Dynamic modeling of soft continuum manipulators using lie group variational integration.

Area of Science:

• Robotics
• Mechanical Engineering
• Control Theory

Background:

• Current models for soft continuum manipulators often lack dynamic validation.
• Dynamic behavior is crucial for many real-world applications of these manipulators.
• Existing methods are typically limited to static or quasi-static analyses.

Purpose of the Study:

• To derive and experimentally validate dynamic models for soft continuum manipulators.
• To incorporate external forces and torques (magnetic fields, friction, gravity) into the modeling.
• To improve the estimation accuracy of manipulator tip position during dynamic motion.

Main Methods:

• Utilizing Lie group variational integration for modeling and estimation.
• Deriving a discrete variational principle for enhanced conservation properties.
• Conducting experiments with metal and polymer rods to validate the dynamic model.
• Employing distributed estimation filters to account for dissipative forces.

Main Results:

• The derived dynamic model shows good agreement with experimental data.
• Accurate estimation of tip position for a Polydimethylsiloxane (PDMS) rod manipulator.
• Achieved average absolute error of 0.13 mm and maximum error of 0.58 mm for a 60.55 mm manipulator.

Conclusions:

• The proposed Lie group variational integration approach effectively models and estimates the dynamic behavior of soft continuum manipulators.
• Experimental validation confirms the model's accuracy in the presence of external forces.
• This work provides a robust framework for dynamic control and simulation of soft robotic systems.