Feature Extraction of Down-hole Drilling-plug Impact Signal Based on Empirical Mode Decomposition

For analyzing the non-stationary drilling-plug impact signal under the complicated drilling environment, use of conventional discrete Fourier Transform (DFT) has been known to be less efficient. To extract useful information contained in the drilling-plug impact signal, a new extraction method based on empirical mode decomposition (EMD) is presented. In the signal processing, EMD is used to decompose the original drilling-plug impact signal into the intrinsic mode functions (IMFs). By analyzing kurtosis and energy spectrum of the IMFs, the first three IMFs are found that contain the characteristic of the drilling-plug, and the specific frequency band of the drilling-plug impact signal is determined after power spectrum analyzing. The reconstruction drilling-plug impact signal after band-pass filter can clearly show the detailed process of the drilling-plug such as drilling the rubber plug, drilling the float collars and drilling the formation. (paper)