A Review of Data-driven UAV Anomaly Detection Algorithms
With the continuous development of Unmanned Aerial Vehicle(UAV)integration and intelligence,UAV are widely used in military and civilian fields.Therefore,higher requirements are placed on the safe flight of UAV,and UAV anomaly detection has an important role in ensuring safe flight and reducing economic losses.In recent years,the advantages of data-driven methods in feature extraction,nonlinear problem solving,and accuracy have made them the mainstream algorithms for UAV anomaly detection.The types of UAV anomalies and the characteristics of the anomaly data are analyzed and summarized.Current research status of UAV data-driven anomaly detection algorithms at home and abroad is summarized.The three aspects of UAV anomaly detection from supervised learning,semi-supervised learning and unsupervised learning are summarized and concluded,and the advantages and disadvantages of various algorithms are analyzed.The development trend of future UAV anomaly detection is prospected in view of current research status of existing algorithms,aiming to provide reference for the subsequent related research.
国家科技期刊平台
NSTL
万方数据
