Data-driven cascade control system: Response estimation and controller design.

Area of Science:

• Control Systems Engineering
• Data-Driven Modeling
• Automation Technology

Background:

• Cascade control systems are crucial for precise industrial process management.
• Traditional design methods often require extensive closed-loop system data, which can be difficult to obtain.
• Data-driven approaches offer potential for more efficient controller design.

Purpose of the Study:

• To propose a novel data-driven design method for cascade control systems.
• To enable controller parameter optimization using only open-loop input-output data.
• To optimize the reference model's time constant for reduced control error.

Main Methods:

• Estimating the controlled plant's input-output response directly from open-loop data.
• Optimizing controller parameters by minimizing the difference between a reference model and the estimated closed-loop response.
• Estimating the response of a fictitious reference input to determine closed-loop behavior.
• Simultaneously optimizing the reference model time constant.

Main Results:

• The proposed method successfully determines cascade controller parameters using only open-loop data.
• Eliminates the need for complex and potentially unavailable closed-loop system data.
• Demonstrated effectiveness through numerical examples compared to conventional methods.

Conclusions:

• The developed data-driven method offers an efficient and practical approach for designing cascade control systems.
• This technique simplifies controller design by leveraging readily available open-loop data.
• The optimization of both controller parameters and reference model time constant enhances control performance.