Advance soil and Deep learning.pptx

Advance soil and Deep learning.pptx

• 1.
  Soil mechanices The primaryobjective of this research is to develop an appropriate connection for the effective stress parameter via the use of a potent machine learning strategy known as Multi-Gene Genetic Programming (MGGP), which is a subfield of artificial intelligence (AI).
• 2.
  Introduction • The capacityof the MGGP method to forecast equations without the need for any previous assumptions about the nature of the sought-after connection is the primary benefit of the approach when compared to regression and other forms of soft computing. Independent variables in this investigation include soil suction, residual volumetric water content, saturated volumetric water content, air entry pressure (bubbling pressure), and saturated volumetric water content. Additionally, residual volumetric water content, saturated volumetric water content, and net stress are als
• 3.
  Introduction • where isthe so-called effective stress (χ) parameter developed by Bishop and advised to be equal to the degree of saturation, P stands for net stress, and S indicates suction. Bishop stated that this parameter should be equal to the degree of saturation. This approximation, on the other hand, has not been proven to be highly suitable and accurate by his successors who have done the study. It is important to highlight that one of the primary focuses of researchers is still on developing an appropriate equation that can reliably measure the value of the effective stress parameter in unsaturated soils.
• 4.
  Methodology • In orderto perform multi-gene genetic programming for the purpose of finding an appropriate formula for the effective stress parameter, the GPtips 2.0 software and programming in the MATLAB environment were used. Each and every one of the SWRC parameters, as well as any combination of these parameters, may be taken into consideration for use as a potential input to the model. As a result, there were a total of six characteristics that were taken into consideration as model input variables. These parameters were net stress, suction, SWRC slope, air entry pressure, residual, and saturated volumetric water content. In order to use these variables as input variables for the model, they were first transformed into dimensionless quantities, which are represented by the equation (2). The effective stress parameter was the output variable in this equation.
• 5.
  Methodology • For thepurpose of creating the model in this investigation, a total of 101 pieces of data were employed. Following an examination of earlier research and a search for studies with datasets that included all of the target variables that were necessary for this investigation, data points were gathered from the relevant literature. These data were retrieved based on the outcomes of triaxial shear tests, from which the effective stress parameter values were obtained; pressure plate studies; and filter paper experiments; all of which were conducted in order to define the parameters of the soil water retention curve [1–10].
• 6.
  Methodology • The datasetconsists of seven unsaturated soil properties, namely, suction (S), air entry suction (hb), net stress (P), residual water content (θr), volumetric water content at saturation (θs), soil water retention curve slope (λ), and the effective stress • arameter (χ). Input variables were made dimensionless/normalized as: θ /θ , P/h and h /S and used together with λ. The dataset was divided into two separate groups: an 80% training dataset and the remaining data were selected for testing the accuracy and effectiveness of the model.
• 7.
  • that isshown below was taken into consideration. The value of function f was determined using multi-gene genetic programming.
• 8.
  that is shownbelow was taken into consideration. The value of function f was determined using multi-gene genetic programming.
• 12.
  Conclusion • This studyused a multi-gene genetic programming methodology to derive an equation for the effective stress parameter in unsaturated soils. The generated equation exhibited a satisfactory level of accuracy when compared to a dataset consisting of 101 test points obtained from laboratory experiments. Additionally, it demonstrated favourable performance in comparison to other existing connections documented in the academic literature. The conducted parametric analysis of the suggested equation revealed that both air entry suction and the net stress ratio have a substantial impact on variations in the effective stress parameter in relation to suction. The influence of the slope of the water retention curve and the ratio of residual to saturated volumetric water content on the fluctuation of the effective stress parameter with suction was found to be comparatively less significant than the other factors stated before.
• 13.
  References • N. Khalili,F. Geiser, G. Blight, Effective stress in unsaturated soils: Review with new evidence, International journal of Geomechanics, 4(2) (2004) 115- 126. • A. Russell, N. Khalili, A unified bounding surface plasticity model for unsaturated soils, International Journal for Numerical and Analytical Methods in Geomechanics, 30(3) (2006) 181-212 •L. Miao, S. Liu, Y. Lai, Research of soil–water characteristics and shear strength features of Nanyang expansive soil, Engineering Geology, 65(4) (2002) 261- • 267 • C. Rampino, C. Mancuso, F. Vinale, Experimental behaviour and modelling of an unsaturated compacted soil, Canadian Geotechnical Journal, 37(4) (2000) 748-