PHM for air craft power supply generators using real-time frequency estimation based on 3-line DFT

With the advancement of technology and incorporation of sophisticated systems in the aircrafts, the modern aircrafts are highly dependent on the reliable electrical power systems. Prognostic and Health Management (PHM) is the evolving field for fault diagnosis, failure prevention and cost cutting maintenance. In this paper, a real-time frequency estimation algorithm based on 3-line Discrete Fourier transform (DFT) is presented for variable frequency measurement in aircraft power system to monitor the health state of the frequency generated. The signal is firstly weighed by means of Hanning window, and the spectrum amplitude in main lobe area is calculated according to DFT. The numerical value relation between fundamental frequency and 3-spectrum amplitude is then derived, performing long-range frequency estimation after iterative calculation of initial value. The number of samples is determined with no main lobe spectrum leakage; the spectrum amplitude is extracted in rotating reference frame to effectively reduce the computation effort. The proposed technique can simultaneously track the frequency with fast step respond and high steady-state accuracy. In the end, based on this measurement it is suggested to develop a PHM model on this OCM.