File(s) under embargo
until file(s) become available
PARKING FACILITY LOCATION AND USER PRICING IN THE ERA OF AUTONOMOUS VEHICLE OPERATIONS
thesisposted on 10.12.2020, 21:22 by Mahmood Tarighati TabeshMahmood Tarighati Tabesh
Parking continues to pose a frustrating problem for travelers and commuters at large metropolitan areas. Despite a significant number of parking spaces, as they struggle to find appropriate parking spots and in doing so, waste enormous amounts of time and money finding and using available parking. With the looming advent of autonomous vehicles (AVs), there is a great opportunity to identify sustainable solutions to the parking problem because AV travelers can travel directly to their destinations to drop them off and then proceed to park at a relatively distant but lower-priced parking facility. Hence, the parking demand in downtown areas is expected to drop significantly as the market penetration rates of AVs increase. This could lead to the decommissioning and repurposing of some existing parking facilities in the downtown areas. However, this raises some social inequity concerns regarding the parking needs of human-driven vehicle (HDV) travelers particularly if the parking facilities are decommissioned at a time of low AV penetration. This thesis presents and demonstrates a comprehensive bi-level optimization framework for locating/relocating/decommissioning, and pricing parking facilities to serve a mixed fleet of AVs and HDVs in long-term. In the upper-level, the transportation decision-maker seeks to minimize the total travel cost of the travelers, to maximize the total parking fee revenues, and to maximize the monetary benefits of decommissioning and repurposing the existing parking facilities. In the lower-level, the AV and HDV travelers seek to minimize their travel cost given the decisions made by the transportation decision-makers in the upper-level. The problem is formulated mathematically as a mixed-integer nonlinear program and is solved using a hybrid approach consisted of machine learning and optimization heuristics. The numerical results indicate that the algorithm is capable of solving the problem in an efficient manner. It is found that as the budget for constructing new parking facilities in the outskirts increases, total cost of the travelers increases since the downtown parking facilities can be decommissioned at faster paces. Also, even without any new parking facility construction, it is possible to decommission some of the existing parking facilities at a time of high AV penetration due to the AV’s requirement of smaller parking spaces compared to HDVs. Further, we found that in high construction budget levels, it is recommended to construct large-sized parking facilities in the outskirts but relatively close to the downtown area, and then, construct small-sized parking facilities in relatively farther locations to fulfill the parking needs of AVs. The numerical results also suggest that similar parking facility decommissioning plans are proposed when the monetary benefits of decommissioning each parking facility is greater than a relatively low threshold. Further, despite the increases in the total travel demand, higher AV penetration rates result in lower total travelers’ costs and parking fee revenues.