Data-Based Robust Tracking Control for Learning Systems Under Disturbance Observers.

Area of Science:

• Control Systems Engineering
• Robotics
• Signal Processing

Background:

• High-precision tracking is crucial for iterative learning control (ILC) systems.
• Iteration-varying disturbances pose significant challenges to achieving robust tracking.
• Accurate system model information is often unavailable in practical ILC applications.

Purpose of the Study:

• To develop a robust tracking control strategy for ILC systems with iteration-varying disturbances and unknown models.
• To enhance the tracking precision of ILC systems despite unpredictable, changing disturbances.
• To address the limitations of traditional ILC methods in uncertain environments.

Main Methods:

• Constructing a nominal ILC system model using input-output data from test iterations.
• Establishing a disturbance observer (DOB) to estimate iteration-varying disturbances and model uncertainties.
• Developing a DOB-based ILC updating law incorporating disturbance and uncertainty estimations.

Main Results:

• The proposed DOB-based ILC updating law significantly improves tracking performance under iteration-varying disturbances.
• Tracking error is shown to be continuously dependent on the bound of the disturbance's second-order variation rate.
• Perfect tracking is achievable when the iteration-varying disturbance has a convergent variation rate.

Conclusions:

• The developed DOB-based ILC method effectively handles iteration-varying disturbances without requiring an accurate system model.
• The approach relies solely on input and output data from test iterations for designing the DOB and updating law.
• This method offers a practical solution for enhancing ILC system robustness and tracking accuracy in complex scenarios.