Prediction of the gradation stability of granular soils using machine learning techniques

IF 2.9 3区工程技术

Granular Matter Pub Date : 2025-07-22 DOI:10.1007/s10035-025-01562-3

Pingfan Wang, Xianqi Luo, Yunwei Shi

{"title":"Prediction of the gradation stability of granular soils using machine learning techniques","authors":"Pingfan Wang, Xianqi Luo, Yunwei Shi","doi":"10.1007/s10035-025-01562-3","DOIUrl":null,"url":null,"abstract":"<div><p>An innovative methodology for predicting gradation stability using integrated machine learning technologies is introduced. Current geometric criteria for suffusion assessment rely on a limited set of characteristic particle sizes, which results in a loss of detailed gradation information embedded in grading curves. This study proposes a new framework for evaluating the suffusion sensitivity through predicting the gradation stability of granular soil with a specified grading curve. Two distinct integrated machine learning models are developed to quantitatively assess soil internal stability. The predicted results and performance analysis demonstrate that the PCA-SVM model achieves superior classification accuracy for internal stability, while the PCA-ANN exhibits enhanced predictive capability in estimating the probability of internal stability within the given dataset. The proposed methodology provides a novel application for investigating the relationship between gradation characteristics and stability. This study will facilitate further research on establishing the accurate gradation stability criteria and predicting the soil suffusion sensitivity.</p><h3>Graphical Abstract</h3><div><figure><div><div><picture><source><img></source></picture></div></div></figure></div></div>","PeriodicalId":49323,"journal":{"name":"Granular Matter","volume":"27 4","pages":""},"PeriodicalIF":2.9000,"publicationDate":"2025-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Granular Matter","FirstCategoryId":"5","ListUrlMain":"https://link.springer.com/article/10.1007/s10035-025-01562-3","RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 0

Abstract

An innovative methodology for predicting gradation stability using integrated machine learning technologies is introduced. Current geometric criteria for suffusion assessment rely on a limited set of characteristic particle sizes, which results in a loss of detailed gradation information embedded in grading curves. This study proposes a new framework for evaluating the suffusion sensitivity through predicting the gradation stability of granular soil with a specified grading curve. Two distinct integrated machine learning models are developed to quantitatively assess soil internal stability. The predicted results and performance analysis demonstrate that the PCA-SVM model achieves superior classification accuracy for internal stability, while the PCA-ANN exhibits enhanced predictive capability in estimating the probability of internal stability within the given dataset. The proposed methodology provides a novel application for investigating the relationship between gradation characteristics and stability. This study will facilitate further research on establishing the accurate gradation stability criteria and predicting the soil suffusion sensitivity.

Graphical Abstract

Abstract Image

查看原文本刊更多论文

用机器学习技术预测颗粒土的级配稳定性

介绍了一种利用集成机器学习技术预测级配稳定性的创新方法。目前的渗透评估几何标准依赖于一组有限的特征粒径，这导致了在分级曲线中嵌入的详细分级信息的丢失。本研究提出了一种通过预测具有特定级配曲线的颗粒土级配稳定性来评价渗透敏感性的新框架。开发了两种不同的集成机器学习模型来定量评估土壤内部稳定性。预测结果和性能分析表明，PCA-SVM模型在内部稳定性方面具有较好的分类精度，而PCA-ANN模型在给定数据集内估计内部稳定性概率方面具有较强的预测能力。所提出的方法为研究级配特性与稳定性之间的关系提供了一种新的应用。本研究将为建立准确的级配稳定性判据和预测土体的渗透敏感性提供依据。图形抽象

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Granular Matter MATERIALS SCIENCE, MULTIDISCIPLINARY-MECHANICS

CiteScore

4.30

自引率

8.30%

发文量

期刊介绍： Although many phenomena observed in granular materials are still not yet fully understood, important contributions have been made to further our understanding using modern tools from statistical mechanics, micro-mechanics, and computational science. These modern tools apply to disordered systems, phase transitions, instabilities or intermittent behavior and the performance of discrete particle simulations. >> Until now, however, many of these results were only to be found scattered throughout the literature. Physicists are often unaware of the theories and results published by engineers or other fields - and vice versa. The journal Granular Matter thus serves as an interdisciplinary platform of communication among researchers of various disciplines who are involved in the basic research on granular media. It helps to establish a common language and gather articles under one single roof that up to now have been spread over many journals in a variety of fields. Notwithstanding, highly applied or technical work is beyond the scope of this journal.