Online Pricing for Balancing Parking Demand and Supply in a Shared Parking Case

This paper investigates the online pricing problem for balancing parking demand and supply in a case of shared parking, which aim to enhance the efficient utilization of underused parking spaces by connecting drivers seeking parking with parking slot owners willing to share their spaces. A key challenge for such case lies in designing parking prices for drivers and subsidies for parking slot owners that incentivize both sides to participate, thereby maximizing the total revenue over the operational horizon. Given the lack of prior knowledge of the demand and supply distributions, which depend on the parking prices and subsidies respectively, we propose a real-time decision-making framework that jointly determines parking prices and slot subsidies while simultaneously learning the unknown demand and supply distributions. We formulate the problem as a combinatorial multi-armed bandit (CMAB) model and develop a data-driven online learning algorithm based on the upper confidence bound (UCB) method. Numerical experiments validate the efficiency and effectiveness of the proposed online pricing algorithm in balancing parking demand and supply while optimizing platform revenue.