Rebalancing the car-sharing system: A reinforcement learning method

With the boom of sharing economy, more and more car-sharing corporations sprout up, providing more travel options and convenience. Due to similar travel patterns of urban dwellers, the car-sharing system results in an imbalance of shared cars in spatial distribution, especially during the rush hours. To redress this imbalance faces many challenges, such as insufficient data and the enormous state space. In this study, we propose a new reward method called Double P (Picking & Parking) Bonus (DPB). We model the research problem as a Markov Decision Process (MDP) problem and introduce Deep Deterministic Policy Gradient, a state-of-the-art reinforcement learning framework, to find a solution. The results show that the rewarding mechanism embodied in the DPB method can indeed guide the users' behaviors through price leverage, increase user stickiness, cultivate user habits, and thus boost the service provider's long-term profit.