Frequency response adaptation of PI controllers based on recursive least-squares process identification.

Area of Science:

• Control Engineering
• Process Automation
• Adaptive Systems

Background:

• Adaptive control and autotuning are crucial for optimizing industrial processes.
• Proportional-Integral (PI) controllers are widely used but require precise tuning.
• Existing tuning methods can be sensitive to process variations and disturbances.

Purpose of the Study:

• To develop and validate a robust frequency domain approach for adaptive PI controller autotuning.
• To improve the reliability and convergence speed of controller tuning parameters.
• To assess the method's performance under significant process variations.

Main Methods:

• Utilizing an overparametrized recursive least-squares (RLS) model to estimate process frequency response.
• Employing frequency response estimates at sample intervals for PI controller tuning.
• Validating the approach through simulations and experimental implementation on a pilot distillation column.

Main Results:

• Frequency response estimates show faster convergence and greater constancy under disturbances compared to RLS coefficients.
• The proposed method demonstrates reliability for tuning PI controllers.
• The autotuning technique effectively handles concurrent gain variations up to 50% and time constant/delay variations up to 100%.

Conclusions:

• The frequency domain approach offers a more reliable method for adaptive PI controller autotuning.
• This technique provides robust performance even with significant process parameter changes.
• Experimental verification confirms the practical applicability of the method in real-world systems like distillation columns.