Designing intelligent public parking locations for autonomous vehicles

Abstract

Looking for parking spaces in crowded areas can be a stressful and time-consuming challenge. With the arrival of technologies like autonomous vehicles (AVs), users can park far away from their destinations. AVs can move from the city center, which has high congestion, into a parking spot outside the area, and thus have more flexibility in choosing the parking location. This study aims to provide a rent bid for the daytime parking of AVs that considers urban land use to evaluate parking strategies possibly chosen by AV users. We determine an actual parking demand function by integrating individual preferences into a p-median problem that controls user optimality. Combining all these together, a novel dynamic optimization formulation is proposed to design the location of parking facilities for AVs that examines the driver's parking preference including rent bid, waiting time for searching parking lots, and travel costs. A Lagrangian relaxation (LR) algorithm is presented to solve the stochastic parking location problem that integrates a reliability strategy to balance demand and supply for parking spaces. The average gap between the exact and the LR solution is about 6% which is very reasonable. We provide a detailed case study of our model with real data generated from daily household travel in New York City. The results show that the average parking price is reduced by 34%, and the average empty trips by AVs decrease by 22% using the proposed public parking strategy.