Intelligent complementary sliding-mode control for LUSMS-based X-Y-theta motion control stage

Area of Science:

• Robotics and Control Systems
• Artificial Intelligence
• Neural Networks

Background:

• Linear ultrasonic motors (LUSMs) are crucial for precise motion control stages.
• Contour tracking requires sophisticated control systems to manage complex movements.
• Existing sliding-mode control (SMC) methods have limitations in tracking accuracy and error bounds.

Purpose of the Study:

• To propose an intelligent complementary sliding-mode control (ICSMC) system for LUSM-based X-Y-theta motion control stages.
• To improve contour tracking performance by reducing tracking errors.
• To develop an on-line uncertainty estimation method using a novel neural network.

Main Methods:

• Development of an ICSMC system incorporating a recurrent wavelet-based Elman neural network (RWENN) estimator.
• Utilizing a complementary generalized error transformation to reduce tracking error bounds.
• Employing wavelet functions as activation functions in the RWENN for enhanced convergence.
• Deriving RWENN estimation laws via Lyapunov stability theorem for on-line training.
• Introducing a robust compensator to handle system uncertainties.

Main Results:

• The proposed ICSMC system demonstrated significantly improved tracking performance compared to conventional SMC and CSMC.
• The RWENN estimator effectively estimated lumped uncertainties on-line.
• The use of wavelet functions in RWENN improved convergent precision and time.
• The complementary generalized error transformation halved the guaranteed ultimate bound of the tracking error.

Conclusions:

• The ICSMC system offers superior contour tracking capabilities for LUSM-based motion stages.
• The RWENN estimator provides an effective solution for on-line uncertainty estimation in control systems.
• The integration of wavelet neural networks enhances the precision and speed of control systems.