Extended Least Squares Algorithm for identification of linear CARARMA MISO model: Application To Distillation Column

In this work, a linear controlled autoregressive autoregressive moving average (CARARMA) model was developed to predict the overhead temperature of the distillation column system. The parameters of the achieved linear model are estimated by applying extended least squares algorithms. To determine the best model, various model structures were compared through the calculation of statistical criterions: (NSE) Nash-Sutcliffe Efficiency, (AIC) Akaike's Information Criterion and (RMSE) Root Mean Square Error.