| Texto completo | |
| Autor(es): |
Xavier, Daniel M.
;
Silva, Natassya B. F.
;
Branco, Kalinka R. L. J. C.
Número total de Autores: 3
|
| Tipo de documento: | Artigo Científico |
| Fonte: | JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS; v. 106, n. 3, p. 17-pg., 2022-11-01. |
| Resumo | |
Unmanned Aerial Vehicles (UAVs) are aircraft that can be manually operated or autonomously guided through an autopilot. In the last case, the system is responsible for stabilising the aircraft and executing guidance tasks such as path-following. In literature, several path-following algorithms were proposed for straight lines and loiter paths. Previous comparisons generally consider straight line algorithms in a 2D space using the kinematic model of an aircraft. In order to complement existing research, this paper compares four 3D path-following algorithms for loiter paths under different wind intensities. Tests are made through Software-in-the-Loop (SiL) simulations using the dynamic model of a fixed-wing UAV. Furthermore, a genetic algorithm is employed to tune the parameters and the analysis is carried out under varying wind conditions. The algorithms compared are four well-known geometric methods: Carrot-Chasing (CC), Non-Linear Guidance Law (NLGL), Pure Pursuit and Line-of-Sight (PLOS) and Vector Field (VF). Results show that Vector Field has the smallest errors, while PLOS is the most resistant to wind disturbance. (AU) | |
| Processo FAPESP: | 17/21303-2 - Avaliação de algoritmos de seguimento de trajetórias retas e circulares para veículos aéreos não tripulados |
| Beneficiário: | Daniel Martin Xavier |
| Modalidade de apoio: | Bolsas no Brasil - Iniciação Científica |