

• etdadmin_upload_929450
• http://dissertations.umi.com/northwestern:16274

Date created

• 2022-01-01

Resource type

• Dissertation

Rights statement

• In Copyright