Model Predictive Control

Only available on StudyMode
  • Download(s) : 108
  • Published : March 10, 2013
Open Document
Text Preview
i ABSTRACT MPC is the most widely applied advanced control technique in industry. This project provides a brief overview of Dynamic Matrix Control which is the backbone of Model Predictive Control presents the reliable tuning strategy and its effects for unconstrained SISO dynamic matrix control (DMC),which has achieved a great deal of success in the petroleum refining and petrochemicals industries.. A brief history of early industrial MPC applications is given followed by some of its industrial uses. Then some basic structure of model predictive control is discussed. A brief idea about the DMC tuning is given. Diabatic Nonlinear CSTR process is investigated in MPC control technique. Firstly, step change is given to the nonlinear process and process reaction curve is obtained. After obtaining process reaction curve, transfer function of the process which is used in MPC controller as a model equation is determined. Then, the model predictive control technique is applied to a diabatic continously-stirred tank reactor (CSTR) system with to control the temperature and the concentration of the tank by manipulating the input flow rate. To control the CSTR system by the model predictive control technique, a Matlab code is generated to implement the MPC to the CSTR system models. The effect of each of the dynamic matrix controller tuning parameters over the CSTR system is observed and discussed throughly.

ii TABLE OF CONTENTS ABSTRACT ..................................................................................................................................... i LIST OF FIGURES ........................................................................................................................ iv LIST OF TABLES .......................................................................................................................... v 1. 2. INTRODUCTION ................................................................................................................... 1 MODEL PREDICTIVE CONTROL THEORY ..................................................................... 2 2.1. 2.2. 2.3. 2.4. Motivation ........................................................................................................................ 2 Benefits of MPC (Meadows and Rawlings,1997) ............................................................ 2 Industrial Implemtation (Qui and Badgwell, 1996).......................................................... 2 Commercial MPC Software (Qui and Badgwell, 1996) ................................................... 3

Basic Description ........................................................................................................................ 4 General Concept of MPC ............................................................................................................ 5 2.5. 2.6. 3. Optimization Problem (Bequette, 2003) ........................................................................... 6 Dynamic Matrix Control (Bequette,2003)........................................................................ 9

PROJECT STEPS ................................................................................................................. 15 3.1. 3.2. 3.3. Process Description of Diabatic CSTR with an Exothermic Irriversible Reaction ........ 15 Nonlinear Process Reaction (Diabatic CSTR) Curve with a Step Change.................... 18 Controlling Nonlinear Diabatic CSTR Process with MPC (Mismatch) ......................... 19 Effect of Control Weight (W) ................................................................................. 21 Effect of Prediction Horizon (P) ............................................................................. 22 Effect of Control Horizon (M) ................................................................................ 23 Effect of Model Length (N) .................................................................................... 24 Effect of...
tracking img