Intelligent displacement back analysis for excavation of an underground powerhouse in China

Abstract

Back analysis is an effective method to obtain the rock mass mechanical parameters with measured displacements. But the traditional back analysis methods have some shortcomings, such as narrow scope of application and instability. The intelligent back analysis method which incorporates a neural network and a genetic algorithm can overcome the drawbacks mentioned above and give satisfactory results. In this paper, based on orthogonal design, neural network and genetic algorithms, the intelligent displacement back analysis was carried out for the excavation of an underground powerhouse of a pumped storage power station in China. First, a series of samples were selected to train the neural network so that the relations between displacement of rock mass and parameters were erected. Then the optimum values of parameters were gotten taking advantage of optimization of genetic algorithms. Substituting the obtained parameters into FDM software for forward computation, it was found that the calculated displacements agreed the measured data well. The intelligent back analysis method can be used as a powerful tool to find out the optimum mechanical parameters of rock mass.