Application of Kalman filter in microseismic data denoising based on identified signal model

Kalman Filter has been used for microseismic data filtering based on mechanism signal model. However, the model was built with many assumptions and simplifications, which did not fully represent the characteristics of microseismic signals and affected the effect of Kalman filtering. In order to establish a more accurate mathematical model of microseismic signals and improve the effect of Kalman filtering, this paper builds an ARMA model for a typical microseismic event by identification methods. The simulation of the theoretical model and the synthetic signals shows that the identification method and model are accurate. Through the processing and analysis with the synthetic signals and the practical ground monitoring data of microseismic, the Gaussian white noise added in the microseismic signals is suppressed and the SNR is improved significantly by Kalman filtering based on identification model, which verifies the feasibility of the identification model and the filtering algorithm.