A Kriging-based hybrid optimization algorithm for slope reliability analysis

First, the artificial bee colony (ABC) algorithm was used to optimize the Kriging model. A typical example was given to validate a good function fitting performance of the ABC-optimized Kriging model. Next, a structural reliability optimization method was proposed on the basis of the Kriging surrogate model, combined with the artificial bee colony optimization method and penalty function method. A nonlinear limit-state function example was provided to validate the accuracy and efficiency of the hybrid algorithm by comparing the algorithm’s solution with the exact solution of the example. Finally, the proposed method was applied to earth slope reliability analysis using a commercial finite difference program to calculate the slope’s factor of safety. Comparison of the results between the proposed method and the other methods shows that the proposed algorithm is the most suitable for slope reliability analysis.