Re-design of Smith predictor systems for performance enhancement

Area of Science:

• Control Systems Engineering
• Automation and Process Control

Background:

• Smith Predictor systems are widely used for controlling systems with time delays.
• Traditional Smith Predictor designs often rely on perfectly matched models, which can limit performance in real-world applications.

Purpose of the Study:

• To present a new design for Smith Predictor systems that enhances performance.
• To investigate the use of a deliberately mismatched model to achieve superior control outcomes.
• To develop a robust methodology for designing controllers for these enhanced systems.

Main Methods:

• The design methodology is formulated as a frequency-domain optimization problem.
• Linear least squares methods are utilized to find an approximate solution for controller design.
• A modified Smith Predictor structure is proposed, simplifying the controller design to second-order dynamics.

Main Results:

• The proposed mismatched model approach enhances performance compared to perfectly matched systems.
• The use of linear least squares provides an effective approximate solution for controller design.
• The modified structure simplifies the primary controller design, maintaining high performance.

Conclusions:

• The novel Smith Predictor design with a mismatched model offers improved control performance.
• The frequency-domain optimization and linear least squares method provide a practical design approach.
• The modified system structure facilitates simpler yet effective controller design for time-delay systems.