Anomaly detection method of flight data of unmanned aerial vehicle based on over-sampling projection approximation basis pursuit

Abstract

The invention provides an anomaly detection method of flight data of an unmanned aerial vehicle based on over-sampling projection approximation basis pursuit. The method introduces the signal sub-space projection approximation thought during over-sampling and signal processing and detects anomaly of dataflow by estimating and pursuit matching of base vector direction of the data projection approximation subspace after over-sampling and utilizes the feature of subspace direction insensitive to the switching of flight modes in order to inhibit influence on flight modes on the anomaly detection result. Therefore, a framework for an on-line anomaly detection method for flight data of the unmanned aerial vehicle is provided. The detection framework lowers its false drop data by over 53.82% compared with four on-line anomaly detection methods such as Online-MD, BN, CCA and KOAD during on-line anomaly detection of flight data on the condition of consuming the same Mu-s calculating time. The detection accuracy AUC is improved by over 5.28% and reaches 0.9836, close to the theoretical value of 1. The method is capable of effectively solving the problem of one-line anomaly detection of flight data.

Description

technical field [0001] The invention relates to an abnormal detection method of unmanned aerial vehicle flight data based on oversampling projection approximate basis tracking, and belongs to the technical field of abnormal detection methods of unmanned aerial vehicle flight data. Background technique [0002] The deployment and application of UAV (Unmanned Aerial Vehicle, UAV) needs to ensure sufficient safety for itself, other aircraft, ground facilities and personnel. Sensors are key components for UAVs to perceive their own flight status and interact with the external environment, but they are prone to failure, which can lead to safety accidents such as flight out of control. At the same time, limited by the size, weight and power consumption of drones, it is difficult to use physical redundancy technology in the sensor system, so there is a big gap between drones in terms of reliability and safety compared with manned machines. In order to isolate faults in time and pr...

AI Technical Summary

Problems solved by technology

[0008] The purpose of the present invention is to solve the problems in the above-mentioned prior art, that is, the problem of low accuracy rate of online abnormal detection of UAV flight data caused by flight mode switching

Images