Prediction of Unsaturated Soil Water Characteristic Curve Based on Machine Learning Algorithms

ZHANG Xiayang; GAO You; YU Xiang; HE Wei

doi:10.13409/j.cnki.jdpme.20240531003

您当前的位置：

首页 >

文章列表页 >

Prediction of Unsaturated Soil Water Characteristic Curve Based on Machine Learning Algorithms

更新时间：2025-03-10

- Prediction of Unsaturated Soil Water Characteristic Curve Based on Machine Learning Algorithms
- Journal of Disaster Prevention and Mitigation Engineering Vol. 45, Issue 1, Pages: 104-109(2025)
- 作者机构：
  
  1.宁波大学土木工程与地理环境学院，浙江宁波 315211
  2.湖南智谋规划工程设计咨询有限责任公司，湖南株洲 412000
- 作者简介：
- 基金信息：
- DOI：10.13409/j.cnki.jdpme.20240531003
  CLC： TU411
- Received：31 May 2024，
  
  Revised：2024-07-11，
  
  Published：15 February 2025
- 稿件说明：
移动端阅览
张夏阳,高游,于响等.基于机器学习算法的非饱和土水特征曲线预测[J].防灾减灾工程学报,2025,45(01):104-109.

ZHANG Xiayang,GAO You,YU Xiang,et al.Prediction of Unsaturated Soil Water Characteristic Curve Based on Machine Learning Algorithms[J].Journal of Disaster Prevention and Mitigation Engineering,2025,45(01):104-109.
张夏阳,高游,于响等.基于机器学习算法的非饱和土水特征曲线预测[J].防灾减灾工程学报,2025,45(01):104-109. DOI： 10.13409/j.cnki.jdpme.20240531003.

ZHANG Xiayang,GAO You,YU Xiang,et al.Prediction of Unsaturated Soil Water Characteristic Curve Based on Machine Learning Algorithms[J].Journal of Disaster Prevention and Mitigation Engineering,2025,45(01):104-109. DOI： 10.13409/j.cnki.jdpme.20240531003.

摘要

土水特征曲线（SWCC）是研究非饱和土渗透、强度预测与本构关系的基础。机器学习算法具有高效处理大量数据和特征提取等特点。采用六种机器学习算法（四种集成学习和两种传统机器学习算法）对美国非饱和土数据库中的154条SWCC包含1976个数据点进行模拟；并使用四个性能评价指标（

、EVS、MAE和RMSE）评价算法的性能。选取两种数据输入的方式：对压力水头进行对数处理和未处理两类。结果表明，在两种输入情况下，对LightGBM、XGB、RF和AdaBoost算法的影响很小；但是对GPR和SVM两种传统机器学习算法的影响很大，在未进行对数处理情况下，

降低明显甚至会出现无法模拟SWCC的情况。此外，LightGBM对SWCC测试集的模拟效果上均优于其他模型，拥有高的趋势评价指标（

和EVS）和低的误差测量指标（MAE和RMSE）；六种算法对SWCC模拟的优劣的排列顺序依次为：LightGBM、GPR、XGB、RF、AdaBoost和SVM。最后，利用已训练好的LightGBM模型对9条不包含在数据库内的SWCC数据进行预测，结果显示LightGBM能够较好地预测非饱和土的土水特性。研究结果对提升不同类型土的SWCC预测具有重要的指导意义。

Abstract

The soil water characteristic c

urve (SWCC) is fundamental for studying the permeability

strength prediction

and constitutive relationships of unsaturated soils. Machine learning algorithms are characterized by their efficiency in large dataset processing and feature extraction. This study used six machine learning algorithms (four ensemble learning and two traditional machine learning algorithms) to simulate 154 SWCCs with 1976 data points from the United States Unsaturated Soil Database. Four performance evaluation indicators (

EVS

MAE

and RMSE) were used to assess the algorithms' performance. Two types of data input methods were selected: one with logarithmic processing of matric suction

and the other without any transformation. The results showed that

under both input types

the effect on the LightGBM

XGB

and AdaBoost algorithms was minimal. However

the two traditional machine learning algorithms

GPR and SVM

were significantly affected. Without logarithmic transformation

decreased noticeably

and in some cases

the SWCC could not be simulated. Additionally

LightGBM outperformed other models in simulating the SWCC for the test set

with higher trend evaluation indicators (

and EVS) and lower error measurement indicators (MAE and RMSE). The ranking of the six algorithms in terms of SWCC simulation performance was as follows: LightGBM

GPR

XGB

AdaBoost

and SVM. Finally

the trained LightGBM model was used to predict 9 SWCC datasets not included in the original database. The results showed that LightGBM could effectively predict the soil water characteristics of unsaturated soils. The findings provide important guidance for improving SWCC predictions for different types of soils.

