Velocity Feedback-Free Synchronous Control for Pneumatic Artificial Muscle-Actuated Parallel Robots With Fractional-Order Hysteresis

Abstract

Pneumatic artificial muscles (PAMs) can reduce the weight of robots and endow them with safer flexibility, by leveraging advantages such as large force-weight ratios and controllable stiffness. However, their input-output relationship also exhibits hysteresis, which is detrimental to the control performance of robots. The existing hysteresis models of PAMs usually have complex structures, and the accuracy depends on the number of model parameters. Therefore, by establishing a modified fractional-order Bouc-Wen (MFOBW) model, this article develops a synchronous controller with hysteresis compensation for PAM-actuated parallel robots (PAMPRs), which quickly reconstructs the velocities of master arms and achieves their synchronization and trajectory tracking. Notably, fractional calculus is employed to capture the rate-dependent hysteresis of PAMs, effectively simplifying the model structure, and the added auxiliary term further improves model accuracy; meanwhile, the designed velocity estimation system containing sign functions not only reconstructs unmeasurable velocities, but also accelerates the convergence rate without producing chattering. To our knowledge, this article provides the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">first</i> MFOBW model for PAMs’ hysteresis, and the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">first</i> finite-time synchronous controller for PAMPRs <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">without</i> velocity information. Additionally, the closed-loop stability is proven through theoretical analysis, and the MFOBW model’s effectiveness and the proposed method’s feasibility are verified through hardware experiments.