An adaptive fuzzy model predictive control method for underwater robotic arms

In this study, we proposed a novel adaptive fuzzy model predictive controller for underwater robotic arms, which can operate efficiently in underwater environment where external environmental disturbances greatly impact the robotic arm's performance and the dynamics of the robotic arms are highly nonlinear. When the amount of data is not sufficient to build a model using data-driven techniques, T-S fuzzy model is used to obtain several linear subsystems for the dynamics modeling of the robotic arms. Designing predictive controllers for these linear subsystem, a significant computational burden is reduced. Furthermore, the proposed approach uses the adaptive interaction theory to counteract the effect of underwater disturbances, such as underwater currents, by adjusting the controllers' weight matrices with respect to environment and system state. experiments were performed with a 2-DOF robotic arm operating in a simulated underwater environment with disturbances. Through simulation experiments, the feasibility and effectiveness of the algorithm is verified.