基于人工蜂群算法的岩土边坡稳定性测度分析

传统算法存在对初值选择敏感或者容易陷入局部最优的弊端，求解准确性较差。为此提出一种新的基于人工蜂群算法的岩土边坡稳定性测度分析方法。针对某研究岩土边坡，将坡脚看作坐标原点，构造直角坐标系，计算边坡稳定性系数。针对若干圆弧滑动面，构造优化数学模型，将其看作适应度函数，通过人工蜂群算法对其进行求解。在寻找最优解时，形成含若干解的初始种群，不同蜂群首先对解进行一次邻域搜寻，把新得到的解和之前解比较，保留适应度更高的解。人工蜂群算法随机因子多，在寻优后期，收敛速度过快，影响收敛准确性。为此，引入细菌趋化思想对其进行改进，在蜂群密度较小的情况下，蜂群开始吸引操作，在种群密度较大的情况下，蜂群开始排斥操作，引入自适应步长，增强蜂群全局搜索能力。经实验验证，所提算法可搜索获取全局最优解，有效完成岩土边坡稳定性测度分析。

Measurement and analysis of rock slope stability based on artificial bee colony algorithm

traditional algorithms are sensitive to initial value selection or easy to fall into local optimum, and the accuracy of solution is poor. Therefore, a new method based on artificial bee colony algorithm is proposed to measure the stability of rock slope. In view of a study of rock slope, the slope toe is regarded as the origin of coordinates, and a rectangular coordinate system is constructed to calculate the slope stability coefficient. For some circular sliding surfaces, an optimization mathematical model is constructed. It is regarded as fitness function and solved by artificial bee colony algorithm. In finding the optimal solution, the initial population with some solutions is formed, and the different bees first search the solution in a neighborhood, and compare the new solution and the previous solution, so as to retain the higher fitness of the solution. Artificial bee colony algorithm has many random factors. In the later stage of optimization, the convergence rate is too fast, which affects the accuracy of convergence. To this end, the idea of bacterial chemotaxis was introduced to improve it. In the case of small colony density, the colony began to attract the operation. In the case of large population density, the colony began to repel operation, and the adaptive step length was introduced to enhance the global search ability of the colony. Experimental results show that the proposed algorithm can search the global optimal solution and effectively measure the stability of rock slope.