Time-Varying Water Temperature Modelling of Steam Distillation Pilot Plant using NARX-based Binary Particle Swarm Optimisation Structure Selection

Many studies in current years has concentrated on both linear and nonlinear modelling in the real nonlinear system applications. This study reports a nonlinear modelling for a time-varying process of water temperature by utilising a Binary Particle Swarm Optimisation (BPSO) algorithm based on Nonlinear Auto-Regressive with eXogenous input (NARX) structure. The model structure selection of polynomial NARX has been concentrated on BPSO algorithm for system identification of Steam Distillation Pilot Plant (SDPP). Several model’s selection criteria such as Akaike Information Criterion (AIC), Model Descriptor Length (MDL), and Final Prediction Error (FPE) were investigated. The results demonstrated that all criterion models were considered valid and accurate representations of the system. The accuracy was evaluated by the high R-squared, small MSE value and passed all the correlation and histogram tests.