Doutorado em Engenharia Elétrica
URI Permanente para esta coleção
Nível: Doutorado
Ano de início:
Conceito atual na CAPES:
Ato normativo:
Periodicidade de seleção:
Área(s) de concentração:
Url do curso:
Navegar
Navegando Doutorado em Engenharia Elétrica por Autor "Amaral, Paulo Farias Santos"
Agora exibindo 1 - 3 de 3
Resultados por página
Opções de Ordenação
- ItemModelagem e compensação da dinâmica de robôs móveis e sua aplicação em controle de formação(Universidade Federal do Espírito Santo, 2009-03-06) Martins, Felipe Nascimento; Bastos Filho, Teodiano Freire; Carelli, Ricardo; Sarcinelli Filho, Mário; Fardin, Jussara Farias; Alsina, Pablo Javier; Amaral, Paulo Farias Santos; Dynnikov, Vladimir IvanovitchA new dynamic modeling approach for unicycle-like mobile robots is proposed, which is applied in the design of controllers for this type of robot. The dynamic model thus generated accepts linear and angular velocities as inputs, which is usual in commercial robots. Some of its properties are studied and proved, and are then used in the design of adaptive controllers that compensate for the robot dynamics while tracking a desired trajectory, following a leader or being part of a group in formation control problems. The Lyapunov theory is used on the stability analysis of the equilibrium in every case. A robustness analysis considering possible parameter variation and non-modeled disturbance is also performed. The influence of the dynamic compensation is studied, and its importance is illustrated by a performance index measured for both simulation and experimentation. Three formation control strategies with dynamic compensation are presented: one is a decentralized leader-follower control, and the other two are centralized virtual structure control. A Multi-Layer Scheme for formation control is here presented using one of the centralized formation control strategies. Such scheme is flexible in the sense that each part of the formation control problem is solved by an independent module. The proposed formation controller is capable of making the robots achieve a fixed desired formation, and to follow a desired formation having time-varying position and shape. The influence of the dynamic compensation on this formation control scheme is analyzed and illustrated through both simulation and experimental results.
- ItemProjeto de controladores não lineares para voo autônomo de veículos aéreos de pás rotativas(Universidade Federal do Espírito Santo, 2013-04-03) Brandão, Alexandre Santos; Sarcinelli Filho, Mario; Carelli, Ricardo; Munaro, Celso José; Borges, Geovany Araújo; Raffo, Guilherme Vianna; Amaral, Paulo Farias SantosThis Ph.D. Thesis proposes nonlinear controllers to guide rotary-wing unmanned aerial vehicles (UAV) to accomplish tasks of positioning, trajectory tracking and path following in the 3D space. Two cases are addressed: the UAV is navigating alone or in cooperation with an unmanned ground vehicle (UGV). Initially the dynamic model of the rotorcraft is presented, obtained by using the Euler-Lagrange formulation and the Newton-Euler formulation as well. In addition, it is presented the representation of a UAV (helicopter or quadrotor) focusing on the underactuated characteristic of the model developed for the aircraft (the variables to be controlled are more than the control signals available). Considering the flight missions themselves, as a first step some restrictions of movement are applied to the aircraft, such that its movement become restricted to the Z axis and, in the sequel, to the XZ and Y Z planes, referenced to the inertial frame. For such cases, PVTOL (Planar Vertical Takeoff and Landing) controllers capable of guiding the aircraft in taking-off, hovering and landing are proposed. The stability of the control system implemented using such controllers, in Lyapunov’s sense, is demonstrated, and the controllers are validated through simulated and experimental results. In the sequel, the flight restrictions are relaxed, and the UAV becomes able to fly in the 3D space. At this point, a important contribution of this work is the proposal of a nonlinear controller based on partial feedback linearization, considering the high coupling between the active/actuated and passive/nonactuated parts of the underactuated system. Simulated and experimental results validate the proposed controller to be used in the classical classes of movement control in Robotics. Following, it is presented a proposal of a switching scheme associated to the PVTOL controllers previously proposed, so that it becomes possible to use simpler controllers to guide the aircraft in a 3D flight mission. Finally, this Thesis also presents a control scheme to guide the navigation of a UAV in coordination with a UGV (or a group of UGVs). The leader-follower control strategy is adopted, to allow the UAV to track the UGV, which is labeled the leader of the formation. Notice that the controllers adopted for guiding the UAV and the UGV work in a completely independent way, with the leader-follower approach being an upper layer responsible for coordinating the poses of the two vehicles. The stability of the control system using such controller is proven, using Lyapunov’s theory, and simulated and experimental results also shown validate the proposed control scheme.
- ItemUm sistema autônomo para navegação de cadeiras de rodas robóticas orientadas a pessoas com deficiência motora severa(Universidade Federal do Espírito Santo, 2009-08-05) Celeste, Wanderley Cardoso; Sarcinelli Filho, Mário; Carelli, Ricardo; Bastos Filho, Teodiano Freire; Ferreira, Edson de Paula; Salles, Jose Leandro Félix; Alsina, Pablo Javier; Amaral, Paulo Farias SantosThis dissertation proposes a solution to the problem met by people with severe motordiseases, which have full incapacity of moving around by themselves, but keep intact their cognitive abilities. Such a solution is a navigating system for a Robotic Wheelchair, which consists of a control system, a reference generating system and a supervisory system. The control system comprises a set of kinematic controllers to execute speci c tasks, besides a dynamic compensation controller capable to adapt itself in the presence of structured uncertainties and being robust to unstructured uncertainties. The robust adaptive dynamic compensator is based on a dynamic model of the robotic wheelchair carrying a user seated. Speci c tasks are accomplished by the reference generator based on a user command and vehicle information. User commands establish poses of interest to be reached by the robotic wheelchair after following a safe route. To split a general task in speci c ones is the strategy here adopted to reduce the complexity of the navigating system, which should also take care of the safety and comfort of the user. The supervisory system is responsible for an adequate coordination of the kinematic controllers, in addition to perform a sequence of speci c tasks. Simulation and experimentation results show the good performance of the system, even when submitted to changes caused by a user on board the vehicle and by external e ects.