摘要
为有效掌握三峡库区典型滑坡的成因机制及其发展趋势,该文以老蛇窝滑坡为背景,结合其区域地质条件,开展成因机制分析;并利用卷积神经网络及长短时记忆神经网络构建变形预测模型,评价其发展趋势。分析表明:老蛇窝滑坡属典型堆积层滑坡,成因较为复杂,是在固有因素的前提下,外部因素诱发产生的;同时,通过变形预测研究,得出其预测结果的最大平均相对误差为1.72%,具有较高的预测精度,且外推预测认为该滑坡变形将持续增加,发展趋势趋于不稳定,需采取适当的防治措施。
In order to effectively grasp the genetic mechanism and development trend of typical landslides in the Three Gorges Reservoir area, this paper takes the Laoshewo Landslide as background. The genetic mechanism is analyzed by combination with its regional geological conditions;And the convolution neural network and long short term memory neural network are used to build the deformation prediction model, so as to evaluate the development trend. The analysis shows that the Laoshewo Landslide is a typical accumulated layer landslide, and its origin is relatively complex, which is induced by external factors on the premise of its inherent factors;Meanwhile, through the deformation prediction research, the maximum average relative error of the prediction result is 1.72%, which has a high prediction accuracy.And the extrapolation prediction result considers the deformation of this slop will continue to increase.Its development trend tends to be unstable, and the appropriate prevention and control measures should be taken.
作者
王辉
黄鑫
张静
柴慧鹏
郭贵强
Wang Hui;Huang Xin;Zhang Jing;Chai Huipeng;Guo Guiqiang(Hydrogeological and Geothermal Geological Key Laboratory of Qinghai Province(Hydro Geology and EngineeringGeology and Enviromental Geology Survey Institute of Qinghai Province))
出处
《勘察科学技术》
2020年第3期22-27,共6页
Site Investigation Science and Technology
关键词
三峡库区
成因机制
卷积神经网络
长短时记忆神经网络
发展趋势
Three Gorges Reservoir area
genetic mechanism
convolution neural network
long short term memory neural network
development trend
