摘要
针对当前露天矿山边坡变形难以准确获取的问题,建立了粒子群优化算法(PSO)及相关向量机(RVM)相结合的矿山边坡变形量预测模型。利用RVM对矿山边坡变形量非线性系统进行建模,借助PSO对RVM模型核参数寻优,构建基于PSO-RVM的矿山边坡变形预测模型。利用该模型对工程实例进行预测,并在相同学习样本下与MFOA-SVR模型进行对比。结果表明:PSO-RVM模型预测结果精度更高、离散性更小,在平均相对误差和均方根误差上远低于MFOA-SVR模型。PSO-RVM模型为准确预测矿山边坡变形量提供了一条新途径,对矿山开采工作具有一定的参考价值。
Aiming at the problem that it is difficult to accurately obtain slope deformation in open-pit mine,a prediction model of mine slope deformation combined with particle swarm optimization algorithm(PSO)and correlation vector machine(RVM)was established.The nonlinear system of mine slope deformation was modeled by using RVM and the kernel parameter of RVM model was optimized by using PSO.The prediction model of mine slope deformation was built based on PSO-RVM.The model is used to predict the engineering examples and compared with the MFOA-SVR model under the same learning sample.The results show that the prediction results of the PSO-RVM model are more accurate and less discrete,and are far lower than the MFOA-SVR model in terms of average relative error and root mean square error.The PSO-RVM model provides a new way to accurately predict the deformation of the mine slope,and has a certain reference value for the mining work.
作者
张研
范聪
吴哲康
刘晶
邝贺伟
ZHANG Yan;FAN Cong;WU Zhekang;LIU Jing;KUANG Hewei(Guangxi Key Laboratory of Geomechanics and Geotechnical Engineering,Guilin 541004,China;School of Civil and Architectural Engineering,Guilin University of Technology,Guilin 541004,China)
出处
《金属矿山》
CAS
北大核心
2022年第10期191-196,共6页
Metal Mine
基金
国家自然科学基金项目(编号:52068016)
广西自然科学基金项目(编号:2020GXNSFAA297118,2020GXNSFAA159125)
广西岩土力学与工程重点实验室项目(编号:桂科能20-Y-XT-01)。
关键词
粒子群优化算法
相关向量机
露天矿山
边坡变形
预测模型
sparticle swarm optimization algorithm
relevance vector machine
open pit mine
slope deformation
prediction model
