Neural-Network-Based Sliding-Mode Control of an Uncertain Robot Using Dynamic Model Approximated Switching Gain

Area of Science:

• Robotics
• Control Systems
• Artificial Intelligence

Background:

• Conventional sliding-mode control (SMC) methods often suffer from chattering and high-speed switching of control input.
• Uncertainties and disturbances in robotic systems pose significant challenges for precise trajectory tracking.

Purpose of the Study:

• To develop a novel neural-network-based sliding-mode control (SMC) scheme for uncertain robotic systems.
• To address the limitations of conventional SMC, specifically chattering and control input switching.
• To ensure robust trajectory tracking with guaranteed precision.

Main Methods:

• A neural-network strategy is employed to approximate the unknown dynamics and disturbances, designing the switching gain as a dynamic model approximation.
• The control scheme requires minimal prior modeling information, needing only one parameter estimation per joint.
• Lyapunov stability theory is used to rigorously prove the convergence of trajectory tracking errors.

Main Results:

• The proposed control scheme effectively adapts to unknown dynamics and external disturbances.
• Simulation studies demonstrate the successful mitigation of chattering and high-speed switching issues inherent in traditional SMC.
• The novel approach guarantees satisfactory control precision for uncertain robotic systems.

Conclusions:

• The neural-network-based SMC offers a robust and effective solution for controlling uncertain robots.
• This method overcomes key limitations of conventional SMC, providing superior performance and precision.
• The adaptive nature of the control scheme ensures reliable trajectory tracking even with unmodeled dynamics.