Predefined-Time Super-Twisting Sliding Mode Control for Construction Robot with Arbitrary Initial Values

Area of Science:

• Robotics
• Control Systems Engineering
• Artificial Intelligence

Background:

• Construction robots face challenges in precisely tracking reference trajectories due to complex operational environments.
• Existing control methods may not fully address the dynamic and uncertain nature of construction robot systems.

Purpose of the Study:

• To develop an advanced control scheme for precise trajectory tracking in construction robots.
• To design a novel super-twisting sliding mode controller with predefined-time convergence.
• To address the influence of fluid materials and composite disturbances on robot control.

Main Methods:

• Exploration of fluid material influence and trajectory tracking control features for construction robots.
• Design of an extended reference trajectory and a nonsingular sliding surface with predefined-time convergence.
• Development of a Radial Basis Function (RBF) neural network for approximating system disturbances.
• Implementation of a super-twisting sliding mode controller based on predefined-time convergent control theory.

Main Results:

• The proposed controller guarantees accurate movement of construction robot angles along the reference trajectory.
• Achieved angular tracking error precision of 3 × 10^-6 rad.
• Theoretical validation of the closed-loop system's predefined-time convergence performance.

Conclusions:

• The developed control scheme and controller are feasible and effective for precise trajectory tracking in construction robots.
• The method significantly enhances the accuracy and reliability of construction robot operations.
• This research contributes to advancing autonomous capabilities in the construction industry.