基于无损检测和机器学习方法的粘土水力压裂裂缝图像分类研究

IF 8.4 1区工程技术 Q1 ENGINEERING, GEOLOGICAL

Engineering Geology Pub Date : 2025-05-22 DOI:10.1016/j.enggeo.2025.108149

Jia-He Zhang , Shi-Jin Feng , Qi-Teng Zheng , Xiao-Lei Zhang

{"title":"基于无损检测和机器学习方法的粘土水力压裂裂缝图像分类研究","authors":"Jia-He Zhang , Shi-Jin Feng , Qi-Teng Zheng , Xiao-Lei Zhang","doi":"10.1016/j.enggeo.2025.108149","DOIUrl":null,"url":null,"abstract":"<div><div>High-frequency ground penetrating radar (GPR) offers a precise and non-destructive method for assessing the distribution of internal soil fractures. This research develops a non-destructive fracture GPR testing platform for low-permeability contaminated soil to acquire GPR B-scan images under authentic environmental conditions. The reliable dataset of soil GPR image is collected. Furthermore, an Improved ResNet50 Version 3 (IRV3) network, featuring embedded self-attention modules and an enhanced bottleneck design, is presented and applied to real hydraulic fracturing laboratory testing using GPR. Comparisons of GPR images before and after fracturing revealed significant alterations in fracture distribution. Under the complex conditions of fracturing, the IRV3 network achieved a classification accuracy of 86.3 %. These results validate the reliability of the GPR testing platform constructed for simulating soil internal fractures and demonstrate the IRV3 network's applicability in experimental fracturing scenarios.</div></div>","PeriodicalId":11567,"journal":{"name":"Engineering Geology","volume":"354 ","pages":"Article 108149"},"PeriodicalIF":8.4000,"publicationDate":"2025-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Fracture image classification study of clay hydraulic fracturing based on non-destructive testing and machine learning methods\",\"authors\":\"Jia-He Zhang , Shi-Jin Feng , Qi-Teng Zheng , Xiao-Lei Zhang\",\"doi\":\"10.1016/j.enggeo.2025.108149\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>High-frequency ground penetrating radar (GPR) offers a precise and non-destructive method for assessing the distribution of internal soil fractures. This research develops a non-destructive fracture GPR testing platform for low-permeability contaminated soil to acquire GPR B-scan images under authentic environmental conditions. The reliable dataset of soil GPR image is collected. Furthermore, an Improved ResNet50 Version 3 (IRV3) network, featuring embedded self-attention modules and an enhanced bottleneck design, is presented and applied to real hydraulic fracturing laboratory testing using GPR. Comparisons of GPR images before and after fracturing revealed significant alterations in fracture distribution. Under the complex conditions of fracturing, the IRV3 network achieved a classification accuracy of 86.3 %. These results validate the reliability of the GPR testing platform constructed for simulating soil internal fractures and demonstrate the IRV3 network's applicability in experimental fracturing scenarios.</div></div>\",\"PeriodicalId\":11567,\"journal\":{\"name\":\"Engineering Geology\",\"volume\":\"354 \",\"pages\":\"Article 108149\"},\"PeriodicalIF\":8.4000,\"publicationDate\":\"2025-05-22\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Engineering Geology\",\"FirstCategoryId\":\"89\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S0013795225002455\",\"RegionNum\":1,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"ENGINEERING, GEOLOGICAL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Engineering Geology","FirstCategoryId":"89","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0013795225002455","RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, GEOLOGICAL","Score":null,"Total":0}

引用次数: 0

摘要

高频探地雷达（GPR）为评估土体内部裂缝的分布提供了一种精确、无损的方法。本研究开发了一种低渗透污染土壤无损裂缝探地雷达检测平台，以获取真实环境条件下的探地雷达b扫描图像。获得了可靠的土壤探地雷达图像数据集。此外，改进的ResNet50版本3 （IRV3）网络具有嵌入式自关注模块和增强的瓶颈设计，并应用于使用GPR进行的实际水力压裂实验室测试。对比压裂前后的GPR图像，发现裂缝分布发生了显著变化。在复杂压裂条件下，IRV3网络的分类准确率达到了86.3%。这些结果验证了模拟土壤内部裂缝的GPR测试平台的可靠性，验证了IRV3网络在压裂实验场景中的适用性。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

Fracture image classification study of clay hydraulic fracturing based on non-destructive testing and machine learning methods

High-frequency ground penetrating radar (GPR) offers a precise and non-destructive method for assessing the distribution of internal soil fractures. This research develops a non-destructive fracture GPR testing platform for low-permeability contaminated soil to acquire GPR B-scan images under authentic environmental conditions. The reliable dataset of soil GPR image is collected. Furthermore, an Improved ResNet50 Version 3 (IRV3) network, featuring embedded self-attention modules and an enhanced bottleneck design, is presented and applied to real hydraulic fracturing laboratory testing using GPR. Comparisons of GPR images before and after fracturing revealed significant alterations in fracture distribution. Under the complex conditions of fracturing, the IRV3 network achieved a classification accuracy of 86.3 %. These results validate the reliability of the GPR testing platform constructed for simulating soil internal fractures and demonstrate the IRV3 network's applicability in experimental fracturing scenarios.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

Engineering Geology 地学-地球科学综合

CiteScore

13.70

自引率

12.20%

发文量

327

审稿时长

5.6 months

期刊介绍： Engineering Geology, an international interdisciplinary journal, serves as a bridge between earth sciences and engineering, focusing on geological and geotechnical engineering. It welcomes studies with relevance to engineering, environmental concerns, and safety, catering to engineering geologists with backgrounds in geology or civil/mining engineering. Topics include applied geomorphology, structural geology, geophysics, geochemistry, environmental geology, hydrogeology, land use planning, natural hazards, remote sensing, soil and rock mechanics, and applied geotechnical engineering. The journal provides a platform for research at the intersection of geology and engineering disciplines.