Abstract: Genetic expression programming (GEP) is a new method of simulating the evolutional process of biology. It makes encoding more convenient and simple by combining the advantages of genetic algorithm and genetic programming. In this article, a mixed GEP is presented by introducing simplex algorithm and backtracking mechanism into GEP to improve local searching capability and avoid precociousness. On this basis, a new method which integrates the mixed GEP algorithm and Imbalanced Thrust Force Method and takes safety factor as objective function is proposed to search for the non-circular critical slip surface. The results of two classical examples show that the proposed method could exactly find the non-circular critical slip surface and safety factor of slope. Moreover, the mixed GEP is superior to standard GEP in terms of local searching precision and global searching ability as well as convergence rate.

Key words: slope, critical slip surface, GEP, backtracking mechanism, simplex algorithm