“…Performance of an MPC scheme depends predominantly on the choice of the model, that is, how good the model can predict the dynamic behaviour of the actual plant/process. There are variety of models available in the literature which are commonly used in MPC such as Hammerstein‐Wiener based model, input‐output data based model, operator theory based linear model, ARX model, ARMAX model, second‐order Volterra series model, ARX‐Volterra model, fuzzy rule based model, NN‐based model, ARX‐NN model, grey‐box model using state and output feedback, event‐driven observer‐based output‐feedback model, models obtained using support vector regression, and support vector machine techniques. Even though these models are quite useful and well established for predicting the behaviour of the actual system/process, they do not explicitly consider the effects of stochastic uncertainties like process and measurement noises, random variation in process parameters, stochastic disturbances, and unmodelled dynamics.…”