Extended Six-Stage Recursive Least Squares Algorithm for linear Box Jenkins MISO model: Application To Distillation Unit

In this paper, a linear Box Jenkins model was developed, and the parameters of this model were estimated using the extended least squares algorithm. The development process involved a decomposition divided into six stages. The model was then applied to the distillation column system to predict the overhead temperature. To determine the best model, it was compared with other linear models through the calculation of performance metrics such as Root Mean Square Error (RMSE), and Nash-Sutcliffe Efficiency (NSE).