Auto Regressive eXogenous (ARX) System Identification of Batch Milk Cooling Process

The dynamic model of the milk cooling process from 36°C to 4°C using chilled water available at 2°C has been carried out. temperature is kept constant by a refrigeration unit. being studied was Packo brand tank belonging KUD SAE Pujon (Malang - Indonesia). A fundamental heat balance method used derive model, leading first-order transfer function process. For 2 hours then, gain and time values are 1.00 42.3548 mins respectively, or G(s)=1/(42.3548s+1) (first order process). Deriving system through mechanistic considered difficult; therefore, in this paper, we explored identification via Auto Regressive eXogenous (ARX). Transient simulations could then be performed identify behavior identified several orders (ARX) results were re-tested on different forms perturbations obtained quite accurate results. ARX111 G(s)=1/(42.3729s+1) process), while ARX441, 5th obtained: G(s)=(0.02361s^4+0.000371s^3+0.2331s^2+9.27×10^(-7) s+0.0005826)/(s^5+0.03932s^4+9.873s^3+0.2331s^2+0.02468s+0.0005826). These models particularly useful for control design analysis.