Model-Based Feedforward Control of a Robotic Manipulator With Pneumatic Artificial Muscles

Pneumatic artificial muscles (PAMs) are lightweight, flexible actuators capable of higher specific work than comparably-sized hydraulic actuators at the same pressure and electric motors. PAMs are composed of an elastomeric bladder surrounded by a helically braided sleeve. Lightweight, compliant actuators are particularly desirable in portable, heavy-lift robotic systems intended for interaction with humans, such as those envisioned for patient assistance in hospitals and battlefield casualty extraction. However, smooth and precise control remains difficult because of nonlinearities in the dynamic response. The objective of this paper is to develop a control algorithm that satisfies accuracy and smooth motion requirements for a two degree-of-freedom manipulator actuated by pneumatic artificial muscles and intended for interaction with humans, such as lifting a human. This control strategy must be capable of responding to large, abrupt variations in payload weight over a high range of motion. In previous work, the authors detailed the design and construction of a proof-of-concept PAM-based manipulator. The present work investigates the feasibility of combining output feedback using proportional-integral-derivative control or fuzzy logic control with model-based feedforward compensation to achieve improved closed-loop performance. The model upon which the controller is based incorporates the internal airflow dynamics, the geometric parameters of the pneumatic actuators, and the arm dynamics. Simulations were performed in order to validate the control algorithm, guide controller design, and predict optimal gains. Using real-time interface software and hardware, the controller was implemented and experimentally tested on the manipulator. Performance was evaluated for several trajectories, and different payload weights. The effect of varying the feedforward gain was also analyzed. Model refinement further improved performance.