A proposed soft pneumatic actuator control based on angle estimation from data-driven model.

Area of Science:

• Robotics
• Biomedical Engineering
• Control Systems

Background:

• Soft pneumatic actuators (SPAs) are crucial for developing assistive devices like rehabilitation gloves.
• Accurate control of SPA bending angles is essential for effective hand function assistance in patients with impairments.
• Existing control methods often struggle with the non-linear dynamics of SPAs and their components.

Purpose of the Study:

• To propose and validate a novel bending angle controller for SPAs.
• To enhance the performance of soft robotic rehabilitation gloves for patients with hand impairment, including stroke survivors.
• To improve the accuracy and responsiveness of SPA angle estimation.

Main Methods:

• Development of a data-driven model using linear regression to predict actuator angle based on pneumatic pressure.
• Optimization of actuator dimensions using a finite element model.
• Integration of an embedded flex sensor and a custom testing rig with a pneumatic control circuit and camera for data acquisition.
• Implementation and validation of a modified proportional-integral-derivative (PID) controller to address non-linearities.

Main Results:

• The data-driven model accurately estimates the bending angle of the soft pneumatic actuator.
• Experimental validation confirmed the model's accuracy and the performance of the modified PID controller.
• The combined system demonstrated more accurate angle estimation and quicker response times compared to baseline methods.
• The controller effectively mitigated the non-linear response of solenoid valves.

Conclusions:

• The proposed data-driven controller and modified PID approach offer a robust solution for precise control of soft pneumatic actuators.
• This technology holds significant potential for advancing soft robotic rehabilitation gloves and improving patient outcomes.
• The system's accuracy and responsiveness pave the way for more sophisticated and effective upper limb assistive devices.