Robust Controller Design for an Autonomous Underwater Vehicle

• Autor(a) principal: Mendes, Carlos Hugo Ribeiro
• Data de Publicação: 2017
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10400.6/7956
• topic: Controlo Robusto; Proportional Integral Derivativo (Pid); Regulador Linear Quadrático (Lqr); Simulação; Veículo Autónomo Subaquático; Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Aeronáutica

Worldwide there has been a surge of interest in Autonomous Underwater Vehicles (AUV). The ability to operate without human intervention is what makes this technology so appealing. On the other hand, the absence of the human narrows the AUV operation to its control system, computing, and sensing capabilities. Therefore, devising a robust control is mandatory to allow the feasibility of the AUV. Motivated by this fact, this thesis aims to present, discuss and evaluate two linear control solutions being proposed for an AUV developed by a consortium led by CEiiA. To allow the controller design, the dynamic model of this vehicle and respective considerations are firstly addressed. Since the purpose is to enable the vehicle’s operation, devising suitable guidance laws becomes essential. A simple waypoint following and station keeping algorithm, and a path following algorithms are presented. To devise the controllers, a linear version of the dynamic model is derived considering a single operational point. Then, through the decoupling of the linear system into three lightly interactive subsystems, four Proportional Integral Derivative controllers (PIDs) are devised for each Degree Of Freedom (DOF) of the vehicle. A Linear Quadratic Regulator (LQR) design, based on the decoupling of the linear model into longitudinal and lateral subsystems is also devised. To allocate the controller output throughout the actuators, a control allocation law is devised, which improves maneuverability of the vehicle. The results present a solid performance for both control methods, however, in this work, LQR proved to be slightly faster than PID.