Nonlinear Model-Based Control

Nonlinear Model-Based Control: Using First-Principles Models in Process Control

First-principles models (engineering models) are used in industry for process design, troubleshooting, training, online analysis and supervisory optimization. The author's vision is to use them for control. Why? They effectively handle nonlinearity, nonstationary behavior and interacting variables with just one tuning coefficient per controlled variable (CV). Using optimization, the controller can handle constraints and shape the manipulated variables to achieve desired controlled variable trajectories. Using first-principles models for control can also enhance the operational staff's understanding of the process, support auxiliary process management, and keep the mathematics at the engineers' comfort level. In addition, unifying all models across diverse process management operations ensures continuity and compatibility.

Control Techniques Using First-Principles Models

The book explains four control techniques using first-principles models that have been credibly demonstrated for industrial practice: generic model control, process-model-based control, predictive functional control and horizon predictive control. It illustrates their applications and discusses the pros and cons of each.

Understanding First-Principles Models

To provide a better understanding of first-principles models, the book includes examples of setting up functions for controllers and discusses inherent properties such as ease of tuning, the handling of nonlinearity and interaction, feedforward constraints and the range of operation.