Stability Assessment of Silin Powerhouse Cavern Based on Displacement Back Analysis Method

Abstract

A back analysis procedure synthetically using parameters sensitivity analysis based on orthogonal design variance analysis, support vector regression modeling of nonlinear relation between mechanical parameters and displacement, and a hybrid optimization algorithm of particle swarm optimizer (PSO) and differential evolution (DE), was proposed. Optimum values of elasto-plastic mechanic parameters of rock mass were identified using the back analysis procedure and the measured displacement data. Assigning to the numerical model the optimum values of parameters, numerical simulation of powerhouse construction process was carried out to investigate the reliability of result of back analysis and the stability of powerhouse cavern. The result shows that calculated displacements are in good agreement with the measured, the powerhouse cavern is stable and has a considerable safety margin, and the existing support system can satisfy the need of local stability of powerhouse.