The invention discloses a spacecraft intelligent anomaly detection method supporting a multi-source mixed heterogeneous data type, and the method comprises the steps: firstly, employing a discrete data dictionary to detect data anomaly through employing a reconstruction residual error for discrete data; then, aiming at continuous data, pruning a search space by using an analysis result of discrete data to detect anomalies; and finally, aiming at the mixed type of data, adjusting an algorithm optimization target by combining the two methods, and realizing rapid and accurate detection on the anomaly. Compared with a traditional method, the method provided by the invention fully explores the relation between data sparse representation and abnormal characteristics according to the discrete data processing requirements of autonomous health management and intelligent precise operation and maintenance scenes of spacecrafts, and establishes an anomaly detection method supporting discrete signals. In addition, based on the idea of discrete and continuous data fusion, an anomaly detection algorithm based on sparse representation is designed, and anomaly feature detection on two types of heterogeneous data at the same time is supported.
本发明公开了一种支持多源混合异构数据类型的航天器智能异常检测方法,首先,针对离散型数据,通过采用离散型数据字典使用重构残差检测数据异常;然后,针对连续型数据,使用离散数据的分析结果剪枝搜索空间来检测异常;最后,针对混合类型的数据,结合以上两种方法调整算法优化目标,对异常实现快速精准检测。而与传统的方法相比,本发明专利针对航天器自主健康管理与智能精准运维场景的离散数据处理需求,充分发掘数据稀疏表示和异常特征间的关系,建立支持离散信号的异常检测方法。另外本发明基于离散和连续数据融合的思想,设计基于稀疏表示的异常检测算法,支持同时对两种类型异构数据的异常特征检测。
Spacecraft intelligent anomaly detection method supporting multi-source mixed heterogeneous data type
支持多源混合异构数据类型的航天器智能异常检测方法
22.11.2024
Patent
Elektronische Ressource
Chinesisch
IPC: | B64F GROUND OR AIRCRAFT-CARRIER-DECK INSTALLATIONS SPECIALLY ADAPTED FOR USE IN CONNECTION WITH AIRCRAFT , Boden- oder Flugzeugträgerdeckeinrichtungen besonders ausgebildet für die Verwendung in Verbindung mit Luftfahrzeugen / G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung |
NOAA's spacecraft anomaly data base
AIAA | 1990
|Multi-source heterogeneous sensor data processing method for intelligent driving
Europäisches Patentamt | 2023
|Spacecraft Anomaly Detection via Transformer Reconstruction Error
Springer Verlag | 2020
|Spacecraft Anomaly Detection via Transformer Reconstruction Error
British Library Conference Proceedings | 2020
|SADDLE: Spacecraft Anomaly Detection using Deep Learning
IEEE | 2024
|