基于细粒度时区的泊位共享动态定价研究

Dynamic Pricing Strategy Based on Granular Space-time

随着共享经济的蓬勃发展，作为缓解城市拥堵有效途径的泊位共享应运而生，它能够在很大程度上提高泊位闲置资源的利用率，从而解决停车难问题．但现有的定价策略难以合理调度泊位资源，基于粗粒度的分区泊位定价模式也存在着较为明显的缺陷，因而造成了大量泊位资源的浪费．为此，提出了满足细粒度时区的动态定价策略，通过建立泊位价格和泊位空闲率之间的关系模型，将某一区域内的泊位使用率控制在一定阈值．以武汉某大学校区内的所有停车场为例，结合离散型选择模型和优化后的RAF算法，旨在通过价格调度使得单一时间段内每个停车场的使用率达到理想值，从而更好地实现泊位资源的最优化配置．

With the booming development of sharing economy, parking space sharing comes into being which can ease urban traffic congestion effectively and improve the utilization of parking resources to a great degree so as to solve parking problems. However, nowadays, there are obvious defects in the pricing strategy and pricing model based on regions which account for the huge waste of parking space resources. Thus, we propose a dynamic pricing strategy based on granular space-time and establish a relational model between parking price and occupancy rate. Taking all parking spaces in Wuhan University of Technology （Jian Hu campus） as examples, In this paper and an optimized real-time availability forecast we use a calibrated discrete choice model （RAF） algorithm. By adjusting the parking prices of different areas in a certain time, we aim at optimizing the occupancy rate of every parking spaces, thus achieving the best utilization of parking resources.