The purpose of this article is to provide an effective approach to evaluate slope stability in real-time in a reservoir area, which is significant for carrying out risk management for landslide disaster prevention in various engineering practices. A comprehensive idea for stability estimation of bank slope under the influence of rainfall or the reservoir water level is presented in this work. Slope stability analysis and back analysis of soil parameters are both included based on numerical simulation. The mechanical parameters of the bank slope were first back-analyzed using particle swarm optimization (PSO), and real-time stability analysis with high accuracy and efficiency was then established based on multiple continuously monitored displacements. Two case studies were carried out in this study. The results show that (1) based on the real-time monitored displacement and numerical simulation, the mechanical parameters of the slope can be reasonably retrieved through PSO; and (2) based on the inverse mechanical parameters, the safety factors of the slope can be numerically obtained, so that the real-time estimation of slope stability can be realized.

中文翻译：

---

基于粒子群优化的数值反分析实时估算边坡稳定性的有效方法

本文旨在提供一种实时评价库区边坡稳定性的有效方法，对在各种工程实践中开展滑坡灾害防治风险管理具有重要意义。本文提出了在降雨或水库水位影响下岸坡稳定性估算的综合思路。基于数值模拟，包括边坡稳定性分析和土参数反分析。首先使用粒子群优化 (PSO) 对岸坡的力学参数进行反分析，然后基于多个连续监测的位移建立高精度和高效的实时稳定性分析。本研究进行了两个案例研究。结果表明：（1）基于实时监测位移和数值模拟，可以通过粒子群算法合理反演边坡的力学参数；(2)基于逆向力学参数，可以数值化得到边坡的安全系数，从而实现边坡稳定性的实时估计。