Compensation function observer based model-free adaptive boundary layer sliding mode control for hydraulic...

Area of Science:

• Robotics and Control Systems
• Hydraulic Engineering
• Adaptive Control Theory

Background:

• Hydraulic manipulators are crucial in various industries but suffer from complex dynamics and uncertainties like dead-zones.
• Existing control methods often struggle with precise tracking and robustness against these nonlinearities.
• Model-free control offers a promising alternative by avoiding complex system modeling.

Purpose of the Study:

• To propose a novel model-free adaptive control strategy for hydraulic manipulators.
• To effectively compensate for dead-zone uncertainties in directional valves.
• To achieve precise force tracking and robust position control.

Main Methods:

• A dual-loop control structure combining force and position sub-control.
• Utilizing an ultra-local model (ULM) for a model-free framework.
• Implementing a compensation function observer (CFO) for uncertainty estimation.
• Employing a nonsingular fast terminal sliding mode (NFTSM) control law.

Main Results:

• The proposed CFO-DIBLSMC achieved zero error estimation for lumped uncertainties.
• Demonstrated robustness against high-frequency disturbances and dead-zone/friction effects.
• Comparative simulations confirmed superior performance over conventional methods.

Conclusions:

• The CFO-DIBLSMC is effective for hydraulic manipulators with dead-zone compensation.
• The model-free approach simplifies control design while maintaining high performance.
• This method offers enhanced precision and robustness for complex robotic systems.