Evaluation of gap-metric based multi-model control schemes for nonlinear systems: An experimental study

Area of Science:

• Control Engineering
• Nonlinear Process Control
• Automation Systems

Background:

• Multi-model approaches are common for nonlinear system control, but often simulation-based.
• Experimental validation is crucial for understanding practical challenges in nonlinear process control.
• Existing methods lack comprehensive experimental comparison of weighting strategies.

Purpose of the Study:

• To experimentally evaluate and compare gap metric-based weighting methods for multi-model control.
• To design and implement multi-model control schemes for level control in spherical and conical tank processes.
• To assess the practical effectiveness of different weighting functions (1-δ and 1/δ) in a real-world setting.

Main Methods:

• Design of Internal Model Control (IMC) Proportional-Integral (PI) controllers for linearized models.
• Simulation study to examine initial performance of level control on nonlinear systems.
• Calculation and application of gap metric values to determine weighting functions for controller combination.
• Experimental implementation and comparison of multi-model control schemes using (1-δ) and (1/δ) weighting functions.

Main Results:

• Successful experimental implementation of multi-model control schemes for level control in both tank types.
• Demonstrated effectiveness of gap metric-based weighting functions in constructing global multi-model controllers.
• Comparative performance analysis highlighting the practical applicability and challenges of the evaluated methods.

Conclusions:

• Experimental evaluation confirms the viability of gap metric-based weighting for multi-model control design.
• The study provides practical insights into the implementation of nonlinear process control strategies.
• The presented methods offer a robust approach for level control in challenging tank systems.