System identification, estimation and controller design of a hot air blower system

Abstract

This project presents an importance task of System Identification, parameter estimation and model validation to develop a mathematical model that describes the dynamics of a hot air blower system. A PT326 process trainer is a hot air blower system used in this project. The scope of work for this project consists of modeling and controller design of a PT326 process trainer. A heating ventilation model is the system to be modeled and is perturbed by a Pseudo Random Binary Sequences (PRBS) signal. Parametric approach using Auto Regressive with Exogenous Input (ARX) model structure is used to estimate the mathematical model of PT326 process trainer. The System Identification Toolbox GUI in MATLAB environment is used to estimate this approximated plant model. Once the estimated plant model is validated using Model Validity Criterion method, the behavior of the system without applied any controller have been analyzed using MATLAB Simulink and result shows that the output responds does not corresponds to its input; the output temperature of air flowing is not maintained at a desired level. Several controllers such as Pole-Assignment Servo-Regulator controller, Proportional-Integral-Derivative (PID) controller, and Generalized Minimum Variance (GMV) controller were designed using the approximated plant model obtained and the performance of each controller was compared and justified by running a simulation. Simulation results demonstrated that in most cases, a Self-Tuning Pole Assignment Servo-Regulator controller with a small value of pole provide relatively high ability in controlling the system and a GMV controller using PSO tuning method obviously has improved the performance of the Self-Tuning GMV controller in term of rise time (Tr) and settling time (Ts).