Abstract
The importance of Transfer Function Model comes from its use in prediction . The model has been constructed with multi-inputs because of the importance of multi-input in contructing models. Such multi-inputs give better results because the more effective inputs are used , the more precise are the results.In order to get the best prediction , the study studies the construction of transfer function model by black box models.
This method consists of constructing many models for the basic models of the black box Model,namely (ARX, ARMAX, BJ, OE ). The model which gave the best result was the model "BJ" depending , for comparison , on a group of statistical and geometrical standards .