关键词: 刚性承压板试验, 变形模量, 反演分析, 数值模拟, PSO-LSSVM模型

Abstract: The aim of this study is to determine the deformation modulus under two circumstances: the distance from the edge of bearing plate to the sidewall of test tunnel and the distance from the center of bearing plate to the fulcrum of measuring support cannot meet the requirements of specification for rigid bearing plate test. The influence regularity of the two circumstances on test results was investigated first by using FLAC3D numerical simulation, and the learning samples of the relationship between deformation modulus and displacement were obtained. On this basis, an inverse analysis method for obtaining deformation modulus under non-standard conditions was developed based on the learning and prediction ability of PSO-LSSVM model. The inversion analysis method was applied to an engineering example. Results suggest that the distance from the edge of bearing plate to the sidewall of test tunnel and the distance from the center of bearing plate to the fulcrum of measuring support would both lead to deformation error, and the errors under the two circumstances would decrease at different rates with the increase of the distance. The deformation modulus of rock mass determined by the optimization inversion is GPa, with an average value of 2.30 GPa and a corresponding deformation error of 0.1%. The result is accurate and reasonable, and the proposed method provides a new idea for calculating deformation results of bearing plate test under special conditions.

Key words: rigid bearing plate test, deformation modulus, inversion analysis, numerical simulation, PSO-LSSVM model