ICUAS'17 Paper Abstract


Paper WeA3.2

Noble, Deleena (Cal Poly Pomona), Bhandari, Subodh (Cal Poly Pomona)

Neural Network Based Nonlinear Model Reference Adaptive Controller for an Unmanned Aerial Vehicle

Scheduled for presentation during the "Control Architecture - I" (WeA3), Wednesday, June 14, 2017, 10:20−10:40, Salon CD

2017 International Conference on Unmanned Aircraft Systems, June 13-16, 2017, Miami Marriott Biscayne Bay, Miami, FL,

This information is tentative and subject to change. Compiled on April 12, 2021

Keywords Control Architectures, Autonomy, Simulation


This paper presents a neural network based model reference adaptive controller (MRAC) for the control of a fixed-wing unmanned aerial vehicle (UAV). An adaptive neural network is trained using the error between the UAV response and the desired response as given by the reference model. The design of a suitable reference model for the desired aircraft performance is investigated and developed. Unknown nonlinearities of the vehicle dynamics not accounted for by the reference model are compensated in real-time by the adaptive neural network approximation, which also provides online adaptation during off-nominal flight conditions. The nonlinear dynamics of a twin-engine UAV are modeled in Simulink to test the controller in a software-in-the-loop simulation environment. Simulation results for a number of flight maneuvers show the feasibility and performance of the proposed controller.



All Content © PaperCept, Inc.

This site is protected by copyright and trademark laws under US and International law.
All rights reserved. © 2002-2021 PaperCept, Inc.
Page generated 2021-04-12  05:30:31 PST  Terms of use