RBF neural network dynamic sliding mode control based on lambert W function for piezoelectric stick-slip actuator

Area of Science:

• Control Engineering
• Robotics
• Nanotechnology

Background:

• Piezoelectric actuators are crucial for high-precision applications.
• Traditional control methods struggle with jitter and disturbances.
• Time delays can significantly degrade positioning accuracy.

Purpose of the Study:

• To develop a novel control strategy for enhancing the precision of piezoelectric stick-slip actuator systems.
• To address jitter, model uncertainties, and time delays in control systems.
• To improve the robustness and accuracy of high-precision positioning.

Main Methods:

• Dynamic sliding mode control combined with a radial basis function neural network (RBFNN).
• Lyapunov stability analysis to verify controller performance.
• Lambert W function optimization to mitigate time-lag effects.

Main Results:

• Achieved positioning control accuracy of <40 nm in scanning mode.
• Demonstrated robustness against external loads.
• Successfully controlled long-distance positioning (40 μm) with maintained accuracy.
• Reduced the adverse impact of time delays on system performance.

Conclusions:

• The proposed RBFNN-based dynamic sliding mode control strategy significantly enhances positioning precision for piezoelectric actuators.
• The method effectively handles uncertainties, disturbances, and time delays, improving system robustness.
• This approach offers a promising solution for demanding high-precision positioning applications.