ORTHOGONAL LEAST SQUARE BASED NON-LINEAR SYSTEM IDENTIFICATION OF A REFRIGERATION SYSTEM

摘要

Chillers are important part of several processes in thernChemical, Petro-Chemical, Pharmaceutical, Beverage andrnFood industries. Controlling these processes at anrnadvantageous operating point is essential to achieve highrnproductivity and profitability. Ultimately control systemrndesign and controller tuning depend on accurate processrnknowledge in the form of dynamic mathematical models.rnBut attempts to develop analytical models often stumblernupon problems such as unknown physical parameters. Inrnthis work, system identification, an established modelingrntechnique, is used to build a non-linear dynamic model ofrna chiller from raw Input-Output data. Two dynamic nonlinearrnstochastic models where obtained, one morerncompact and the other with more terms but more precise,rnshowing good simulation results and average predictionrnerrors between 3.65%-5.23%.