Hyper-complex signal alignment for well log correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current dynamic time warping (DTW) techniques face challenges in effectively aligning and correlating well log signals from multiple boreholes, especially when dealing with missing or null data points, which complicates the process of identifying geological formations and features in subterranean hydrocarbon exploration.
Innovation Solution
A robust signal alignment scheme is introduced that incorporates user-assigned picks or constraints, converted into complex number values, to generate hyper-complex signals. These signals allow for a single application of DTW, aligning sequences while accommodating user reliability weights to resolve conflicts, and extending the method to various applications involving multiple types of samples or components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional DTW techniques are used to align well log signals, then the alignment process can handle basic signal variations, but the method fails to effectively accommodate missing or null data points and user-defined constraints
Solution Approach 1:
The patent extends traditional DTW from operating in real-value signal space to operating in hyper-complex signal space. By incorporating user picks as additional dimensional constraints within the hyper-complex framework, the method can simultaneously process multiple signal types and user-defined alignment requirements, effectively resolving the contradiction between maintaining alignment accuracy and adapting to missing data with user constraints.
Solution Approach 2:
The patent creates composite signal structures by combining multiple signal types (real and imaginary components) into hyper-complex signals. This composite approach allows the DTW algorithm to process diverse data including missing values and user constraints simultaneously, improving both reliability in alignment and adaptability to various data conditions.
2Manufacturing precision
If multiple signal types and user picks are integrated into DTW, then user-defined alignments can be prioritized, but the computational complexity increases
Solution Approach 1:
The patent merges multiple signal types and user picks into a unified hyper-complex signal representation. By combining these elements into a single integrated framework rather than processing them separately, the method achieves precise user-defined alignments while managing computational complexity through consolidation of processing steps.
3Reliability
If user picks are converted into complex number values to create hyper-complex signals, then user reliability weights can be incorporated, but the data processing complexity increases
Solution Approach 1:
The patent transforms user picks into complex number values, changing the parameter representation from simple constraints to multi-dimensional complex values. This parameter transformation enables the incorporation of user reliability weights while managing processing complexity through the mathematical properties of complex numbers, which can encode multiple pieces of information in a unified structure.
Data Source
AI summary
A method for correlating data comprises acquiring a first sequence signal and a second sequence signal, wherein the first sequence signal comprises at least a first data point including a first set of components and the second sequence signal comprises at least a second data point including a second set of components; acquiring a first set of user picks and a second set of user picks, wherein the first and the second sets of user picks each contain a respective first and second correspondence between a component in the first set of components and a component in the second set of components; combining the first and second sets of user picks with the first and second sequence signals to create a first hyper-complex signal and a second hyper-complex signal; and performing signal alignment on the first and second hyper-complex signals.