Ho P.H. Anh
NARX fuzzy scheme, internal model controller, pneumatic artificial muscle manipulator, identified process, fuzzy-based internal model controller, particle swarm optimization
The paper investigates the fuzzy non-linear auto-regressive with eXogenous input (NARX) model used to model the dynamic pneumatic artificial muscle (PAM) manipulator. The NARX fuzzy model is then introduced in the proposed fuzzy-based internal model controller (IMC) control scheme, to effectively track the two-axis PAM manipulator rotation. Parameters of the fuzzy NARX models are optimized using particle swarm optimization. Good performance of the novel control method combined the robustness of IMC control and the prediction capability of the proposed fuzzy scheme. Efficiency of the novel control method is proven by experiments for two different payloads as well as two typical control schemes.
Important Links:
Go Back
