Back-Analysis of Stress Field for Underground Powerhouse Using Difference Evolutionary Self-Adaptive BP

Parameter back-analysis of the initial stress state of underground rock is studied by the differential evolution (DE) algorithm and artificial neural network (ANN). Based on the observed deformation of underground workshop, a three-dimensional numerical model for back-analysis in the initial stress field of underground rock is put forward in the case study of an engineering project. In the process of analyzing, orthogonal experimental design and self-adaptive back propagation (BP) neural networks are used to back-analyze differential directions deformation of rock, and the initial stress field distribution of rock is accessed. It shows that the calculation results agree well with the measured values, and the calculated 3-D stress field in this engineering case is reasonable.