Métodos de controle de posição para manipuladores robóticos

Abstract

Some tasks of robotic manipulators are performed using trajectory tracking control techniques. These methods ensure that the steady-state error between desired and executed trajectories is minimized. This study proposes a hybrid control scheme that enhances a conventional control approach with optimization using computational intelligence techniques. The Linear Quadratic Regulator (LQR) is designed through the dynamic model in state spaces obtained by linearizing the plant to be controlled. The optimization of this controller is achieved by tuning the weighting variables of the cost function. Computational tuning is employed using fuzzy logic (FL) techniques. Results are derived from transient response analysis and measurement of mean squared error. The simulation conducted with MATLAB software tools allows for visualization of the positioning of the two-degree-of-freedom planar manipulator. The results demonstrate that the LQR controller optimized with FL is more effective for complex tasks than the conventional LQR controller. The LQR-FL controller not only improves precision but also enhances performance in controlling continuous trajectories, providing more accurate adjustment in tracking such trajectories.