A Punishment Mechanism-Combined Recurrent Neural Network to Solve Motion-Planning Problem of Redundant Robot...

Area of Science:

• Robotics
• Artificial Intelligence
• Control Systems

Background:

• Redundant robot manipulators (RRMs) require sophisticated motion planning for complex trajectory tracking.
• Existing methods for solving time-varying quadratic programming (TVQP) problems in motion planning can be computationally intensive and less accurate.

Purpose of the Study:

• To develop a novel and efficient method for motion planning in RRMs.
• To address the challenge of tracking complex time-varying trajectories using RRMs.
• To improve the speed and accuracy of solving the TVQP problem in robotic motion planning.

Main Methods:

• Formulating the RRM motion-planning problem as a constrained time-varying quadratic programming (TVQP) problem.
• Proposing a new punishment mechanism-combined recurrent neural network (PMRNN) inspired by varying-gain neural-dynamic design (VG-NDD).
• Theoretically analyzing the convergence performance of the PMRNN model for solving TVQP problems.

Main Results:

• The PMRNN model effectively solves the TVQP problem for RRM motion planning.
• Optimal joint angles and velocities for RRMs are obtained in the working space.
• The proposed PMRNN method demonstrates significantly faster calculation speeds and higher accuracy compared to traditional approaches.
• Successful application of the PMRNN model to an actual RRM for end-effector trajectory tracking.

Conclusions:

• The PMRNN model offers a superior solution for motion planning in RRMs, enabling complex trajectory tracking.
• This novel approach enhances computational efficiency and accuracy in robotic motion control.
• The PMRNN method has practical applicability, validated by successful implementation on a real RRM system.