Data-Driven Indirect Iterative Learning Control.

Area of Science:

• Control Engineering
• Automation Systems
• Nonlinear System Dynamics

Background:

• Repetitive nonlinear systems often require advanced control strategies for precise execution.
• Existing methods may struggle with systems lacking explicit model information or exhibiting nonaffine dynamics.
• Iterative learning control (ILC) offers a framework for improving performance over repeated trials.

Purpose of the Study:

• To develop a data-driven indirect iterative learning control (DD-iILC) scheme for nonlinear, nonaffine, repetitive systems.
• To design an adaptive iterative tuning algorithm for set-point control using theoretical nonlinear learning functions.
• To validate the proposed control strategy's convergence and effectiveness through simulations.

Main Methods:

• Integration of a proportional-integral-derivative (PID) controller in the inner feedback loop.
• Application of iterative dynamic linearization (IDL) to derive a linear parametric iterative tuning algorithm from ideal nonlinear functions.
• Development of an adaptive iterative updating strategy for parameter tuning by optimizing an objective function.

Main Results:

• Successful implementation of a DD-iILC scheme for nonlinear systems without prior model knowledge.
• Demonstrated convergence of the control strategy using contraction mapping and mathematical induction.
• Validation through simulations on both a general numerical example and a specific permanent magnet linear motor.

Conclusions:

• The proposed DD-iILC, combined with PID control and adaptive tuning, effectively manages repetitive nonlinear systems.
• The IDL technique enables model-free control design, making the approach broadly applicable.
• The control scheme shows robust performance and guaranteed convergence for the targeted applications.