Locating the critical failure surface in a slope stability analysis by genetic algorithm

The slope stability analysis is routinely performed by engineers to evaluate the stability of embankment dams, road embankments, river training works, excavations and retaining walls. Locating the critical failure surface of a soil slope is rendered erroneous and cumbersome due to the existence of local minima points. In case of large soil slopes, engineers face with a search space too large to employ the trial and error method in a computationally efficient fashion. A genetic algorithm is proposed to locate the critical surface under general conditions with general constraints. Convergence to any prescribed degree of precision was achieved with the algorithm. The algorithm has been demonstrated to be computationally superior to other optimization routines, like, Monte-Carlo method and grid-points approach. (C) 2008 Elsevier B.V. All rights reserved.