While-Drilling Porosity Inversion from Small-Sample Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining rock porosity in underground engineering, such as in subways and tunnels, are economically costly and operationally cumbersome, and there is a lack of effective methods to invert and predict porosity using small-sample while-drilling data due to complexity and external interference during data collection.
Innovation Solution
A method utilizing small-sample while-drilling data inversion, involving multi-stage denoising and a backpropagation regression-genetic algorithm (BP-GA) model, followed by constructing a rock porosity convolutional neural network (VG-CNN) prediction model to analyze time-frequency domain graphs of drilling parameters, optimizing the structural parameters for improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional laboratory testing methods are used to measure rock porosity, then measurement precision is improved, but productivity deteriorates due to time-consuming sample processing and testing
Solution Approach 1:
The patent replaces traditional mechanical laboratory testing methods with an intelligent algorithmic system. The VG-CNN model processes while-drilling data to predict porosity, substituting physical sample processing and laboratory measurement with computational analysis, thereby achieving rapid prediction without sacrificing accuracy
Solution Approach 2:
The patent performs porosity prediction during the drilling process itself, before the drilling operation is completed. By analyzing while-drilling parameters in real-time, the system obtains porosity information proactively, eliminating the need for subsequent sample processing and laboratory testing
2Measurement precision
If well logging and mud logging are conducted to characterize porosity, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent makes the drilling system multi-functional by enabling it to perform both drilling and porosity measurement functions simultaneously. The while-drilling parameters collected during normal drilling operations are reused for porosity prediction, eliminating the need for separate logging equipment and operations
Solution Approach 2:
The drilling system serves itself by using its own operational parameters (torque, propulsion, rotational speed, penetration rate, drill pipe amplitude, vibration acceleration) to determine rock porosity. The system leverages data already being collected during drilling, making the measurement process self-sufficient without external logging equipment
3Productivity
If while-drilling parameters are used for porosity inversion, then productivity is improved, but measurement precision deteriorates due to data noise and limited sample size
Solution Approach 1:
The patent employs dynamic data processing techniques including multi-stage denoising that adapts to varying signal characteristics, and a VG-CNN model that dynamically learns from data patterns. The system adjusts its processing approach based on the specific characteristics of while-drilling data, improving precision without sacrificing the rapidity of in-situ measurement
Solution Approach 2:
The patent transforms while-drilling parameters from their raw noisy state through multi-stage denoising processing, changing their quality parameters to make them suitable for accurate porosity inversion. The VG-CNN model further transforms the relationship between drilling parameters and porosity through learned mappings, improving measurement precision
4Device complexity
If small-sample while-drilling data is used for inversion, then device complexity is reduced, but measurement precision deteriorates due to insufficient data volume
Solution Approach 1:
The VG-CNN model acts as an intermediary that bridges the gap between limited while-drilling data and accurate porosity prediction. The model compensates for insufficient data volume by learning robust patterns from the available samples and generalizing to make accurate predictions, maintaining precision despite small sample size
Data Source
AI summary
A rock porosity prediction method based on small-sample while-drilling data inversion is provided, including: collecting while-drilling parameters during laboratory or field drilling processes, including parameters such as torque M, propulsion F, rotational speed N, drilling speed V, drill pipe amplitude A, and vibration acceleration a; inputting the different types of multi-stage denoised while-drilling parameter data into a backpropagation regression-genetic algorithm (BP-GA) model for discrete point elimination, data augmentation, and iterative calculation to form a new while-drilling dataset; inputting the time-frequency domain feature graphs of while-drilling parameters for rocks with different porosities into a rock porosity convolutional neural network (VG-CNN) prediction model, obtaining an inversion model between the time-frequency domain features of while-drilling parameters and porosity through training and learning, ultimately achieving inversion prediction of rock porosity according to real-time field while-drilling parameters.