Reverse pneumatic artificial muscles (rPAMs): Modeling, integration, and control

Area of Science:

• Robotics
• Materials Science
• Control Systems Engineering

Background:

• Soft pneumatic actuators offer advantages over rigid designs but lack efficient models and control.
• Existing systems often require bulky valves and extensive hardware, limiting their practical application.
• The reverse pneumatic artificial muscle (rPAM) presents a novel, inexpensive soft linear actuator solution.

Purpose of the Study:

• To develop and validate comprehensive analytical and numerical models for the rPAM.
• To investigate the application of rPAMs in controlling kinematic structures, specifically a revolute joint.
• To design and test advanced control schemes for precise rPAM operation using miniature valves.

Main Methods:

• Development of analytical and numerical static models for the rPAM, validated against experimental data.
• Derivation of an analytical model for a single-degree-of-freedom revolute joint driven by antagonistic rPAMs.
• Implementation and testing of a sliding-mode controller and an augmented sliding-mode controller with feed-forward for solenoid valve modulation.

Main Results:

• Analytical and numerical models accurately predict rPAM static behavior, aligning with experimental findings.
• The derived analytical model effectively predicts the static joint angle based on input pressures.
• Both proposed controllers demonstrated effective operation, with the feed-forward augmented controller showing superior performance in dynamic trajectory tracking.

Conclusions:

• The rPAM is a viable and reliable soft actuator with accurate modeling capabilities.
• The developed analytical models facilitate the design and control of soft robotic systems incorporating rPAMs.
• Advanced control strategies, particularly with feed-forward compensation, enable precise dynamic control of rPAM-driven mechanisms.