实时停车可用性预测

Real-time Parking Availability Predictions

针对目前机动车数量的急剧增长,在停车抉择过程中很难获得实时停车信息,提出一种基于智能停车预约(IPR)架构的停车设施信息系统的预测停车可用性方法。该方法利用实时可用性预测(RAF)算法评价停车请求,采用聚合方法迭代分配停车请求;并通过实际场景验证。结果表明该聚合方法的耗时更短;实际可用性与RAF算法预测可用性比较结果也表明采用RAF算法预测停车可用性的平均误差更小。

In view of the present sharp increase number of vehicles,it is difficult to obtain real-time parking information in the process of parking choice.A methodology is proposed for predicting real-time parking availability in Intelligent parking reservation （IPR） architectures.This method uses the real-time availability forecast （RAF） algorithm to evaluate parking request and employ an aggregated approach to allocate iteratively parking requests.And it was verified by a real case scenario.The results show that the aggregation method is less time-consuming,by comparing the actual availability and RAF algorithm to predict availability.The result also shows that the average error using RAF algorithm to predict parking availability is smaller.