Predefined-Time Sliding Mode Control of Robotic Manipulators via Artificial Delay Feedback and Reinforcement...

Area of Science:

• Robotics
• Control Systems Engineering
• Artificial Intelligence

Background:

• Modern industrial automation demands high-precision trajectory tracking and rigid temporal control.
• Existing predefined-time control methods struggle with disturbances, causing actuator chattering and performance degradation.

Purpose of the Study:

• To develop a novel composite control strategy for robotic manipulators.
• To enhance temporal determinism and reduce chattering in control systems.
• To improve trajectory tracking accuracy under disturbances.

Main Methods:

• Integration of Actor-Critic reinforcement learning (RL) with predefined-time sliding mode control (PTC).
• Design of a novel predefined-time sliding surface using artificial delay feedback.
• Online compensation of unknown dynamics and external disturbances via RL agent.
• Lyapunov stability theory for system boundedness proof.

Main Results:

• Achieved high-precision, singularity-free trajectory tracking within a user-defined time (Tc).
• Effectively suppressed high-frequency chattering compared to traditional non-singular terminal sliding mode control (NTSMC).
• Demonstrated robustness under parameter mismatches and time-varying disturbances.

Conclusions:

• The proposed PTC-RLC strategy offers superior temporal determinism and tracking accuracy.
• The method significantly reduces actuator chattering, leading to smoother control torque.
• The strategy shows strong potential for practical implementation in industrial automation.