Adaptive Wiener Filtering for Non-Stationary Microseismic Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to effectively recover microseismic signals from noisy observations, particularly in non-stationary environments, due to the challenge of distinguishing signal from noise and the assumption of stationary noise statistics, which is not valid in microseismic data.
Innovation Solution
A data-driven method using a linear Wiener filter that estimates noise and observation autocorrelations by defining filter design and correlation estimation windows, allowing for iterative signal recovery without prior knowledge of noise statistics, suitable for both single trace and array processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If conventional filtering methods (band pass filtering, spectral filtering) are used to enhance SNR, then noise removal is achieved, but signal attenuation occurs when signal and noise share the same frequency content
Solution Approach 1:
The patent applies dynamic adaptive filtering where filter parameters are continuously adjusted based on local signal characteristics. The filter adapts its impulse response to match the local signal properties, allowing effective noise suppression while preserving signal integrity in non-stationary environments where signal and noise frequency content may overlap.
Solution Approach 2:
The patent changes filter parameters dynamically based on local signal statistics. By computing adaptive filter coefficients from local signal segments and adjusting the filter impulse response accordingly, the system optimizes the balance between noise removal and signal preservation for each specific segment of the microseismic signal.
2Ease of operation
If noise statistics are assumed to be stationary for filtering, then filtering simplicity is maintained, but accuracy deteriorates in non-stationary microseismic environments
Solution Approach 1:
The patent segments the microseismic signal into local frames or segments and computes adaptive filter coefficients for each segment independently. This segmentation allows the filter to adapt to non-stationary characteristics while maintaining computational simplicity through localized processing, avoiding the need for complex global non-stationary models.
Solution Approach 2:
The patent implements dynamic adaptation of filter parameters for each local segment, allowing the filtering operation to respond to changing signal and noise characteristics over time. This dynamic approach maintains ease of operation through standardized filter algorithms while achieving high accuracy by adapting to non-stationary conditions in each segment.
3Measurement precision
If more complicated denoising methods are used to preserve signal strength, then signal attenuation is reduced, but device complexity increases
Solution Approach 1:
The patent employs self-adaptive filtering where the filter automatically computes its own parameters from the signal itself without requiring external calibration or complex manual configuration. The adaptive filter coefficients are derived directly from local signal statistics, enabling the system to achieve high signal preservation performance through automated, self-configuring processing rather than complex fixed algorithms.
4Measurement precision
If surface arrays with several thousand geophones are deployed to improve horizontal location accuracy, then location precision is enhanced, but cost and system complexity increase significantly
Solution Approach 1:
The patent applies local adaptive filtering to individual geophone traces or small groups, allowing each channel to be optimized independently for its local noise characteristics. This approach improves signal quality and location accuracy without requiring a large-scale array, as the enhancement comes from local signal processing rather than increased spatial sampling density.
Data Source
AI summary
A data-driven linear filtering method to recover microseismic signals from noisy data/observations based on statistics of background noise and observation, which are directly extracted from recorded data without prior statistical knowledge of the microseismic source signal. The method does not depend on any specific underlying noise statistics and works for any type of noise, e.g., uncorrelated (random/white Gaussian), temporally correlated and spatially correlated noises. The method is suitable for microquake data sets that are recorded in contrastive noise environments. The method is demonstrated with both field and synthetic data sets and shows a robust performance.