The invention discloses a vehicle abnormal data screening method for trajectory data, and the method comprises the steps: S10, obtaining a highway vehicle trajectory data set, and carrying out the preprocessing of the vehicle trajectory data set; s20, the preprocessed vehicle track data set is judged, if the sampling time interval is lower than a preset threshold value, the step S30 is executed, and otherwise, the step S40 is executed; s30, carrying out resampling on the vehicle track data set; s40, dividing the vehicle track data set into sub-data sets according to the time features and the space features; and S50, analyzing the data features of the vehicles in each sub-data set, and marking the abnormal index data as abnormal data according to the analysis result of the data features. According to the method, sub-data sets are divided from the trajectory data through time and space features, feature analysis is carried out on three indexes including the vehicle speed, the acceleration and the time headway in the sub-data sets, space-time feature interference is reduced, and abnormal data occurring in the vehicle driving process are screened out.
本发明公开一种面向轨迹数据的车辆异常数据筛选方法,包括:S10,获取高速公路车辆轨迹数据集,对车辆轨迹数据集进行预处理;S20,对预处理后的车辆轨迹数据集进行判断,若采样时间间隔低于预设阈值,则进入步骤S30,否则进入步骤S40;S30,对车辆轨迹数据集进行重采样;S40,按照时间特征和空间特征,将车辆轨迹数据集分成子数据集;S50,分析每一个子数据集中车辆的数据特征,根据数据特征的分析结果,将异常指标数据标记为异常数据。本发明将轨迹数据通过时间、空间特征划分出子数据集,在子数据集中对车辆速度、加速度以及车头时距这三个指标进行特征分析,将时空特性干扰降低,筛选出车辆行驶过程中出现的异常数据。
Trajectory data-oriented vehicle abnormal data screening method
一种面向轨迹数据的车辆异常数据筛选方法
2023-07-21
Patent
Electronic Resource
Chinese
Vehicle abnormal behavior detection method and system based on trajectory data
European Patent Office | 2024
|Real-time abnormal trajectory detection method for urban traffic data
European Patent Office | 2023
|Abnormal vehicle data identification method, device and equipment
European Patent Office | 2023
|European Patent Office | 2017
|European Patent Office | 2021
|