Murdoch University Research Repository

Welcome to the Murdoch University Research Repository

The Murdoch University Research Repository is an open access digital collection of research
created by Murdoch University staff, researchers and postgraduate students.

Learn more

Parallel Development of a Real-time Co-Simulation and MPC Control System for the Universal Water System

Allsopp, Joseph (2018) Parallel Development of a Real-time Co-Simulation and MPC Control System for the Universal Water System. Honours thesis, Murdoch University.

[img]
Preview
PDF - Whole Thesis
Download (7MB) | Preview

Abstract

Model Predictive Control (MPC) has had great success within the process industry. With its optimising capabilities and ability to handle complex interactions, MPC helps reduce operating costs while increasing production rates. A comprehensive internal process model is a prerequisite for MPC and consequently the process must be disturbed extensively with test signals during development or following modifications. This project focuses on the development of a cooperative simulation (co-simulation) to facilitate the end to end design of an MPC control system to be ultimately used on its real-world counterpart.

Murdoch University’s Universal Water System (UWS) was used as the test case with Honeywell’s Profit Suite selected as the MPC software platform. Following the derivation of a process model from first principles and empirical data, offline and real-time simulations were developed in Matrix Laboratory (MATLAB) and connected to Profit Suite using an OLE for Process Control (OPC) server. The offline simulation was used to generate data for controller model identification and the real-time version was used to virtually commission the MPC controller and for operator training. Profit Stepper was used to recreate the controller’s internal model by applying test signals directly to the real-world process. Analysis of controller performance for both controller models and conventional feedback control indicated that even with an imperfect controller model, MPC performance was superior to Proportional and Integral control, particularly when control loop interaction and disturbances were prominent. However, it was noted that MPC performance did improve with model accuracy. The benefit of the simulated design approach is that the controller model can be created without having to disturb the plant but it is limited by the accuracy of the process model.

Item Type: Thesis (Honours)
Murdoch Affiliation: School of Engineering and Information Technology
Supervisor(s): Bahri, Parisa
URI: http://researchrepository.murdoch.edu.au/id/eprint/44812
Item Control Page Item Control Page

Downloads

Downloads per month over past year