关键词

Keywords

references

毕银强，张茂省，谷天峰，等 . 非饱和黄土土水特征及其强度指标［J］. 防灾减灾工程学报， 2018 ， 38 （ 3 ）： 520 - 527 .

Bi Y Q ， Zhang M S ， Gu T F ， et al . Soil-water characteristics and strength indexes of unsaturated loess ［J］. Journal of Disaster Prevention and Mitigation Engineering ， 2018 ， 38 （ 3 ）： 520 - 527 . （in Chinese）

黎澄生，孔令伟，柏巍，等 . 土-水特征曲线滞后阻塞模型［J］. 岩土力学， 2018 ， 39 （ 2 ）： 598 - 604 .

Li C S ， Kong L W ， Bai W ， et al . Hysteresis model of soil-water characteristic curve ［J］. Rock and Soil Mechanics ， 2018 ， 39 （ 2 ）： 598 - 604 . （in Chinese）

王叶娇，王有为，靳奉雨，等 . 石灰改良土的土水特征曲线及其冻结特征曲线［J］. 防灾减灾工程学报， 2020 ， 40 （ 6 ）： 967 - 973 .

Wang Y J ， Wang Y W ， Jin F Y ， et al . Soil⁃water characteristic curve and freezing characteristic curve of lime improved Soil ［J］. Journal of Disaster Prevention and Mitigation Engineering ， 2020 ， 40 （ 6 ）： 967 - 973 . （in Chinese）

Fredlund D G ， Rahardjo H . Soil mechanics for unsaturated soils ［M］. New York ： John Wiley and Sons Inc ， 1993 .

刘倩倩，李舰，蔡国庆，等 . 全吸力范围的盐渍土持水特性的试验研究［J］. 岩土力学， 2021 ， 42 （ 3 ）： 713 - 722 .

Liu Q Q ， Li J ， Cai G Q ， et al . Experimental study on water retention characteristics of saline soil in the full suction range ［J］. Rock and Soil Mechanics ， 2021 ， 42 （ 3 ）： 713 - 722 . （in Chinese）

桂跃，刘锐，赵振兴，等 . 高分解度泥炭土-土水特征曲线研究［J］. 防灾减灾工程学报， 2021 ， 41 （ 3 ）： 622 - 628 .

Gui Y ， Liu R ， Zhao Z X ， et al . Study on soil water characteristic curve of peaty soil with high degree of decomposition ［J］. Journal of Disaster Prevention and Mitigation Engineering ， 2021 ， 41 （ 3 ）： 622 - 628 . （in Chinese）

Gao Y ， Sun D A . Soil-water retention behavior of compacted soil with different densities over a wide suction range and its prediction ［J］. Computers and Geotechnics ， 2017 ， 91 ： 17 - 26 .

Van Genuchten M T . A closed‑form equation for predicting the hydraulic conductivity of unsaturated soils ［J］. Soil Science Society of America Journal ， 1980 ， 44 （ 5 ）： 892 - 898 .

Fredlund D G ， Xing A . Equations for the soil-water characteristic curve ［J］. Canadian Geotechnical Journal ， 1994 ， 31 （ 4 ）： 521 - 532 .

Or D ， Tuller M . Liquid retention and interfacial area in variably saturated porous media： Upscaling from single‐pore to sample‑scale model ［J］. Water Resources Research ， 1999 ， 35 （ 12 ）： 3591 - 3605 .

刘樟荣，叶为民，崔玉军，等 . 基于微孔填充和毛细管凝聚理论的持水曲线模型［J］. 岩土力学， 2021 ， 42 （ 6 ）： 1549 - 1556 .

Liu Z R ， Ye W M ， Cui Y J ， et al . Water retention curve model based on micro-pore filling and capillary condensation theories ［J］. Rock and Soil Mechanics ， 2021 ， 42 （ 6 ）： 1549 - 1556 . （in Chinese）

Zhai Q ， Rahardjo H ， Satyanaga A ， et al . Estimation of wetting hydraulic conductivity function for unsaturated sandy soil ［J］. Engineering Geology ， 2021 ， 285 ： 106034 .

