Soil Classification Methods Research Articles

An Advanced Soil Classification Method Employing the Random Forest Technique in Machine Learning

An Advanced Soil Classification Method Employing the Random Forest Technique in Machine Learning

Automatic soil classification method from CPTU data based on convolutional neural networks

Automatic soil classification method from CPTU data based on convolutional neural networks

Multi-fidelity fusion for soil classification via LSTM and multi-head self-attention CNN model

An effective soil classification method is essential for soil stratigraphy interpretation. Traditional methods rely on classification systems in design codes or empirical formulae based on single source data, which may not be accurate or suitable for every site. Recent advancements in data fusion techniques and deep learning have shown promising results in a wide range of practical domains, which provides motivation to explore its potential application in soil classification. To that end, this paper introduces a novel machine learning architecture using boosting long-short term memory (LSTM) machine to map low fidelity cone penetration test (CPT) data to high fidelity laboratory test (LT) data which are then fused and fed into a multi-head self-attention convolutional neural network (MSCNN) for soil classification. A rigorous loss function is developed to train a LSTM-MSCNN model in an end-to-end workflow. The MSCNN classifier is benchmarked against traditional machine learning methods to demonstrate its superior outperformance, with ablation analyses exploring the related merits of the attention mechanism and the influence of attention heads on classification results. The results reveal a significant improvement in the prediction of true ground profiles using the proposed LSTM-MSCNN model over conventional methods.

A comparative study of soil classification machine learning models for construction management

Prior to the construction of buildings and roads, soil classification plays a vital role to ensure proper project management and durability of the structure. Civil engineers need to possess knowledge about soil behavior and properties of soil classification to construct reliable and long-lasting infrastructures. However, traditional soil classification methods are both expensive and time-consuming. Recently, machine learning (ML) has become increasingly popular in solving complex problems in Geotechnical Engineering, leading to novel approaches for automating soil classification. This research evaluates the effectiveness of various machine learning algorithms, including Multinomial Logistic Regression (MLR), Gaussian Naive Bayes (GNB), Extreme Gradient Boosting (XGBoost), Random Forest, and Artificial Neural Network-Multilayer Layer Perceptron (ANN-MLP), in classifying soils. The paper also implemented Hard and Soft Voting Ensemble models. Each model was quantitatively evaluated and compared using various metrics. The models achieved high scores in multiple performance metrics across the soil classes with accuracies ranging between 87% and 93%, except the Naive Bayes model with a balanced accuracy of 77.6% while the hard-voting model outperformed the others with a balanced accuracy of 93%. The study’s results benefit industry practitioners and academic researchers for integrating and assessing ML algorithms in soil classification.

Soil Classification and Crop Suggestion using Machine Learning Techniques

In agriculture, the optimization of crop selection based on soil characteristics plays a pivotal role in maximizing yields and ensuring sustainable agricultural practices. Traditional methods of soil classification and crop suggestion rely heavily on manual labour and expertise, often resulting in time-consuming processes with limited accuracy. However, the integration of machine learning techniques presents a promising solution to this challenge

Knowledge-based U-Net and transfer learning for automatic boundary segmentation

An effective soil classification method is essential for soil layering interpretation. Traditional methods rely on empirical formulae or a general unified system, which may not be accurate for every site. Recent advancements in deep learning have shown promising results in a wide range of practical domains, which provides motivation to explore its potential application in soil classification. A “U-shaped” convolutional neural network (U-Net), coupled with multiple source data to incorporate prior knowledge, is adopted in this study to predict the depths of soil stratum boundaries based on cone penetration testing (CPT), standard penetration testing (SPT) and laboratory index testing data. The U-Net model is first pre-trained on open-access data from other sites around the world and then trained through transfer learning on datasets gathered specifically for the Suzhou No.6 metroline project. Comparison is made from different metrics to investigate how pre-training benefits model performance. The soil boundaries in nine selected output images are extracted and reversed to their original depth for visualization of a subsurface profile. Comparisons to the benchmark interpretations show that the predicted soil profiles have reasonably good agreement with the benchmark profiles developed by an engineering expert. The results indicate that the proposed method is effective and efficient for the prediction of soil stratification.

A new empirical model for the undrained shear strength of undisturbed marine fine-grained soil based on piezocone penetration tests

Simple, precise, and economical evaluation of the undrained shear strength Su of undisturbed marine soil can facilitate the construction of offshore projects. In this study, piezocone penetration tests (CPTu) were conducted at the site of a wind power project, and consolidated undrained triaxial tests were performed on specimens of undisturbed marine fine-grained soil taken from the site. Analysis shows that the Eslami–Fellenius and Chinese soil classification methods based on CPTu data are more applicable for marine soils. The test results indicate that the stress–strain curve exhibits strain softening if the clay content is less than 30 %, otherwise it exhibits either strain hardening or strain stability. Furthermore, the specimens exhibit two deformation failure modes: softening contraction and hardening dilatancy. From the correlation between the consolidated undrained triaxial test data and the CPTu data, Su can be determined simply via an explicit expression involving the measured pore pressure behind the cone (u2), the corrected cone tip resistance (qt), and the soil depth H, and this expression is significantly more precise than the existing empirical formulas. The proposed method offers a sound basis for designing the foundations of marine engineering and infrastructure projects in different sea region.

