Abstract:

The support vector machine (SVM)is applied to reliability analysis of underground engineering. A new method of tunnel reliability analysis is proposed by combining the SVM, numerical simulation, first order second moment (FORM) and Monte Carlo simulation. The learning samples are built by numerical simulation; then the relationship between displacement of tunnel and the random variables is built. The performance function and its derivatives in tunnel reliability analysis were approximated by SVM. The statistical moments calculated from the performance function values and the corresponding gradients using SVM were then used in the calculation of the reliability index in tunnel reliability analysis. An example is given for illustrating the applicability of the proposed approach. The new method is effective in tunnel reliability analysis. And it has potential application to other reliability problems of complicated engineering structure with a considerably large number of random variables.

Key words: underground engineering, reliability analysis, support vector machine, Monte Carlo simulation