Hafez H ， Teirelbar A ， Kurda R ， et al . Pre-bcc： A novel integrated machine learning framework for predicting mechanical and durability properties of blended cement concrete ［J］. Construction and Building Materials ， 2022 ， 352 ： 129019 .

黄志，周芙蓉，陈娟，等 . 基于IPOA的巨型组合框架结构震损快速预测模型研究［J］. 防灾减灾工程学报， 2024 ， 44 （ 2 ）： 263 - 272 .

Huang Z ， Zhou F R ， Chen J ， et al . Rapid prediction model for seismic damage in mega composite frame structures based on IPOA methods ［J］. Journal of Disaster Prevention and Mitigation Engineering ， 2024 ， 44 （ 2 ）： 263 - 272 . （in Chinese）

Schaap M G ， Bouten W . Modeling water retention curves of sandy soils using neural networks ［J］. Water Resources Research ， 1996 ， 32 （ 10 ）： 3033 - 3040 .

Lamorski K ， Pachepsky Y ， Sławiński C ， et al . Using support vector machines to develop pedotransfer functions for water retention of soils in Poland ［J］. Soil Science Society of America Journal ， 2008 ， 72 （ 5 ）： 1243 - 1247 .

陈勇，邹皓天，苏剑，等 . 基于数据挖掘的土水特征曲线影响分析及预测研究［J］. 应用基础与工程科学学报， 2023 ， 31 （ 2 ）： 451 - 466 .

Chen Y ， Zou H T ， Su J ， et al . Impact analysis and prediction research of soil-water characteristic curves based on data Mining ［J］. Journal of Basic Science and Engineering ， 2023 ， 31 （ 2 ）： 451 - 466 . （in Chinese）

Qin W ， Fan G . Estimation and predicting of soil water characteristic curve using the support vector machine method ［J］. Earth Science Informatics ， 2023 ， 16 （ 1 ）： 1061 - 1072 .

Zhan J ， Li Z ， Yu X ， et al . Pedo-transfer functions of the soil water characteristic curves of the vadose zone in a typical alluvial plain area in the lower reaches of the Yellow River using machine learning methods ［J］. Environmental Monitoring and Assessment ， 2022 ， 194 （ 12 ）： 850 .

Nemes A ， Schaap M G ， Leij F J ， et al . Description of the unsaturated soil hydraulic database UNSODA version 2.0 ［J］. Journal of Hydrology ， 2001 ， 251 （ 3-4 ）： 151 - 162 .

Pham K ， Kim D ， Le C V ， et al . Machine learning-based pedotransfer functions to predict soil water characteristics curves ［J］. Transportation Geotechnics ， 2023 ： 101052 .

Zhang W ， Wu C ， Zhong H ， et al . Prediction of undrained shear strength using extreme gradient boosting and random forest based on Bayesian optimization ［J］. Geoscience Frontiers ， 2021 ， 12 （ 1 ）： 469 - 477 .

Chen Y ， Xu Y ， Jamhiri B ， et al . Predicting uniaxial tensile strength of expansive soil with ensemble learning methods ［J］. Computers and Geotechnics ， 2022 ， 150 ： 104904 .

Sanders R . The pareto principle： its use and abuse ［J］. Journal of Services Marketing ， 1987 ， 1 （ 2 ）： 37 - 40 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Research on Construction Alignment Prediction of Continuous Rigid‑frame Bridges Based on Machine Learning

Research on Slope Stability Prediction Based on PSO‑LightGBM Model

Research Progress on Fire Resistance of Reinforced Concrete Structures

Ground Settlement Prediction Method for Shield Tunneling Considering Rigid Pavement Thickness

Rapid Prediction Model for Seismic Damage in Mega Composite Frame Structures Based on IPOA Methods

Related Author

WANG Li

ZHANG Hongyan

CHEN Peng

WANG Bing

ZUO Hongpeng

ZHANG Ning

ZHANG Yu

ZHANG Shijie

Related Institution

School of Civil Engineering, Lanzhou Jiaotong University

Shanxi Traffic Construction Engineering Quality Inspection Center Co., Ltd.

Anhui Provincial Key Laboratory of Intelligent Construction and Disaster Prevention and Control of Geotechnical Engineering , Anhui Jianzhu University

Collega of Civil Engineering , Anhui Jianzhu University

School of Civil Engineering and Transportation, South China University of Technology

⁰