摘要
使用上门查询、蹲点守候等人工手段对公务用车私用进行监管,效率低下且成本高昂。为自动、客观管理公务车辆,挖掘和监控公车私用行为,提出一种基于fastMCD和DBSCAN的公车私用监管模型。利用fastMCD算法对于异常检测的鲁棒性,提取公务车辆使用时长和行驶里程中的异常联系,并结合历史北斗定位数据,建立DBSCAN异常轨迹检测模型,计算公务车辆行驶轨迹的异常程度。实验结果表明,该模型可以较为有效地解决公车私用监管的问题。
The conventional method uses the manual means such as on-site inquiry and waiting to supervise the private use of official vehicles,which is inefficient and costly.In order to automatically and objectively manage official vehicles,excavate and monitor the private behavior of buses,a public vehicle supervision model based on fastMCD and DBSCAN is proposed.By using the fastness of the fastMCD algorithm for anomaly detection,extracting the abnormal relationship between the duration of the official vehicle and the mileage,and combining the historical Beidou positioning data,a DBSCAN anomaly trajectory detection model is established to calculate the abnormality of the trajectory of the official vehicle.The experimental results show that the model can solve the problem of bus private supervision more efficiently.
作者
聂启阳
朱峰
NIE Qiyang;ZHU Feng(State-owned Assets Management Center of Sichuan Provincial Government Offices,Chengdu 610000,China)
出处
《技术与市场》
2019年第10期176-180,共5页
Technology and Market
基金
四川省科技计划项目(2007FZ0073)
关键词
公车私用
数据挖掘
fastMCD
异常检测
anomaly detection
data mining
fast MCD
private use of official vehicles
