Innovative inverse kinematics algorithm for 6-DOF robotic manipulators with offset wrists

Area of Science:

• Robotics
• Numerical Analysis
• Control Systems

Background:

• Inverse kinematics (IK) for robotic arms with offset wrists presents challenges due to biased parameters.
• Existing IK solutions may struggle with accuracy and convergence in complex robotic configurations.

Purpose of the Study:

• To develop a novel numerical algorithm for solving inverse kinematics in robotic arms with offset wrists.
• To enhance the accuracy and robustness of IK solutions by addressing biased parameters and avoiding local optima.

Main Methods:

• Simplification of the IK problem by categorizing robotic arms and transforming the terminal link.
• Correction of the Hessian matrix using Jacobian matrix and regularization terms.
• Dynamic adjustment of the scaled memoryless augmented Broyden-Fletcher-Goldfarb-Shanno (SMABFGS) algorithm's step size with momentum-based gradient descent (GD) for iterative refinement.

Main Results:

• The algorithm successfully approximates desired joint angles for robotic arms with biased wrists.
• Simulation experiments validate the accuracy of the IK solutions for discrete end-effector poses.
• Real-time trajectory tracking experiments on a physical robotic arm demonstrate practical effectiveness.

Conclusions:

• The proposed numerical algorithm offers an effective solution for inverse kinematics in robotic arms with offset wrists.
• The integration of adaptive step size adjustment and matrix correction improves convergence and accuracy.
• The algorithm demonstrates robust performance for both discrete pose solving and continuous trajectory tracking in practical applications.