A Novel Supertwisting Zeroing Neural Network With Application to Mobile Robot Manipulators.

Area of Science:

• Robotics and Control Systems
• Artificial Intelligence
• Nonlinear Dynamics

Background:

• Existing Zeroing Neural Network (ZNN) models face challenges with slow convergence and limited robustness analysis for robot manipulator tracking control.
• Current ZNN models often exhibit infinite convergence times and lack comprehensive studies on their stability and asymptotic convergence robustness.

Purpose of the Study:

• To propose a new ZNN model that enhances both convergence speed and robustness for mobile robot manipulator tracking control.
• To develop a Super-Twisting Zeroing Neural Network (STZNN) model that addresses the limitations of existing ZNN approaches.

Main Methods:

• Integration of the Super-Twisting (ST) algorithm into the ZNN framework to create the STZNN model.
• Development of a unified design process connecting sliding mode control principles with ZNN.
• Mathematical formulation and rigorous proofs for global stability, finite-time convergence, and robustness.

Main Results:

• The proposed STZNN model demonstrates inherent finite-time convergence capabilities.
• The STZNN model exhibits enhanced robustness in tracking control applications.
• The model's effectiveness and superiority are validated through path-tracking simulations and comparative analyses.

Conclusions:

• The STZNN model offers a significant advancement in tracking control for mobile robot manipulators.
• The integration of ST algorithms provides a robust and efficient solution for ZNN-based control.
• The study confirms the STZNN model's potential for fast and reliable robot control applications.