一种基于交通大模型的异常交通事件识别方法及系统
By using a traffic model-based method for identifying abnormal traffic events, real-time data streams and dynamic baselines are utilized, combined with K-means clustering and trajectory deviation analysis, to generate early warning signal sequences. This solves the problem of insufficient early warning capabilities in existing technologies and achieves accurate identification and efficient emergency response.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHENZHEN TUOBIDA TECH CO LTD
- Filing Date
- 2025-12-08
- Publication Date
- 2026-07-17
AI Technical Summary
Existing traffic incident identification technologies lack early warning capabilities, have limited coverage, and use a single data dimension. They are unable to capture the deep connections between traffic participants, and their reliance on fixed-frequency data collection and manual response leads to delays in anomaly identification and false alarms.
The abnormal traffic event identification method based on a large traffic model acquires real-time traffic data streams, extracts abnormal feature vectors, performs grouping and trajectory deviation calculations, establishes mapping relationships, generates warning signal sequences, determines the type of abnormal event through time-series matching and chain integrity analysis, and finally pushes the data to the traffic management platform to obtain emergency response instructions.
It enables accurate identification of abnormal traffic events, reduces false alarm rates, improves early warning capabilities and emergency response efficiency, and avoids the problems of manual intervention and information gaps in traditional solutions.
Smart Images
Figure CN121505876B_ABST