The invention relates to an airport scene semantic trajectory representation and taxiing mode anomaly detection method, which belongs to the field of airport operation management, and is characterized in that semantic transformation is realized by combining scene map topological information based on aircraft motion characteristics, and airport scene aircraft semantic trajectory representation is obtained; on the basis, a track similarity matrix is constructed by using a method for improving a longest common subsequence by distributing weights through road section geographic information, and scene sliding mode recognition is realized by using hierarchical clustering; and finally, obtaining a maximum outlier for judging the abnormal mode by calculating the internal similarity of the identified modes, calculating the similarity between the follow-up input track and each mode, and comparing the maximum outlier to realize the abnormal detection, early warning of the abnormal track in advance, maintenance of the scene scheduling safety and improvement of the scene scheduling efficiency.
本发明涉及一种机场场面语义轨迹表示与滑行模式异常检测方法,属于机场运行管理领域,基于航空器运动特征结合场面地图拓扑信息实现语义转换,得到机场场面航空器语义轨迹表示;在此基础上,利用路段地理信息分配权重改进最长公共子序列方法构建航迹相似度矩阵,并使用分层聚类实现场面滑行模式识别;最终,通过计算识别出的模式的内部相似度得到判断模式异常的最大离群值,将后续输入航迹与各模式计算相似度,通过对比最大离群值实现异常检测,提前预警异常航迹,维护场面调度安全,提高场面调度效率。
Airport scene semantic trajectory representation and sliding mode anomaly detection method
一种机场场面语义轨迹表示与滑行模式异常检测方法
2024-04-30
Patent
Electronic Resource
Chinese
Dynamic scene modelling and anomaly detection based on trajectory analysis
Wiley | 2014
|Trajectory anomaly detection system and online trajectory anomaly detection method
European Patent Office | 2024
|Learning Semantic Scene Models by Trajectory Analysis
British Library Conference Proceedings | 2006
|