Prediction of the outputs of Two-CoupledDistillation Column system by Data-Driven method

Two-coupled distillation systems are mainly used to separate liquids from each other.Modeling and controlling these systems by model-based approaches are complex tasks, sothat’s why we used data-driven methods. Training the network is done by machine learningmethods to estimate the outputs of the system. Training is performed just by sampling theinputs and outputs data of the two-coupled distillation system, which is simulated in theMATLAB simulator. Note that, training the network is done without having a view of thesystem’s noise and disturbances. The linear regression method is used for training the model.The average training loss was ۰.۰۷.