Adaptive notch filter based on LMS algorithm and its application in ground tilt data processing

The basic principles of adaptive notch filter are derived in detail in this paper. The possibility of gaining reference signal by the use of signal frequency estimation through the approach of adaptive notch filter is studied. The numerical experiments extracting the specific harmonic wave by using this notch filter show that when the frequency of the extracted composition is very close to that of the noise composition, the filtering effect of the adaptive notch filter will be obviously decreased, and the close range is 1e-3~1e-4. The idea of restraining various types of tides in sequence is proposed at first and it obtains the long-term trend of tides. The three types of ground tides are effectively filtered out respectively to pick-up the long-term trend of the tilt observed data. The research results show that the adaptive notch filter has a lower frequency-drift range and performs well in process of the observed data so that its application is promising in the signal process of surface-wave data, free oscillation wave data and earthquake precursor acquisition data.