Assessing the Performance of Machine Learning Algorithms for Soil Classification Using Cone Penetration Test Data

Conventional soil classification methods are expensive and demand extensive field and laboratory work. This research evaluates the efficiency of various machine learning (ML) algorithms in classifying soils based on Robertson’s soil behavioral types. This study employs 4 ML algorithms, including artificial neural network (ANN), random forest (RF), support vector machine (SVM), and decision trees (DT), to classify soils from 232 cone penetration test (CPT) datasets. The datasets were randomly split into training and testing datasets to train and test the ML models. Metrics such as overall accuracy, sensitivity, precision, F1_score, and confusion matrices provided quantitative evaluations of each model. Our analysis showed that all the ML models accurately classified most soils. The SVM model achieved the highest accuracy of 99.84%, while the ANN model achieved an overall accuracy of 98.82%. The RF and DT models achieved overall accuracy scores of 99.23% and 95.67%, respectively. Additionally, most of the evaluation metrics indicated high scores, demonstrating that the ML models performed well. The SVM and RF models exhibited outstanding performance on both majority and minority soil classes, while the ANN model achieved lower sensitivity and F1_score for minority soil class. Based on these results, we conclude that the SVM and RF algorithms can be integrated into software programs for rapid and accurate soil classification.

Soil Salinity Monitoring in Irrigated Areas of Rishtan District of Ferghana Valley, Uzbekistan

The salinization of soils impacts many arid areas. In the world, about 62 million hectares of agricultural land are subject to salinization. At present, about 47% of the irrigated lands in Uzbekistan are subject to some degree of salinity. Landsat satellite photos taken across an exact time span (from 1993 to 2021) with timeframes of 8 to 10 years were utilized in this study to monitor soil salinity. To analyze soil salinity level, salinity index, and classification method using Google Earth was performed. During the entire study period, changes in the indicator of soil salinity improvement work were shown. Especially related to non-saline soil, which was around 4,670 ha in 1993, and it increased to 10,533 ha in 2021 and reached 19.2 % after thirty years. Slightly saline soil class has the propensity to decrease and included its largest area of 15,941 ha (52.1 %) in 1993, but then dropped quickly to 9,649 ha (31.5 %) in 2021. Saline soil maximum increased in 2001 to 10.251 ha (33.5 %), then started steadily contracting and reaches a minimum of 7,440 ha (24.3%) in 2011. Strongly saline soil starts reaching from 1,771 ha to 5,069 ha between 1993-2001.

Systematic Review of the Literature on Various Soil Classification Methods

Systematic Review of the Literature on Various Soil Classification Methods

Applicability and improvement of soil classification methods in Delta regions based on the CPTU database

Several soil classification systems have been developed for specific sites based on the CPTU database. However, issues may occur when applying these existing soil classification methods to delta regions, whose construction sites involve deposits that are both weak and compressible. This requires further evaluation and a more detailed dataset. A database has been compiled that includes 85 CPTU soundings from 36 sites in the Yangtze River Delta. This study evaluates the performance of four existing CPTU-based soil classification methods for deltaic soils. The performance of each soil classification method is quantified by the micro F 1 score, the weighted kappa coefficient, and the chi-square value. An analysis and comparison indicate that the SBT chart has the highest accuracy, with a micro F 1 score of 0.72. The weighted kappa values for most of the soil classification methods can reach 0.5, which shows that there is a median degree of consistency in terms of the classified soil type. Hence, current soil classification methods need to be supplemented using local geological conditions and laboratory test results from borehole records. Therefore, a novel modified soil classification chart is presented herein to illustrate the deltaic soil conditions more accurately.

Systematic Review of the Literature on Various Soil Classification Methods

Systematic Review of the Literature on Various Soil Classification Methods

Analysis of the influence of tropical soil classification methods on railway subgrades according to repeated load triaxial (RLT) and light weight deflectometer (LWD) tests

Recently, non-destructive in-situ tests have emerged that can predict the resilient modulus (MR) of materials that make up railway sidewalk layers in an agile and low-cost manner. These are in contrast to laboratory tests that require more infrastructure and are more time-consuming to perform. However, few studies have been directed to the resilient behavior compared to in-situ and laboratory tests for soils classified as tropical. This paper aims to determine the relationship between the estimated modulus of elasticity (ELWD) and the estimated resilient modulus (MR) from the repeated load triaxial test (RLT) for use in tropical soil railway subsoils. The samples of subgrade soils belong to a railway stretch in Brazil, present in the states of São Paulo-SP and Minas Gerais-MG, which has large circulation of minerals and commodities. The soils are classified as lateritic and non-lateritic according to the Brazilian Miniature, Compacted, Tropical (MCT) methodology, in addition to grain size, compaction, LWD and RLT tests. It is noted that the modulus of elasticity (ELWD) tends to decrease according to the increase in strain obtained in LWD. For soils classified as lateritic, as the soil moisture increases the ratio between the moduli obtained by RLT and LWD tend to increase. However, non-lateritic soils have the opposite performance, i.e., the ratio between the modulus of the tests decreases with increasing moisture. In this study, the LWD test is a possible viable alternative with good accuracy and effectiveness for geotechnical investigations in tropical soils.

