Extraction and identification to early fault feature of aircraft

In order to fast identify the early fault of aircraft, a method is presented by combination of wavelet with fractal. It is based on the early fault analysis of aircraft, and the singular characteristic signal can be extracted by the wavelet analysis and the early fault of aircraft can be identified by the fractal correlation dimension. An algorithm of wavelet adaptive de-noising is given and the selection of wavelet threshold is analyzed. At the same time, the extraction of early fault singular characteristic, the calculation of correlation dimension and the identification of early fault are did. The experimental results show that the singular signal of early fault can be effectively extracted by wavelet analysis and adaptive de-noising, and the early fault can be fast identified by fractal correlation dimensions. It provides an effective method for the early fault feature extraction of aircraft and the identification.