摘要
以某土质边坡为研究对象,采用有限差分软件FLAC^(3D)对边坡进行确定性分析,用非线性映射能力较强的神经网络作为响应面函数,并编写Monte-Carlo法MATLAB程序,对边坡进行可靠度分析。结果表明:将神经网络作为响应面函数对于三维复杂边坡的可靠度求解能得到较好的精度,FLAC^(3D)内置的强度折减法可以很方便地实现安全系数的多次求解,计算效率较高。
Taking a soil slope as the research object, certainty analysis is made on the slope using the finite-difference software FLAC^3D. The neural network with strong non-linear projection ability is used as the response surface function and the MATLAB program is written by the Monte-Carlo method to analyze the reliability of the slope. The results show that the accuracy of three- dimensional complex slope reliability can be obtained by using the neural network as the response surface function and the strength reduction method of FLAC^3D can be used to solve the safety factor multiple times with high calculation efficiency.
出处
《化工矿物与加工》
北大核心
2017年第6期53-55,共3页
Industrial Minerals & Processing
关键词
神经网络响应面法
三维高陡边坡
可靠度分析
neural network response surface method
three-dimensional high-steep slope
reliability analysis