The Influence of Dispersing Agents on Soil Particle Size Analysis

Abstract This article explores soil classification methods and the influence of using dispersing agents in various amounts for the soil particle size analysis. Laboratory testing with different dispersing agents was conducted to analyze the effects of sodium hexametaphosphate and lithium carbonate for two different soils with plasticity indices indicating one to be with low plasticity (PI = 8.23 %) and the other one with high plasticity (PI = 39 %). The article highlights differences in the granulometric fractions reported for each soil and usage of the same dispersing agent but in various amount. Thus, the classification of soils is influenced by the amount of the used dispersing agents, recommendations of standards being different in time and countries. These findings contribute to a better understanding of the anticipated behavior of fine soils and provide useful insights for various geotechnical applications.

A Practical Soil Classification Method (Pscm) Based on Grain Size Distribution and Consistency of Soils for Geotechnical Applications

A Practical Soil Classification Method (Pscm) Based on Grain Size Distribution and Consistency of Soils for Geotechnical Applications

Use of Cone Penetration Test for Classification of Some Sudanese Soils; the State of the Art

The cone penetration test (CPT) method has successfully been applied in Sudan for research, foundation design and the solution of geotechnical problems for various engineering structures since 1977. Based on analyses of database accumulated during four decades correlation relationships have been developed to classify local soils from CPT data. This paper presents the state of the art and the main findings of previous studies carried out on the application of CPT for soil classification. A simple and reliable CPT method based on the statistical “discriminant analysis” technique was proposed in 1980 to classify a given soil of known CPT parameters into one of four main USCS soil groups. Another correlation was proposed in a different study between CPT data andrelative consistency of highly plastic clay soils. The original soil classification method was further modified in 2003 to include more database and incorporate additional features relating to the consistency of clay soils and relative density of sandy soils. The accuracy of the CPT based soil classification method was examined amongst worldwide known methods by comparison of the soil types predicted from CPT and the actual soil profiles. The comparison study showed that the method performed quite well in classifying alluvial soils which are similar to those used in their development.

Soil Classification: A New Approach for Grouping Soils Using Unsaturated Hydraulic Conductivity Data

AbstractGrouping soils based on similarities in their textural, taxonomic, and/or structural properties has broad applications to pedology, hydrology, and soil science. In this study, we present a new approach for classifying soils using hydraulic conductivity data. We apply concepts from critical path analysis and calculate critical pore sizes at various water saturations from the unsaturated hydraulic conductivity curves. Soils with similar critical pore size at the same effective water saturation are then grouped into the same class. To demonstrate the practical application of the proposed soil classification method, we use 102 samples including nine soil textures from the UNSODA database. Applying a curve clustering method, we find eight different soil classes within the studied data set.

17世紀の会津地方における土壌の分類評価—『会津農書』と土地帳の比較—

17世紀の会津地方における土壌の分類評価—『会津農書』と土地帳の比較—

Liquefaction resistance behaviours of gravel steel slag

Due to the continuous scarcity of sand and gravel under environmental protection policies, a new geo-backfill material with the good liquefaction resistance is urgently needed. Steel slag is a kind of industrial solid waste. The aged eight-month steel slag usually could be used as backfill material. It is important to investigate the liquefaction resistance behaviours of steel slag. According to the soil classification method, the waste steel slag is also divided into gravel steel slag, coarse steel slag and fine steel slag. Dynamic triaxial tests were carried out to investigate the liquefaction resistance of gravel steel slag in the paper by considering the effects of consolidation stress ratio, loading frequency, confining pressure and gravel-size particle contents. The variations of pore water pressure and cyclic shear resistance were analysed. The calculation model of pore water pressure of saturated sand proposed by Seed and Finn could be applied to analyse the dynamic pore water pressure development of gravel steel slag. Compared with conventional soils, the cyclic liquefaction resistance of gravel steel slag is better than that of fine sand, gravel soil, sand-gravel soils and sand pebbles. Gravel steel slag can be used to replace some conventional cohesionless soils as a geo-backfill material.

Comparison of soil classification methods among domestic and oversea based on CPTU test

This article presents the comparison of two soil classification methods among domestic and oversea based on CPTU test in geotechnical practice of soft deposit subgrade. Although CPTU test exists earlier in the world, However, CPTU test has been not applied in geotechnical practice of road subgrade until in recent years in China. Two soil classification methods among domestic and oversea are applied to soil classification and differences of them are analyzed.