Inverse kinematics for cooperative mobile manipulators based on self-adaptive differential evolution

Area of Science:

• Robotics
• Control Systems
• Optimization Algorithms

Background:

• Cooperative mobile manipulators are crucial for complex manipulation tasks.
• Solving inverse kinematics for these systems is challenging due to singularities and joint constraints.
• Existing methods often rely on Jacobian inversion, which can fail in singular configurations.

Purpose of the Study:

• To present a robust approach for solving the inverse kinematics of cooperative mobile manipulators.
• To address challenges related to singularities and joint limits in coordinate manipulation tasks.
• To validate the proposed method through simulations and experimental setups.

Main Methods:

• A self-adaptive differential evolution algorithm is employed to solve the inverse kinematics as a global constrained optimization problem.
• A kinematics model for a system of two omnidirectional platform manipulators with n degrees of freedom (DOF) is developed.
• An objective function is formulated using forward kinematics equations, incorporating penalty functions for joint limit constraints.

Main Results:

• The proposed approach effectively solves the inverse kinematics without requiring Jacobian matrix inversion, thus avoiding singularities.
• Simulation experiments demonstrate precise and accurate results for coordinate path tracking tasks.
• Experimental validation using KUKA Youbot systems confirms the practical applicability of the method.

Conclusions:

• The self-adaptive differential evolution algorithm provides an effective and singularity-free solution for the inverse kinematics of cooperative mobile manipulators.
• The developed objective function successfully handles joint limit constraints.
• The approach is demonstrated to be accurate and applicable for real-world robotic manipulation tasks.