ICUAS 2020 Paper Abstract

Close

Paper WePS.9

Ahn, Hyojung (Korea Aerospace Research Institute)

Deep Learning Based Anomaly Detection for a Vehicle in Swarm Drone System

Scheduled for presentation during the Poster Session "Poster Papers Session" (WePS), Wednesday, September 2, 2020, 13:00−18:00, Foyer, Mezzanine Level

2020 International Conference on Unmanned Aircraft Systems (ICUAS), September 1-4, 2020 (Postponed from June 9-12, 2020), Athens, Greece

This information is tentative and subject to change. Compiled on March 29, 2024

Keywords Swarms, Reliability of UAS, Control Architectures

Abstract

In this paper, we perform the actual verification of the anomaly detection (AD) model of each drone that indicates the anomaly in swarm drone flight using the actual flight data. For this purpose, we use a model-based AD method that uses data accumulated through actual flight tests. The AD model uses a deep neural network-based generation model to create a training model with normal data and perform tests with abnormal data. As a result, the diagnostic results of mainly three cases are derived and analyzed for validity. The proposed AD method can be integrated with a machine learning based framework that can immediately detect abnormal behavior of swarm drone flights, which can be utilized to improve the reliability of swarm drone flight operations.

 

 

All Content © PaperCept, Inc.

This site is protected by copyright and trademark laws under US and International law.
All rights reserved. © 2002-2024 PaperCept, Inc.
Page generated 2024-03-29  01:20:08 PST  Terms of use