SPA: Smart Parking Algorithm Based on Driver Behavior and Parking Traffic Predictions

Abstract

Smart parking problems have received much attention in recent years. In literature, many smart parking allocation algorithms that considered the parking grid reservation and recommendation have been proposed. However, the parking policies for maximizing parking rate and benefits still can be improved. This paper proposes a smart parking allocation algorithm (SPA), which aims to maximize the benefits created by a given parking lot while guaranteeing the quality of parking services. The proposed SPA algorithm predicts the driver behavior and estimated parking traffic in the near future based on the historical parking records. These predictions help SPA to better match the parking demands and the resource of available parking grids and, hence, improve the utilization and the created benefit of each parking grid. The proposed SPA applies three policies, namely worst-fit (WF-SPA), best-fit (BF-SPA), and parking behavior forecast (PBF-SPA), to allocate the available grids to the vehicles. Performance evaluations reveal that the proposed SPA outperforms exiting work in terms of accumulated parking rate and service quality and, hence, improves the benefits of a given parking lot.