Jet Grouting Pile Control Using Stratum Inversion and XGBoost
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing jet grouting pile construction methods face challenges in accurately determining soil layer distribution and controlling construction parameters, leading to uneven pile strength, material waste, and low efficiency due to complex stratum conditions.
Innovation Solution
An intelligent construction control method and system using stratum information inversion, employing XGBoost machine learning and a self-learning parameter matching model to dynamically adjust construction parameters in real-time based on drilling data, including torque, axial force, and pore water pressure, to optimize the construction process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If field test is performed to determine construction parameters by selecting a section with similar conditions, then construction parameters can be obtained, but the derived parameters may not be suitable for actual soil layer due to complex stratum conditions, resulting in low construction efficiency and poor quality
Solution Approach 1:
The patent replaces traditional mechanical field testing methods with electromagnetic detection technology. The electromagnetic detector measures soil electrical properties to identify soil layers, substituting physical drilling and sampling with non-intrusive electromagnetic field measurement, thereby improving both accuracy and efficiency
Solution Approach 2:
The patent introduces soil electrical properties as an intermediary parameter to infer soil layer characteristics. Instead of directly measuring physical soil properties through field tests, the system uses electrical conductivity and resistivity measurements as intermediate indicators to determine soil layer distribution and optimize construction parameters
2Adaptability or versatility
If construction parameters are determined by field test in similar soil sections, then some construction guidance can be provided, but the parameters are not dynamically adjusted to actual soil conditions, causing uneven pile strength and material waste
Solution Approach 1:
The patent transforms static construction parameters determined from field tests into dynamic parameters that are continuously adjusted during construction. The system real-time monitors soil layer changes through electromagnetic detection and automatically adjusts construction parameters (jetting pressure, lifting speed, grouting amount) to match actual soil conditions at each depth
Solution Approach 2:
The patent implements a closed-loop feedback system where electromagnetic detection continuously monitors soil electrical properties, the control system processes this information to identify soil layers, and construction parameters are automatically adjusted based on the identified soil conditions. This real-time feedback ensures parameters remain adaptive to actual soil variations throughout the construction process
3Loss of information
If current value of power head is monitored to determine soil layer distribution, then some information can be obtained, but the current value is affected by penetration pressure, speed and depth, making it inaccurate for determining soil layer
Solution Approach 1:
The patent replaces mechanical monitoring of power head current with electromagnetic detection technology. Instead of using electrical current as a proxy for soil conditions (which is confounded by multiple factors), the system directly measures soil electrical properties using electromagnetic fields, providing accurate soil layer identification independent of drilling parameters
Solution Approach 2:
The patent uses changes in electrical properties (analogous to color changes in visual detection) to identify soil layers. The electromagnetic detector measures variations in soil electrical conductivity and resistivity, which change characteristically across different soil layers, providing a clear diagnostic signal for soil layer identification
Data Source
AI summary
The present invention relates to pile foundations and provides an intelligent construction control method and system for jet grouting piles based on stratum information inversion. This method integrates multi-source sensing components and local transmission to perceive real-time drilling information to a data integration terminal. Using engineering soil layer data as training samples, a machine learning algorithm correlates perception information with geological conditions, judging and outputting stratum conditions in real time. An XGBoost model quickly and accurately identifies soil layers at the site. Based on big data matching of a cloud platform, an optimal construction parameter database is established, and the optimal combination of parameters is acquired through algorithm matching. A variable-frequency and variable-speed pressure adjusting device links a piling machine intelligent control system with other subsystems to achieve intelligent control, reducing material consumption and improving pile quality and efficiency.