Repetitive control approach towards automatic tuning of Smith predictor controllers.

Area of Science:

• Control Systems Engineering
• Automation and Process Control

Background:

• Smith predictor controllers are effective for systems with time delays.
• Automatic tuning methods are crucial for optimizing controller performance.
• Repetitive Control (RC) offers potential for precise tracking of periodic signals.

Purpose of the Study:

• To propose a novel method for automatic tuning of Smith predictor controllers.
• To leverage Repetitive Control (RC) for enhancing controller performance.
• To enable parameter computation from experimental data.

Main Methods:

• Utilizing a Repetitive Control (RC) approach for automatic tuning.
• Employing a relay feedback experiment to obtain periodic reference signals.
• Applying a nonlinear least squares algorithm for parameter estimation.
• Using experimental data for initial parameter vector estimation.

Main Results:

• Demonstrated tracking error convergence through a modified RC scheme.
• Successfully computed Smith predictor controller parameters from experimental signals.
• Validated the effectiveness of the proposed tuning method via simulations and experiments.

Conclusions:

• The proposed method provides an effective approach for automatic tuning of Smith predictor controllers.
• Repetitive Control (RC) combined with relay feedback experiments offers a robust tuning solution.
• The method is validated through both simulation and experimental results.