Integrated Control Method for STOVL UAV Based on RBF Neural Network and Nonlinear Dynamic Allocation

A short takeoff and vertical landing unmanned aerial vehicle (STOVL UAV) is significantly influenced by factors such as the ship’s surface effect, deck motion, and jet effect during vertical landing on an aircraft carrier. The existing control logic cannot effectively solve the coupling problem of l...

Full description

Saved in:
Bibliographic Details
Main Authors: Shilong Ruan, Shuaibin An, Zhe Dong, Zeyu Jin, Kai Liu
Format: Article
Language:English
Published: MDPI AG 2025-02-01
Series:Drones
Subjects:
Online Access:https://www.mdpi.com/2504-446X/9/3/167
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850089697687109632
author Shilong Ruan
Shuaibin An
Zhe Dong
Zeyu Jin
Kai Liu
author_facet Shilong Ruan
Shuaibin An
Zhe Dong
Zeyu Jin
Kai Liu
author_sort Shilong Ruan
collection DOAJ
description A short takeoff and vertical landing unmanned aerial vehicle (STOVL UAV) is significantly influenced by factors such as the ship’s surface effect, deck motion, and jet effect during vertical landing on an aircraft carrier. The existing control logic cannot effectively solve the coupling problem of longitudinal attitude and trajectory, so it is hard to guarantee the stability and control accuracy of the UAV at low speed. To address the aforementioned interference and coupling problems, a comprehensive control law based on a radial basis function neural network (RBFNN) and nonlinear dynamic optimal allocation is designed in this paper. Firstly, the integrated landing control law of the STOVL UAV is designed. Considering the model uncertainty and complex landing environment, an RBFNN is used for online observation and compensation to improve the robustness of the system. Subsequently, a dynamic control allocation module based on nonlinear optimization is developed to simultaneously satisfy force and moment commands. The simulation results show that the integrated control method effectively decouples the pitch attitude and longitudinal trajectory at low speeds, resulting in effective convergence control of pitch angle, forward flight speed, and altitude. The integration of the RBFNN, as evaluated by the integral of absolute error (IAE), results in a 93% improvement in control accuracy compared to the integrated landing control law designed in this paper without the RBFNN integration.
format Article
id doaj-art-3dbdfaa23f9a49a18b213b7cc2ee77b0
institution DOAJ
issn 2504-446X
language English
publishDate 2025-02-01
publisher MDPI AG
record_format Article
series Drones
spelling doaj-art-3dbdfaa23f9a49a18b213b7cc2ee77b02025-08-20T02:42:42ZengMDPI AGDrones2504-446X2025-02-019316710.3390/drones9030167Integrated Control Method for STOVL UAV Based on RBF Neural Network and Nonlinear Dynamic AllocationShilong Ruan0Shuaibin An1Zhe Dong2Zeyu Jin3Kai Liu4School of Mechanics and Aerospace Engineering, Dalian University of Technology, Dalian 116081, ChinaSchool of Mechanics and Aerospace Engineering, Dalian University of Technology, Dalian 116081, ChinaSchool of Mechanics and Aerospace Engineering, Dalian University of Technology, Dalian 116081, ChinaSchool of Mechanics and Aerospace Engineering, Dalian University of Technology, Dalian 116081, ChinaSchool of Mechanics and Aerospace Engineering, Dalian University of Technology, Dalian 116081, ChinaA short takeoff and vertical landing unmanned aerial vehicle (STOVL UAV) is significantly influenced by factors such as the ship’s surface effect, deck motion, and jet effect during vertical landing on an aircraft carrier. The existing control logic cannot effectively solve the coupling problem of longitudinal attitude and trajectory, so it is hard to guarantee the stability and control accuracy of the UAV at low speed. To address the aforementioned interference and coupling problems, a comprehensive control law based on a radial basis function neural network (RBFNN) and nonlinear dynamic optimal allocation is designed in this paper. Firstly, the integrated landing control law of the STOVL UAV is designed. Considering the model uncertainty and complex landing environment, an RBFNN is used for online observation and compensation to improve the robustness of the system. Subsequently, a dynamic control allocation module based on nonlinear optimization is developed to simultaneously satisfy force and moment commands. The simulation results show that the integrated control method effectively decouples the pitch attitude and longitudinal trajectory at low speeds, resulting in effective convergence control of pitch angle, forward flight speed, and altitude. The integration of the RBFNN, as evaluated by the integral of absolute error (IAE), results in a 93% improvement in control accuracy compared to the integrated landing control law designed in this paper without the RBFNN integration.https://www.mdpi.com/2504-446X/9/3/167STOVL UAVnonlinear optimizationRBFNNcoupling control allocationthrust vector
spellingShingle Shilong Ruan
Shuaibin An
Zhe Dong
Zeyu Jin
Kai Liu
Integrated Control Method for STOVL UAV Based on RBF Neural Network and Nonlinear Dynamic Allocation
Drones
STOVL UAV
nonlinear optimization
RBFNN
coupling control allocation
thrust vector
title Integrated Control Method for STOVL UAV Based on RBF Neural Network and Nonlinear Dynamic Allocation
title_full Integrated Control Method for STOVL UAV Based on RBF Neural Network and Nonlinear Dynamic Allocation
title_fullStr Integrated Control Method for STOVL UAV Based on RBF Neural Network and Nonlinear Dynamic Allocation
title_full_unstemmed Integrated Control Method for STOVL UAV Based on RBF Neural Network and Nonlinear Dynamic Allocation
title_short Integrated Control Method for STOVL UAV Based on RBF Neural Network and Nonlinear Dynamic Allocation
title_sort integrated control method for stovl uav based on rbf neural network and nonlinear dynamic allocation
topic STOVL UAV
nonlinear optimization
RBFNN
coupling control allocation
thrust vector
url https://www.mdpi.com/2504-446X/9/3/167
work_keys_str_mv AT shilongruan integratedcontrolmethodforstovluavbasedonrbfneuralnetworkandnonlineardynamicallocation
AT shuaibinan integratedcontrolmethodforstovluavbasedonrbfneuralnetworkandnonlineardynamicallocation
AT zhedong integratedcontrolmethodforstovluavbasedonrbfneuralnetworkandnonlineardynamicallocation
AT zeyujin integratedcontrolmethodforstovluavbasedonrbfneuralnetworkandnonlineardynamicallocation
AT kailiu integratedcontrolmethodforstovluavbasedonrbfneuralnetworkandnonlineardynamicallocation