Parking pricing strategies in the era of autonomous vehicles

Conventional parking behavior could be completely changed with the arrival of autonomous vehicles (AVs). Unlike conventional vehicles (CVs), which are restricted to parking in specified lots, AVs eliminate the limitation of human driving and provide travelers with a wider range of parking options. This study examines the distinct parking choice behaviors of AVs and CVs and develops pricing strategies for parking lots. To this end, parking choice models for AVs and CVs are separately developed based on their respective parking costs. An agent-based simulation framework is constructed to simulate the parking processes of both vehicle types and capture the macroscopic impacts of individual parking decisions on system. A genetic algorithm (GA) optimization approach is applied to determine the optimal parking pricing. The results indicate that when AVs and CVs each account for 50% of the market, distinct pricing strategies are required to achieve specific objectives. Specifically, under the objective of minimizing vehicle kilometers of travel (VKT), high parking fees should be imposed on parking lots in the city center to guide them toward adjacent areas with low prices. Similarly, achieving vehicle travel time (VTT) minimization requires implementing extremely low prices for peripheral parking lots to attract vehicles, thereby reducing congestion in central areas. Notably, the pricing strategies for VTT minimization and total utilization time (TUT) maximization share certain similarities under this equal-proportion scenario. Moreover, if the penetration rate of CVs exceeds that of AVs, parking revenue can be maximized by establishing high prices across all parking lots. Conversely, when AVs become dominant, it is necessary to lower parking prices appropriately to encourage AVs to utilize parking lots.