Programmable logic controller implementation of an auto-tuned predictive control based on minimal plant information.

Area of Science:

• Control Engineering
• Industrial Automation
• Embedded Systems

Background:

• Constrained predictive control is crucial for low-level control loops but often unavailable.
• Industrial Programmable Logic Controllers (PLCs) are standard for automation tasks.
• Auto-tuning controllers enhance performance and reduce manual tuning effort.

Purpose of the Study:

• To implement constrained predictive control within an industrial PLC using IEC 61131-3.
• To develop and implement a novel auto-tuned predictive controller using pragmatic plant information.
• To validate the effectiveness of the combined hardware and algorithm solution through laboratory experiments.

Main Methods:

• Embedding a constrained control algorithm into an industrial PLC following the IEC 61131-3 standard.
• Defining and implementing a novel auto-tuned predictive controller based on simplified plant models.
• Conducting laboratory experiments on two bench-scale systems to evaluate performance.
• Comparing results against a commercial proportional-integral-derivative (PID) controller with auto-tuning.

Main Results:

• Successful integration of constrained predictive control into an industrial PLC environment.
• Demonstration of a novel auto-tuned predictive controller's effectiveness using basic plant data.
• Experimental validation showing the combined solution's efficacy in laboratory settings.
• Comparative analysis indicating performance advantages over a standard PID controller.

Conclusions:

• The developed constrained predictive control algorithm and hardware integration are effective for low-level control loops.
• The novel auto-tuned predictive controller offers a practical solution for industrial automation.
• The proposed method provides a viable alternative to traditional PID control in specific applications.