Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Evaluation of Group Capacity of Micropile in Soft Clayey Soil from Experimental Analysis Using SVM-Based Prediction Model
Nowadays, micropiles, due to their capability to sustain high loads in a soft cohesive soil, are drawing attention from researchers, geotechnical engineers, and builders. However, the load-carrying capacity of a micropile group in soft cohesive soil is not readily available. This paper describes an investigation of support vector machine (SVM) regression for assessment of micropile group capacity in soft clayey soil considering the complex interaction between the soil and the micropiles and the uncertainty involved in the analysis. A total of 54 large-scale static vertical micropile load tests were conducted in a test pit, and from the load-settlement graphs plotted for these tests, a database of 376 data points was obtained that was used to develop the SVM model. The performance of the SVM model with three different kernel functions was evaluated The results of the SVM models were compared with those of artificial neural network models developed with three different types of algorithms. To determine the relative influence of the different input variables on the load-carrying capacity of micropile groups, a sensitivity analysis was also performed. An empirical equation was developed with the best-fit model for practical application. The developed equation was validated with a set of experimental data not used for generating the empirical equation.
Evaluation of Group Capacity of Micropile in Soft Clayey Soil from Experimental Analysis Using SVM-Based Prediction Model
Nowadays, micropiles, due to their capability to sustain high loads in a soft cohesive soil, are drawing attention from researchers, geotechnical engineers, and builders. However, the load-carrying capacity of a micropile group in soft cohesive soil is not readily available. This paper describes an investigation of support vector machine (SVM) regression for assessment of micropile group capacity in soft clayey soil considering the complex interaction between the soil and the micropiles and the uncertainty involved in the analysis. A total of 54 large-scale static vertical micropile load tests were conducted in a test pit, and from the load-settlement graphs plotted for these tests, a database of 376 data points was obtained that was used to develop the SVM model. The performance of the SVM model with three different kernel functions was evaluated The results of the SVM models were compared with those of artificial neural network models developed with three different types of algorithms. To determine the relative influence of the different input variables on the load-carrying capacity of micropile groups, a sensitivity analysis was also performed. An empirical equation was developed with the best-fit model for practical application. The developed equation was validated with a set of experimental data not used for generating the empirical equation.
Evaluation of Group Capacity of Micropile in Soft Clayey Soil from Experimental Analysis Using SVM-Based Prediction Model
Borthakur, Nirmali (Autor:in) / Dey, Ashim Kanti (Autor:in)
07.01.2020
Aufsatz (Zeitschrift)
Elektronische Ressource
Unbekannt
Predicting Micropile Group Capacity in Soft Cohesive Soil by Artificial Neural Network
| Springer Verlag | 2024
Prediction of uplift capacity of a micropile embedded in soil
| Springer Verlag | 2014
Prediction of uplift capacity of a micropile embedded in soil
| Online Contents | 2015
Bearing capacity improvement using micropile
| British Library Conference Proceedings | 2006