An improved method of local mean decomposition with adaptive noise and its application to microseismic signal processing in rock engineering.

The processing and interpretation of microseismic (MS) signals are the basis for obtaining source information of MS events in rock engineering. However, MS signal processing is subjective and time-consuming, in which different strategies have to be empirically adopted to cope with different issues. To facilitate MS signal analysis, this study proposes an entirely data-driven method based on local mean decomposition (LMD), which has been applied in various signal processing fields. Through some numerical experiments, the superiorities (e.g., smaller reconstruction errors and fewer sifting iterations) of the proposed method over other methods are demonstrated, and the proposed method is identified as an effective filtering and frequency analysis tool to process recorded MS signals. Our analysis results show that the proposed method can effectively remove interfering noise existing in MS signals collected from the Lianghekou hydropower station during underground rock excavation. Moreover, the proposed method can be employed to analyze the frequency revolution features of MS signals before and after rock mass failure. In addition, it is found that MS signal frequency migrates to a low-frequency range when approaching rock failure, which can be regarded as an early warning indicator of rock instability. The proposed method is believed to facilitate MS monitoring and offer a new clue to conduct MS signal processing and interpretation. [ABSTRACT FROM AUTHOR]