Cooperative path following control of ocean vehicles using lyapunov - based model predictive control / moving horizon estimation and consensus theory.

• Ano de defesa: 2023
• Autor(a) principal: Zakartchouk Junior, Alexis
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Controle cooperativo; Controle não-linear; Controle preditivo baseado em modelo; Cooperative control; Estimação baseada em horizonte móvel; Model predictive control; Moving horizon estimation; Nonlinear control; Path following; Seguimento de rota
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/3/3135/tde-02062023-075802/

Boosted by recent developments in control techniques, computing power, communication capabilities and miniaturization technologies, the cooperative deployment of multi-agent systems has been drawing the attention of the scientific community and industry, for applications in distinct areas such as military battle systems, mobile sensor networks, survey & inspection, transport systems and others. Inspired by natural biological systems, the use of a cluster of cooperative distributed agents has proved to be advantageous in terms of cost, efficiency, flexibility and reliability, when compared to conventional monolithic structures relying on a single agent. These advantages become even more preeminent for marine applications, since autonomous marine vehicles are much cheaper than conventional vessels in terms of capital and operational costs, can safely explore previously impenetrable environments of the sea, are less subjected to be hampered by rough weather, and are more discrete in terms of magnetic and acoustic signatures. To solve the associated control problem, this thesis envisages a control strategy relying on two layers of implementation. The lower-level layer, based on the receding horizon concept (model predictive control and moving horizon estimation), controls the vehicles individual motion. The higher-level layer, based on consensus theory, controls the motion of the entire formation. The receding horizon concept is particularly interesting for motion control of marine vehicles due to its capability to deal with nonlinear dynamics, parametric uncertainties and external disturbances, besides its inherent ability to systematically handle practical constraints on control signals and states. Consensus theory is interesting for formation control due to the simplicity of the associated control law. The performance of the designed control system is assessed through numerical simulations.