Composite intelligent learning-based tracking control for discrete-time repetitive process

Area of Science:

• Control Systems Engineering
• Robotics and Automation
• Signal Processing

Background:

• Repetitive discrete-time systems require precise tracking control for finite-duration tasks.
• Existing iterative learning control (ILC) methods may face challenges in achieving both fast convergence and high precision.

Purpose of the Study:

• To develop a novel composite iterative learning control (ILC) algorithm for discrete-time systems with repetitive tasks.
• To enhance tracking performance by integrating two distinct control phases.

Main Methods:

• A two-phase intelligent learning process: gain-adaptive iterative learning control (GAILC) and sliding mode iterative learning control (SMILC).
• Phase switching is dynamically determined by the tracking error.
• GAILC utilizes a prediction-based adaptive gain sequence for rapid error convergence.
• SMILC employs a novel sliding surface and a fractional power term for high tracking precision.

Main Results:

• The proposed composite ILC algorithm demonstrates improved tracking performance.
• Comparative simulations, including a DC motor example, validate the effectiveness of the GAILC-SMILC strategy.
• The two-phase approach achieves both fast convergence and high precision in tracking.

Conclusions:

• The developed composite ILC strategy offers a significant advantage for repetitive discrete-time systems.
• The intelligent two-phase learning process effectively addresses tracking challenges.
• The method shows promise for applications requiring precise repetitive motion control.