基于套索与优化算法的矿山岩质边坡可靠度研究  

Study on the Reliability of Mine Rock Slopes Based on Lasso and Optimization Algorithm

作　　者：郑阿鑫 ZHENG Axin(School of Civil and Surveying Engineering,Jiangxi University of Science and Technology,Ganzhou 341000,China)

机构地区：[1]江西理工大学土木与测绘工程学院,江西赣州341000

出　　处：《河南科技》2024年第5期43-46,共4页Henan Science and Technology

摘　　要：【目的】为解决拟合边坡极限状态方程时存在的过拟合与共线性等问题,且为获得更符合实际的安全系数,在考虑剪胀角的基础上,建立基于LASSO-GA-PSO的矿山岩质边坡可靠度分析模型。【方法】通过将Slide与FLAC3D相结合的方法,计算考虑剪胀角变化的边坡安全系数;采用LASSO对数据进行拟合,获得相应的极限状态方程;编写用遗传算法与LHS优化的自适应粒子群算法程序,建立基于LASSO-GA-PSO的边坡可靠度分析模型,进而确定边坡的可靠度指标。【结果】二次带交叉项的LASSO-GA-PSO模型所得的可靠度指标为2.3257,略小于LASSOMonte-Carlo的2.3879,误差为2.6%。【结论】模型具有较好的优化结果,可为类似工程提供参考。[Purposes]In order to solve the problems of overfitting and collinearity that exist in fitting the slope limit state equations,and to obtain more realistic safety factors,a mine rock slope reliability analysis model based on LASSO-GA-PSO is established based on the consideration of dilation angles.[Methods]By combining Slide with FLAC3D,the slope safety factor considering the change of dilation angle is calculated the data were fitted by LASSO to obtain the corresponding limit state equation;The adaptive particle swarm optimization algorithm program based on genetic algorithm and LHS optimization is compiled,and the slope reliability analysis model based on LASSO-GA-PSO is established to determine the reliability index of the slope.[Findings]The reliability index obtained from the LASSO-GA-PSO model of quadratic polynomial with cross terms is 2.3257,which is slightly smaller than that of 2.3879 of LASSO-Monte-Carlo,and the error is 2.6%.[Conclusions]The model has a good optimization result and can provide reference for similar projects.

关 键 词：矿山岩质边坡 剪胀角 FLAC3D LASSO算法 可靠度指标 

分 类 号：TD804[矿业工程—矿山开采]