Back to Search
Start Over
Integrated method for the UAV navigation sensor anomaly detection
- Source :
- IET Radar, Sonar & Navigation. 11:847-853
- Publication Year :
- 2017
- Publisher :
- Institution of Engineering and Technology (IET), 2017.
-
Abstract
- The rapid development of unmanned aerial vehicles (UAVs) has made great progress for its widespread uses in military and civilian applications in recent years. On-board integrated navigation sensors are essential for UAV flight control systems in that they must operate with robustness and reliability. To achieve this, timely and effectively anomaly detection capabilities for the estimated UAV status from the integrated navigation sensors are required to ensure the UAV flight safety. Extraction of the anomaly information from the real-time navigation sensors and designing a robust and reliable anomaly detection algorithm are major issues for the UAV navigation sensor anomaly detection. This study introduces a novel integrated algorithm for detecting UAV on-board navigation sensor anomaly, by combining particle filter (PF) estimated state residuals with fuzzy inference system (FIS) decision system. The residual information is obtained based on the difference between the collected Global Positioning System measurements and high accuracy PF estimates. The indicators derived from the PF residuals are further made as inputs for the FIS system to output the different anomaly levels. The simulation and filed test results have demonstrated the effectiveness and efficiency of the proposed anomaly detection method in terms of timeliness, recall and precision.
- Subjects :
- Computer science
business.industry
ComputerApplications_COMPUTERSINOTHERSYSTEMS
020206 networking & telecommunications
02 engineering and technology
Residual
Robustness (computer science)
Control system
0202 electrical engineering, electronic engineering, information engineering
Global Positioning System
Flight safety
020201 artificial intelligence & image processing
Anomaly detection
Computer vision
Artificial intelligence
Electrical and Electronic Engineering
Precision and recall
Particle filter
business
Subjects
Details
- ISSN :
- 17518792
- Volume :
- 11
- Database :
- OpenAIRE
- Journal :
- IET Radar, Sonar & Navigation
- Accession number :
- edsair.doi...........f730e147901a46bf2b3f28badad6800b