Diagnosis de Sistemas Dinámicos Basada en Aprendizaje Supervisado Off-Line

摘要: THIS PAPER PRESENTS AN APPROACH TO DIAGNOSE DYNAMICAL SYSTEMS UNSING DECISION RULES GENERATED BY SUPERVISED LEARNING. THE METHODOLOGY CON START WITH REAL SYSTEM MODEL, BOTH OF THEM SHOWING EVOLUTION WHEN DIFFERENT COMPONENTS ARE FAULTY. BEFORE PROCESSING OBTAINNED INFORMATION A CLASSIFICATION TOOL, IT WILL BE TREATED. AFTER THAT, TOOL PROVIDE US CORRESPONDING BEHAVIOUR UNDER TRAINED SITUATIONS. TASKS OBTAIN MADE MAKING DIAGNOSIS. IS RUNNIG, DIAGNOSIS CAN ASSOCIATING OBSERVED OBTAINED RULES. WAY POSSIBLE DETECT WHWN FAULT ACCURS AND FAULTY COM