Abstract

Uncertainties of geotechnical parameters in landslide engineering mainly arise from their spatial variabilities, measuring error and inaccurate constitutive model. Interval limit equilibrium method (ILEM) is used to analyse the stability of landslides with uncertain geotechnical parameters. However, the method suffers from the issue of overestimation in the calculation of the safety factor (FS) interval while using the basic interval operations. To overcome the limitation, an improved ILEM based on particle swarm optimisation (ILEM-PSO) is proposed in the article. In the method, the uncertain parameters and their intervals are taken as the search variables and search space of PSO, and the FS functions are used as the fitness functions. A predefined number of iterations are executed to search the minimal and maximal FS, which are the lower and upper bound of the FS interval. The proposed method is validated with a testing function and ACADS benchmark question EX11, and then applied to the stability analysis of Dahua landslide. Comparisons with ILEM and Monte Carlo simulation are also given in the article. Results show that the proposed method is effective in solving the issue of overestimation in ILEM and obtains accurate FS intervals of landslides.

摘要

滑坡工程中岩土参数的不确定性主要来自其空间变异性、测量误差和不准确的本构模型。区间极限平衡法（ILEM）用于分析岩土参数不确定的滑坡稳定性。然而，该方法在使用基本区间操作时存在安全系数 (FS) 区间计算中高估的问题。为了克服这一局限性，本文提出了一种基于粒子群优化的改进ILEM（ILEM-PSO）。该方法以不确定参数及其区间作为PSO的搜索变量和搜索空间，以FS函数作为适应度函数。执行预定义次数的迭代以搜索最小和最大 FS，这是 FS 间隔的下限和上限。所提出的方法通过测试函数和ACADS基准问题EX11进行验证，然后应用于大化滑坡的稳定性分析。文中还给出了与 ILEM 和 Monte Carlo 仿真的比较。结果表明，所提出的方法有效解决了ILEM中的高估问题，并获得了准确的滑坡FS区间。