An Intelligent Optimization Back-Analysis Method for Geomechanical Parameters in Underground Engineering

Abstract

The geomechanical parameters in underground engineering are usually difficult to determine, which can pose great obstacles in underground engineering. A novel displacement back-analysis method is proposed to determine the geomechanical parameters in underground engineering. In this method, the problem of geomechanical parameter determination is converted into an optimization problem, regarding the geomechanical parameters as the optimization parameters, and the error between the calculated results and the field measurement information as the optimization objective function. The grasshopper optimization algorithm (GOA), which offers excellent global optimization performance, and the Gaussian process regression (GPR) machine learning, offering powerful fitting ability, are combined to address the time-consuming numerical calculations. Furthermore, the proposed method is combined with the 3D numerical calculation software FLAC3D to form the GOA-GPR-FLAC3D method, which can be used in the displacement back-analysis of geomechanical parameters in underground engineering. The results of a case study show that the proposed method can greatly improve computational efficiency while ensuring high precision compared with the GOA. When applied to the Tai’an Pumped Storage Power Station, this method can obtain more accurate results compared with the GOA under the same evaluation times and is more suitable for the back-analysis of rock parameters in underground engineering.