Abstract: The search of non-circular critical slip surface is a key problem in slope stability analysis. Essentially, this problem is a path search problem; so the ant colony algorithm is a very suitable method to solve this problem. To overcome the shortcomings of traditional ant colony algorithm, from the principle of combing one whole road by forward and reverse search of ants, here a new algorithm, meeting ant colony algorithm, is proposed. And then this meeting ant colony algorithm is used to locate the critical slip surface of slopes. Through two typical examples and an engineering application of a reservoir bank slope, the effectiveness of the new algorithm is verified. The results show that, from not only the whole search scope but also one point search scope, the meeting ant colony algorithm are all better than traditional ant colony algorithm; so the solutions diversity to search the critical slip surface of slopes by meeting ant colony algorithm is better. And then, to meeting ant colony algorithm, the global optimal solution can be included in the search scope by the larger probability; and the global optimal solution can be obtained by the larger probability. At last, by meeting ant colony algorithm, the critical slip surface of slopes can be found on larger scope and by larger speed.

Key words: slope, non-circular critical slip surface, search, meeting ant colony algorithm