Three-dimensional inversion analysis of an in situ stress field based on a two-stage optimization algorithm

• Published in: Journal of Zhejiang University. A. Science Vol. 17; no. 10; pp. 782 - 802
• Main Authors: Zhang, She-rong, Hu, An-kui, Wang, Chao
• Format: Journal Article
• Language: English
• Published: Hangzhou Zhejiang University Press 01.10.2016; Springer Nature B.V
• Subjects: Algorithms; Artificial neural networks; cavern; Caverns; Civil Engineering; Classical and Continuum Physics; Compressive properties; DE; Difference; element,Huangdeng; Engineering; evolution; Finite; Geological faults; Geological structures; Geology; Hydroelectric power; Hydroelectric power stations; Industrial Chemistry/Chemical Engineering; Interlayers; machine; Mechanical Engineering; Numerical analysis; Optimization; Optimization algorithms; Parameters; regression; Regression models; Rocks; situ; SR; Stability analysis; Stepwise; Stress; Stress distribution; Support; Support vector machines; SVM; Tectonics; Three dimensional analysis; Tuff; underground; Underground caverns; Underground structures; vector; stress; Support vector machine (SVM); 支持向量机; 逐步回归; 差异进化; Stepwise regression (SR); 有限元; Huangdeng underground cavern; Difference evolution (DE); 初始地应力; Finite element; 黄登地下洞室
• ISSN: 1673-565X; 1862-1775
• DOI: 10.1631/jzus.A1600014

Establishing an accurate in situ stress field is important for analyzing the rock-mass stability of the undergroundcavern at the Huangdeng hydropower station in China. Because of the complexity and importance of the in situ stress field, ex-isting back analysis methods do not provide the necessary accuracy or sufficiently recognize nonlinear relations between thedistribution of the in situ stress field and its formative factors. Those factors are related to the geological structures of high com-pressive tectonic stress regimes, including geological faults and tuff interlayers. The new two-stage optimization algorithm pro-posed in this paper is a combination of stepwise regression （SR）, difference evolution （DE）, support vector machine （SVM）, andnumerical analysis techniques. Stepwise regression is used to find the set of unknown parameters that best match the modelingprediction and determine the range of parameters to be recognized. Difference evolution is used to determine the optimum pa-rameters of the SVM. The SVM is used to create the DE-SVM nonlinear reflection model to obtain the optimal values of theparameters from measured stress data. We compare the new two-stage optimization algorithm to other two popular methods, amultiple linear regression （MLR） analysis method and an artificial neural network （ANN） method, to estimate the in situ stressfield for the actual underground cavern at the Huangdeng hydropower station. The two-stage optimization algorithm produces amore realistic estimate of the stress distribution within the investigated area. Thus, this technique may have practical applica-tions in realistic scenarios requiring efficient and accurate estimations of the in situ stress in a rock-mass.