Multi-Dimensional Logging Outlier Removal Using UMAP and Isolation Forest
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to achieve stable parameter acquisition during drilling due to the presence of outliers in multi-dimensional logging data, affecting the effectiveness of algorithm models used for stratum identification and target layer localization.
Innovation Solution
A method involving Umap dimensionality reduction and an isolation forest algorithm is employed to process multi-dimensional logging data, establishing data point indices and removing outliers, thereby improving the accuracy and speed of data processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual interpretation and machine learning algorithms are used for stratum identification, then detailed lithology distribution can be discriminated, but the algorithm models become complex and are greatly affected by input abnormal data
Solution Approach 1:
The patent applies outlier detection and removal as a preliminary action before feeding data into the machine learning algorithm. By preprocessing the logging data to identify and remove abnormal data points using statistical methods and domain knowledge, the input data quality is improved, which enhances algorithm stability without sacrificing the detailed lithology discrimination capability
Solution Approach 2:
The patent introduces an intermediary data processing layer between the raw logging data and the machine learning algorithm. This intermediary layer includes data cleaning, normalization, and feature selection steps that mediate between the complex multi-dimensional logging data and the algorithm model, improving reliability while preserving measurement precision
2Reliability
If algorithm models are established to remove outliers from multi-dimensional logging curves, then input training data can guide algorithm model updating more effectively, but the data processing time and computational complexity increase
Solution Approach 1:
The patent segments the outlier detection and removal process into distinct modular steps: initial outlier detection using statistical thresholds, iterative refinement based on algorithm feedback, and validation steps. This segmentation allows for efficient processing at each stage and enables parallel computation where applicable, reducing overall processing time while maintaining model stability
Solution Approach 2:
The patent applies partial action by removing only the most significant outliers that have the greatest impact on algorithm stability, rather than attempting to remove all potential abnormal data points. This selective approach maintains reliability while minimizing the time loss associated with extensive data processing
Data Source
AI summary
The present invention is in the field of geological measurement, and particularly relates to a method and system for intelligent removal of multi-dimensional logging data outliers, aiming to solve the problem that it is difficult to achieve stable parameter acquisition in the drilling process when there are more outliers in the prior art. The present invention comprises: acquiring a multi-dimensional logging curve; obtaining two-dimensional logging curve data by using the Umap dimensionality reduction algorithm to perform dimensionality reduction; establishing a first data point index of the multi-dimensional logging curve and the two-dimensional logging curve data; obtaining valid logging data by processing outliers with an isolation forest algorithm based on the two-dimensional logging curve data; acquiring a second data point index based on the first data point index and valid logging data; finding marked data points where the second data point index exists in the multi-dimensional logging curve data.