Improved Adaptive Sliding Mode Control Using Quasi-Convex Functions and Neural Network-Assisted Time-Delay Estimation...

Area of Science:

• Robotics
• Control Systems Engineering
• Artificial Intelligence

Background:

• Robotic manipulators require precise control for complex tasks.
• Traditional control methods often suffer from chattering and instability.
• Existing time-delay estimation (TDE) techniques can have errors that impact performance.

Purpose of the Study:

• To develop an adaptive sliding mode control (SMC) strategy for robotic manipulators.
• To enhance time-delay estimation (TDE) using neural networks and mitigate TDE errors.
• To introduce a quasi-convex function-based continuous control gain to suppress chattering.

Main Methods:

• Implemented an adaptive SMC strategy with a neural network-enhanced TDE.
• Utilized radial basis function neural networks with a damped weight update law to compensate for TDE errors.
• Proposed a quasi-convex function-based continuous gain to replace traditional switching gains.

Main Results:

• The proposed continuous gain function effectively suppressed the chattering phenomenon.
• The adaptive control strategy guaranteed uniform ultimate boundedness, ensuring system stability.
• Simulation and experimental results validated the effectiveness of the proposed method.

Conclusions:

• The novel adaptive sliding mode control strategy enhances robotic manipulator performance by reducing chattering.
• The integration of neural networks for TDE compensation and quasi-convex functions for gain control offers a robust solution.
• The study demonstrates a significant advancement in stable and precise robotic control.