Dynamic Identification of Actuator Frequency Response using Wavelets

Health monitoring on aircraft has traditionally been accomplished through periodic inspections based on probabilistic assessments of failure rates backed by component testing. Thus each component in an aircraft including the airframe has an a priori established inspection and/or replacement schedule. The highly vibratory environment in a helicopter coupled with potential single-point failure modes demands that they be inspected thoroughly very regularly at great cost. Approaches to health monitoring may be sensor-based or model-based. The sensor-based approach monitors parameters such as vibration and temperature at various locations in an aircraft and detects exceedences beyond a preset value. Almost all the existing health monitoring systems including rotorcraft HUMS are sensor based systems. The model-based approach including that proposed herein compares the actual performance or characteristic “dynamic signature” of an aircraft system or component with an appropriate model. The objective of this research is to decrease system down-time and lower the maintenance costs due to current high levels of false alarms. The end product will be a prototype built-in-test prognostics and health monitoring (PHM) application using frequency response identification.