土质边坡失稳判别的CA—AA耦合模型

Evaluation of Soil Slope Stability Based on Cellular Automata Ant Algorithm

解决边坡稳定性问题的关键是确定最危险滑动面(潜在最危险滑动面)和边坡的稳定系数。将蚁群算法用于求解土质的边坡稳定性问题。针对蚁群算法在演化过程中存在停滞和过早收敛的现象,对基本蚁群算法的结构形式和蚂蚁转移概率的计算进行了改进。构建了一种新的蚁群算法——元胞蚁群算法(CA-AA)。在文武坡边坡稳定性判别中,利用CA-AA法计算得到其稳定系数为1.37,略小于《岩土工程勘察规范》不平衡推力法计算值,弥补了不平衡推力法的不足,更接近实际。算例结果表明:元胞蚁群算法(CA-AA)总能搜索到问题的全局最优解,且搜索效率也有较大的提高。

How to ascertain the sliding surface and the factor of safety is the key problem of slope stability, the method of ant colony algorithm（CA）fitting was used to solve the stability analysis of soil slopes, and the structure of CA and the transfer probability formula were modified. An adaptive operator was introduced to change the selection mechanism of ants and prevent the phenomena of stagnation and premature in the evolutionary process. Thus, a new ant colony algorithm, called adaptive ant colony algorithm （CA-AA）, was presented. The stability coefficient of Wenwu slope is 1. 37, which is less than the value of imbalance push method recommended by ＂The Rock Soil Engineering Reconnaissance Specification＂, and its defect is overcomed, and the new method applied to soil slope stability evaluation shows that it can exactly find out the global optimal solution and the search efficiency is higher.