The SI Model Prediction VI determines the response of a system at time tbased on the output information available at time t – k and all the inputsapplied from time t – k to time t. k represents the size of the predictionwindow. Therefore, model prediction can determine how useful a model isin estimating future responses of the system, given all information at time tand an expected input profile in the future. Some control techniques takeadvantage of model prediction to improve control performance. Forexample, model predictive control uses some of the prediction properties ofa model to determine if a particular limitation or constraint is active in thefuture. This method allows the controller to take preventive actions beforesuch constraints become active.If you have the measured input and output of a system, you also canvalidate the model of the system by comparing the predicted output and themeasured output. If the prediction error is small, the model is acceptable.