基于时间序列预测的停车泊位调度算法  

PARKING SPACE SCHEDULING ALGORITHM BASED ON TIME SERIES PREDICTION

作　　者：宗学森 董晓飞 李鹏 熊晓芸 王金龙 Zong Xuesen;Dong Xiaofei;Li Peng;Xiong Xiaoyun;Wang Jinlong(School of Information and Control Engineering,Qingdao University of Technology,Qingdao 266520,Shandong,China;Qingdao Yilian Information Technology Co.,Ltd.,Qingdao 266011,Shandong,China)

机构地区：[1]青岛理工大学信息与控制工程学院,山东青岛266520 [2]青岛亿联信息科技股份有限公司,山东青岛266011

出　　处：《计算机应用与软件》2024年第9期294-303,共10页Computer Applications and Software

基　　金：国家自然科学基金项目(62001262);山东省重点研发计划项目(2019GGX101017)。

摘　　要：针对现有停车方法存在停车冲突和停车泊位利用率低问题,提出一种基于时间序列预测的停车泊位调度算法。算法利用Holt-Winters模型预测月租车辆在下一时段的停车流量;采用两种停车泊位调度策略,即月租服务优先策略和动态分配策略,在保证月租车辆的停车服务质量的前提下,动态规划可用停车泊位的泊位类型。实验结果表明,在不同数量的停车泊位和月租用户条件下,采用该方法比现有停车方法具有更高的累积利用率,可以有效提高写字楼停车场的利用率。In view of the problems of parking conflicts and low parking space utilization in existing parking methods,a parking space scheduling algorithm based on time series prediction is proposed.The algorithm used the Holt-Winters model to predict the parking flow of monthly rental vehicles in the next period.It used two parking scheduling strategies,namely monthly rental service priority strategy and dynamic allocation strategy.It realized the dynamic planning of the berth types of available parking berths under the premise of ensuring the parking service quality of monthly rental vehicles.Experimental results show that under the conditions of different numbers of parking spaces and monthly rented users,this method has a higher cumulative utilization rate than the existing parking methods,and can effectively improve the utilization rate of the office building parking lot.

关 键 词：时间序列预测 Holt-Winters 停车流量预测 停车泊位调度 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]