首页|基于协同过滤的空载电动汽车充电桩推荐方法

基于协同过滤的空载电动汽车充电桩推荐方法

扫码查看
近年来政府大力推广电动汽车,其渗透率越来越高,但存在用户有充电需求时难以找到充电桩和充电导航的问题.因此,结合充电站的运行数据,利用深度置信网络预测到达充电站的车辆数量,并结合用户的历史充电数据,利用协同过滤算法为用户推荐空闲充电桩.综合考虑用户偏好、行驶距离和等待时长,为用户提供充电导航服务.通过算例验证了所提方法在节省时间、降低充电桩闲置率和电动汽车排队率方面的有效性.
Recommendation Method of Idle Charging Piles for EV Based on Collaborative Filtering
In recent years,electric vehicles are vigorously promoted by the government,and the penetration rate is getting high-er.However,it is difficult for users to find charging piles and charging navigation when there is a demand for charging.There-fore,combined with the operation data of the charging station,this paper uses the deep belief network(DBN)algorithm to pre-dict the number of vehicles arriving at the charging station,and combined with the user's historical charging data,uses the col-laborative filtering algorithm to recommend idle charging piles for users.User preference,driving distance and waiting time are comprehensively considered to provide users with charging navigation services.Several example systems are tested,and the re-sults verify the effectiveness of the proposed method in saving time,reducing the idle rate of charging piles and the queuing rate of electric vehicles.

recommended charging pilecollaborative filtering algorithmpath planninguser preference

汪应春、王庆、彭涛、明东岳、魏伟、叶利

展开 >

国网湖北省电力有限公司营销服务中心(计量中心),湖北,武汉 430080

国网电力科学研究院武汉南瑞有限责任公司,湖北,武汉 430078

充电桩推荐 协同过滤算法 路径规划 用户偏好

2024

微型电脑应用
上海市微型电脑应用学会

微型电脑应用

CSTPCD
影响因子:0.359
ISSN:1007-757X
年,卷(期):2024.40(12)