The invention discloses a construction method of an urban traffic anomaly detection model considering a road network topological structure and road attributes. The method comprises: using a GRU-GCN model to predict speed data, combining the predicted speed data and real speed data, using an HTM-Detector algorithm, outputting the time range of traffic abnormal events on a road, and calculating theaccuracy. The method further improves the accuracy of the traffic prediction speed, and has the beneficial effect of providing more reliable data support for urban traffic anomaly detection.
本发明公开了一种兼顾路网拓扑结构和道路属性的城市交通异常探测模型的构建方法,包括:利用GRU‑GCN模型预测速度数据,结合预测速度数据和真实速度数据,采用HTM‑Detector算法,输出道路上的交通异常事件的时间范围,并计算准确率。本发明进一步提升交通预测速度的准确率,具有为城市交通异常探测提供更可靠数据支撑的有益效果。
Construction method of urban traffic anomaly detection model
一种城市交通异常探测模型的构建方法
2020-09-11
Patent
Elektronische Ressource
Chinesisch
Europäisches Patentamt | 2023
|Traffic violation model training process anomaly detection method
Europäisches Patentamt | 2024
|