A fast closed-loop process dynamics characterization

Area of Science:

• Control Engineering
• Process Modeling
• System Identification

Background:

• Proportional-Integral-Derivative (PID) controllers are widely used in industrial automation.
• Accurate process modeling is crucial for effective PID controller tuning.
• Existing methods may struggle with complex processes including dead-time and instability.

Purpose of the Study:

• To develop a method for identifying unknown process dynamics for PID control.
• To enable the selection of PID controllers that optimize sensitivity and noise rejection.
• To validate the proposed identification and controller selection method through simulations and experiments.

Main Methods:

• Estimating the Laplace transform of the set-point step response to determine ultimate gain and frequency.
• Calculating the Nyquist curve tangent angle at the ultimate frequency.
• Determining process gain at zero frequency from control variable measurements.
• Constructing a control-relevant model (G(m)(s)) from estimated parameters.
• Utilizing pre-computed look-up tables for PID controller selection.

Main Results:

• The proposed method successfully estimates key process parameters (ultimate gain, frequency, tangent angle, low-frequency gain).
• A control-relevant model (G(m)(s)) is accurately derived from the estimated parameters.
• PID controllers guaranteeing desired sensitivity and noise immunity were obtained using look-up tables.
• Simulation and experimental results validated the effectiveness of the method on a thermal plant.

Conclusions:

• The developed method provides an effective approach for modeling unknown processes for PID control.
• This technique facilitates the selection of optimal PID controllers for enhanced performance and robustness.
• The findings are applicable to stable, integrating, and unstable processes, including those with dead-time.