Modelo estadístico en series temporales para la predicción de la temperatura de gas de tragante del horno alto

This work, through techniques seldom used in the ironmaking field, establishes a criterion to define the behaviour along the time of significant operating variables of the blast furnace. A time series autoregressive model has been developed, achieving with it an adequate forecasting level. The methodology used has been the Yule-Walker equations to establish the model, the Akaike test to choose the best one and the autocorrelation function analysis to check it.