PROGNOSTIC METHODS FOR ROTORCRAFT CONDITION MONITORING

PROGNOSTIC METHODS (TIME TO FAILURE PREDICTION) REQUIRE KNOWLEDGE OF FAILURE MODES, EARLY DETECTION OF CHANGE, AND LEARNING FROM EXPERIENCE. IN PREVIOUS AND ONGOING GOVERNMENT SUPPORTED RESEARCH, TIDG HAS PROVEN TECHNIQUES FOR EARLY FAULT DETECTION OF ROTORCRAFT MECHANICAL FAULTS AND IS CURRENTLY WORKING ON INTEGRATING MACHINE LEARNING TECHNOLOGY INTO THE DESIGN FOR AN ONBOARD MONITORING SYSTEM. RESEARCH INTO PROGNOSTIC TECHNIQUES INTEGRATES INTO THESE ONGOING PROGRAMS, CREATING A SYNERGY WHERE FUNDED RESEARCH CAN BE MUCH MORE EFFICIENT. IN THIS PHASE I PROGRAM, WE PROPOSE TO IMPLEMENT TRENDBASED PROGNOSTIC ALGORITHMS ON OUR IMADS HELICOPTER MACHINERY MONITORING SYSTEM (AS INSTALLED AT NADEP, PENSACOLA, AND AT TIDG), PERFORM RIG TETS ON BEARINGS, SINCE EARLY FAULT DETECTION ALGORITHMS ARE PROVEN, AND TO INTEGRATE MONITORING MACHINE LEARNING TECHNOLOGY, BEING DEVELOPED IN OTHER PROGRAMS, INTO THE PROGNOSTIC ALGORITHM DEVELOPMENT. THIS WILL RESULT IN AN INTEGRATED APPROACH TO A ROTORCRAFT MONITORING SYSTEM.