Implementing a Process Model for Real-Time Applications
Yu‐Chu Tian, David Levy, Tianlong Gu · QUT ePrints (Queensland University of Technology) · 2004
Model-based advanced control of industrial processes requires well-developed process models. The established models are often complex with the features of multi-dimension, high nonlinearity, strong coupling, etc. Those features make the computational demand of the models conceptually high, especially for large-scale systems, resulting in difficulties in implementation of real-time control systems. This paper reports our efforts in overcoming the difficulties for a large-scale industrial process: a continuous hot dip galvanising production line. The process model is developed first, which is a set of partial differential equations. It is then transformed into a discrete state space model suitable for computation using digital computers. In order to meet the requirements of real-time control, algorithms are designed to significantly reduce the computational time. The process model and its computational algorithms are implemented and embedded into the real-time control system of the galvanising production line in the real plant.