Convolutional Filter Subsurface Inversion Cycle-Skipping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Full-waveform inversion methods in seismic surveying often suffer from cycle-skipped mis-convergence, leading to inaccurate subsurface models due to the requirement for a highly accurate starting model and low-frequency data, which can be costly and impractical to obtain.
Innovation Solution
The method employs convolutional filters to transform seismic data sets, eliminating cycle-skipping by minimizing the misfit between filter coefficients rather than the data sets themselves, allowing for gradient inversion without cyclic issues and enabling accurate convergence with less accurate starting models and higher frequency data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional full-waveform inversion is used to achieve accurate subsurface models, then convergence accuracy is improved, but the requirement for highly accurate starting models and low-frequency data increases complexity and cost
Solution Approach 1:
The patent introduces convolutional filters as an intermediary between the observed and predicted seismic data sets. Instead of directly minimizing the misfit between data sets, the method transforms both data sets using convolutional filters and minimizes the misfit between filter coefficients. This intermediary approach eliminates cycle-skipping issues while maintaining convergence accuracy, avoiding the need for highly accurate starting models and low-frequency data.
Solution Approach 2:
The patent changes the parameter being minimized from data set misfit to filter coefficient misfit. By transforming the optimization target from the raw seismic data to the convolutional filter coefficients, the method fundamentally alters the inversion landscape, eliminating cyclic ambiguities and allowing accurate convergence without stringent requirements on starting model accuracy or data frequency content.
2Ease of operation
If conventional full-waveform inversion minimizes misfit between data sets, then direct comparison is simplified, but cycle-skipped mis-convergence occurs leading to inaccurate models
Solution Approach 1:
The convolutional filter serves as a mediator that transforms the direct data set comparison into a filter coefficient comparison. This intermediary transformation eliminates cycle-skipping while maintaining the essential comparison function, thereby improving convergence reliability without significantly complicating the misfit calculation process.
Solution Approach 2:
The patent substitutes the mechanical approach of direct data set subtraction with a signal processing approach using convolutional filters. Instead of directly comparing seismic traces, the method applies filters to both data sets and compares the filtered results, replacing the straightforward but unreliable mechanical subtraction with a more robust signal processing technique.
3Reliability
If starting models with high accuracy are used to avoid cycle-skipping, then convergence reliability is improved, but the cost and practicality of obtaining such models increases
Solution Approach 1:
The convolutional filter intermediary transforms the inversion problem in such a way that the starting model accuracy requirement is dramatically reduced. The filter coefficient minimization approach inherently handles cycle-skipping, allowing reliable convergence even with inaccurate starting models, thereby greatly easing model preparation requirements.
Solution Approach 2:
Instead of improving starting model accuracy to achieve reliable convergence, the patent inverts the approach: it modifies the inversion methodology itself to be insensitive to starting model accuracy. By minimizing filter coefficient misfit rather than data set misfit, the method achieves reliable convergence without requiring high-quality starting models.
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
Method of, and Apparatus for, Full Waveform Inversion There is provided a method of subsurface exploration, the method comprising generating a geophysical representation of a portion of the volume of the Earth from a seismic measurement of at least one physical parameter. The method comprises the steps of: a) providing an observed seismic data set comprising at least three distinct non-zero data values derived from at least three distinct non-zero seismic measured values of said portion of the volume of the Earth; b)generating, using a subsurface model of a portion of the Earth comprising a plurality of model coefficients, a predicted seismic data set comprising at least three distinct non-zero data values;c)generating at least one non-trivial convolutional filter, the or each filter comprising three or more non-zero filter coefficients;d)generating a convolved observed data set by convolving the or each convolutional filter with said observed seismic data set; e)generating one or more primary objective functions operable to measure the similarity and/or mismatch between said convolved observed dataset and said predicted dataset;f)maximising and/or minimising at least one of said primary objective functions by modifying at least one filter coefficient of the or each convolutional filter;g)generating one or more pre-determined reference filters comprising at least three reference coefficients;h) generating one or more secondary objective functions operable to measure the similarity and/or mismatch between the filter coefficients for the or each non-trivial filter and the reference coefficients for the or each pre-determined reference filters; andi)minimising and/or maximising at least one of said secondary objective functions by modifying at least one model coefficient of said subsurface model of a portion of the Earth to produce an updated subsurface model of a portion of the Earth; and j) providing an updated subsurface model of a portion of the Earth for subsurface exploration.