This project consists of the development of a security incident management model in the Supervision, Control and Data Acquisition (SCADA) systems based on event prediction, that is, through the collection of information on possible computer attacks makes a prediction with the Kalman filter mathematical model (which makes predictions of linear variables by means of recursive least squares), once the prediction of possible safety event is generated, the incident management process is activated; The final result is a security incident management model based on the early detection delivered by the Kalman filter, characterized by levels of impact, which will define the procedure appropriate to the criticality level of the prediction, thus allowing an integration and deployment of the best actions depending on the type of alert that is generated, thereby achieving a possible reduction in the levels of exposure to risks and reduction of possible impacts