Vectorized Disorder Algorithm for Seismic Randomness Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional seismic data processing methods, such as the Disorder Algorithm, struggle with accurately measuring randomness in seismic data due to anisotropy and non-smoothability issues, leading to flawed randomness estimations and difficulties in distinguishing faults from data randomness, especially in areas with significant diagonal structures.
Innovation Solution
The Vectorized Disorder Algorithm employs a vectorized convolution operation with an extra dimension and a nonlinear dimension reduction process to generate a randomness distribution, overcoming the limitations of existing methods by providing isotropic and smooth randomness estimations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional Disorder Algorithm is used for randomness measurement, then the measurement process is simple, but the measurement precision is poor due to anisotropy and non-smoothability issues
Solution Approach 1:
The patent introduces an additional dimension to the disorder operator, transforming it from a conventional 2D operator to a 3D volumetric operator. This dimensional extension enables isotropic randomness measurement by incorporating diagonal structural information that was previously missed, thereby resolving the anisotropy limitation while maintaining computational feasibility through vectorized operations.
Solution Approach 2:
The patent modifies the fundamental parameters of the disorder operator by adding a temporal or depth dimension, changing it from a static 2D filter to a dynamic 3D operator. This parameter change enables the operator to capture randomness in multiple directions simultaneously, achieving isotropic measurement and smoothability that were impossible with conventional 2D operators.
2Reliability
If conventional Disorder Algorithm is used, then computational speed is moderate, but reliability is poor due to inability to distinguish faults from randomness
Solution Approach 1:
By extending the disorder operator to three dimensions, the patent creates a volumetric randomness measurement that captures spatial correlations in all directions. This additional dimension provides more discriminative power to distinguish fault-related randomness from random noise, improving reliability without significantly impacting processing speed due to efficient vectorized implementation.
Solution Approach 2:
The patent creates multiple randomness measurements by applying the 3D vectorized disorder operator at different orientations and scales. These copied measurements from different perspectives are then integrated to produce a comprehensive randomness map that reliably distinguishes faults from random noise, enhancing detection reliability through multi-view analysis.
3Measurement precision
If conventional 2D disorder operator is used, then the operator dimensions match the data dimensions, but measurement precision is poor due to anisotropy in areas with diagonal structures
Solution Approach 1:
The patent resolves the anisotropy problem by adding a third dimension to the disorder operator, transforming it from a 2D to a 3D operator. This additional dimension enables the operator to capture randomness along diagonal directions and other orientations that 2D operators miss, achieving isotropic randomness measurement with uniform accuracy across all structural orientations.
Solution Approach 2:
The 3D vectorized disorder operator serves multiple functions simultaneously: it measures randomness in horizontal, vertical, and diagonal directions; it handles both 2D and 3D seismic data; and it provides both randomness magnitude and directional information. This multi-functionality achieves isotropic measurement accuracy without requiring separate operators for different orientations.
Data Source
AI summary
Systems and methods are disclosed for hydrocarbon exploration using seismic imaging and, more specifically, measuring randomness in seismic data utilizing a vectorized disorder algorithm. The vectorized disorder algorithm is configured to measure the randomness level (e.g., noise) in seismic data to improve seismic data processing/imaging and the ability to expose subsurface geology. The vectorized disorder algorithm includes performing convolution of seismic data with a vectorized disorder operator having an extra dimension than the seismic data. A nonlinear reduction operation is performed on the vectorized output to generate a randomness distribution dataset having the same dimension as the input data. The randomness distribution dataset comprises data points representing the level of randomness for respective seismic data points. A more accurate seismic image is generated from the seismic data as a function of the measured randomness distribution.