Auto-tuning of filtered proportional-integral-derivative controller for industrial processes under routine operating...

Area of Science:

• Control Engineering
• Automation Systems
• Process Control

Background:

• Traditional auto-tuning of PID controllers often requires dual controllers, specific experiments, or prior process knowledge, leading to high implementation costs.
• Existing methods may necessitate external excitation or detailed system information, limiting their applicability in routine operating conditions.

Purpose of the Study:

• To develop a new auto-tuning scheme for fractional-order PID (FPID) controllers for minimum variance tasks.
• To enable auto-tuning under routine operating conditions without external excitation or prior process knowledge.
• To reduce the cost associated with auto-tuning implementation.

Main Methods:

• Online estimation of a closed-loop non-parametric model using impulse response coefficients from one-shot output data.
• Determination of the stochastic disturbance model from impulse response coefficients.
• Estimation of the plant model and updating FPID controller parameters by solving optimization problems (H2 norm).

Main Results:

• A unique stochastic disturbance model can be determined under specific preconditions.
• The proposed scheme successfully estimates the plant model and updates FPID controller parameters online.
• The method avoids reliance on dual controllers, identification experiments, or prior process knowledge.

Conclusions:

• The developed auto-tuning scheme offers a cost-effective alternative for FPID controllers.
• It effectively tunes controllers for minimum variance tasks under normal operating conditions.
• Demonstrated effectiveness through numerical simulations and industrial case studies.