Abstract:A regression model was established through orthogonal experimental design and multiple li-near regression methods to analyze the sensitivity of different support stiffness influencing factors and their advantages and disadvantages in terms of supporting stiffness. A first-order regression equation regarding the supporting stiffness was obtained. By adjusting the supporting parameters, the tunnel supporting stiffness was calculated and optimized. Regarding the influence of sprayed concrete thickness, sprayed concrete strength, long anchor rod length, anchor rod circumferential spacing, anchor rod longitudinal spacing, and steel frame longitudinal spacing on supporting stiffness, a first-order orthogonal regression experiment was conducted using the L25(56) orthogonal table, and the constant values of each parameter and the partial correlation coefficients were obtained. A multiple linear regression equation was constructed. Through range analysis and variance analysis, the test results were analyzed using SPSS software, and the remaining influencing factors of supporting stiffness were substituted into the regression equation for verification and analysis. The research results show that the prediction accuracy of this multiple linear regression analysis model is high, with an average relative error of 1.14%. The significant parameters F values of factors such as sprayed concrete strength and sprayed concrete thickness level are greater than F0.001, which play a decisive role in supporting stiffness; the F value of the longitudinal spacing of the steel arch frame is greater than F0.025, and its influence on supporting stiffness is relatively significant; the F values of the circumferential spacing and longitudinal spacing of the anchor rod are greater than F0.1, and these two factors have a generally significant impact on supporting stiffness; the length of the long anchor rod has no significant influence on the value of supporting stiffness.