The performance function of deep tunnel in the reliability analysis can not be explicitly expressed and has non-linear characteristics, a new method of reliability analysis is proposed by combining intelligent methods and Monte Carlo principle. The support vector machine(SVM) model is employed to establish the nonlinear mapping relationship between basic random variables and structural response action, particle swarm optimization is used to optimize the parameters of SVM model. The learned SVM model is integrated with Monte Carlo method to calculate the failure probability and reliability. At last, the validity of the proposed method is verified by two typical engineering examples, the reliability index of wall rock along Jinping Hydropower Station diversion tunnels is calculated，the results are in good agreement with the actual situations. The proposed method could benefits other similar projects.