REAL-TIME ADAPTIVE IDENTIFICATION AND PREDICTION OF FLUTTER

THE EXISTING METHODS FOR THE IDENTIFICATION AND PREDICATION OF FLUTTER DAMPING CHARACTERISTICS ARE NOT SUFFICIENTLY ACCURATE OR FAST FOR REAL-TIME WIND TUNNEL OR FLIGHT TESTING. THE OBJECTIVE OF THE PROPOSED PHASE I RESEARCH IS TO DEMONSTRATE THAT RECENT DEVELOPMENTS IN SYSTEM IDENTIFICATION AND MODEL STRUCTURE DETERMINATION CAN BE DEVELOPED INTO RELIABLE AND AUTOMATIC SOFTWARE ON MICROPROCESSORS FOR THE IDENTIFICATION, TRACKING, AND PREDICTION OF CHANGING FLUTTER CHARACTERISTICS. THE PROPOSED PHASE I RESEARCH TO DEMONSTRATE THIS FEASIBILITY INCLUDES MODIFICATION OF THE CANONICAL VARIATE SYSTEM IDENTIFICATION AND STRUCTURE DETERMINATION ALGORITHMS TO HANDLE CHANGING SYSTEMS, THE USE OF SIMULTANEOUS CONFIDENCE BANDS FOR PREDICTION OF FLUTTER DAMPING CHARACTERISTICS, DEMONSTRATION OF THE MODIFIED ALGORITHMS OF FLUTTER DATA, AND EVALUATION OF THE COMPUTATIONAL AND STORAGE REQUIREMENTSFOR MICROPROCESSOR IMPLEMENTATION. THE ANTICIPATED RESULTS ARE THE DEMONSTRATION OF AN ALGORITHM SUITABLE FOR MICROPROCESSOR IMPLEMENTATION WHICH RELIABLY AND AUTOMATICALLY IDENTIFIES, TRACKS, AND PREDICTS THE CHANGING FLUTTER DYNAMICS.