土钉墙内部稳定性分析自适应禁忌变异遗传算法

针对土钉墙内部稳定性分析问题，提出一种能同时确定土钉墙任意形状最危险滑动面及相应安全系数的全局优化算法。首先，基于边坡极限平衡法中的Morgenstern-Price法，考虑土钉的加固作用，推导土钉墙安全系数计算公式。然后，分析简单遗传算法（SGA）和禁忌搜索算法（TSA）的不足，引进自适应遗传算法（AGA），用禁忌变异算子替换自适应遗传算法中的标准变异算子，提出一种新型的混合优化算法--自适应禁忌变异遗传搜索算法（ATMGA），并将其应用到土钉墙内部稳定性分析领域，建立土钉墙任意形状临界滑裂面稳定性分析优化算法。算例分析表明，（1）与单纯的SGA、TSA以及AGA相比，文中ATMGA法搜索效率高、收敛速度快且能准确地搜索到土钉墙最危险滑动面及其相应的最小安全系数；（2）土钉支护前、后临界滑面位置相差较大，建议在实际工程中对支护后基坑的临界滑面以及相应的安全系数进行重新搜索。

Application of ATMGA to interior stability analysis of soil-nailing wall

According to the interior stability problem of the soil-nailing wall, a global optimal algorithm, which can locate the critical arbitrary slip surface associated with the minimum safety factor, is proposed. Firstly, based on Morgenstern-Price limited equilibrium method and by considering the consolidation of soil nail, a new computation formula of stability safety factor for the soil-nailing wall is presented. Then, the disadvantages of the simple genetic algorithm (SGA) and the tabu search algorithm (TSA) are following analyzed. Furthermore, the adaptive genetic algorithm (AGA) is also introduced. Through substituting the tabu mutation operator (TM) for the mutation operator of AGA, a new type of optimal algorithm, i.e. the adaptive tabu mutation operator and genetic algorithm (ATMGA) is proposed. Applying the ATMGA into the field of soil-nailing wall interior stability analysis, an optimal algorithm for searching the arbitrary critical slip surface is established. The simulating calculation results show that: (1) compared with SGA, TSA, and AGA, the ATMGA not only has the merits of high searching efficiency and fast converging velocity but also can exactly seek out the critical slip surface and the corresponding global minimal safety factor. (2) There is a significant difference between the critical slip surface of the unconsolidated wall and the soil-nailing wall, it is also suggested that it is necessary to seek out the critical slip surface and its safe factory of the soil-nailing wall in practical foundation pit engineering.