矿山边坡地表变形的PSO-ELM预测模型  

PSO-ELM prediction model for surface deformation of mine slope

作　　者：宁永香 崔希民[2] NING Yongxiang;CUI Ximin(Department of Geology and Environment,Shanxi Institute of Technology,Yangquan 045000,China;College of Geoscience and Surveying Engineering,China University of Mining and 1echnology(Beijing),Beijing 100083,China)

机构地区：[1]山西工程技术学院地质与环境工程系,山西阳泉045000 [2]中国矿业大学(北京)地球科学与测绘工程学院,北京100083

出　　处：《煤田地质与勘探》2020年第6期201-206,216,共7页Coal Geology & Exploration

基　　金：阳泉市重点研发计划项目(2019G14)。

摘　　要：为提高矿山边坡地表变形预测模型的精度,从矿山边坡地表变形影响因素角度考虑,建立了基于粒子群优化(PSO)极限学习机(ELM)的矿山边坡地表变形预测模型。结合经典的粒子群优化算法和极限学习机方法,提出矿山边坡地表变形影响因素同地表变形数值之间的耦合关系;采用中煤平朔安家岭露天矿区矿山边坡地表变形及影响变形因素的采集数据,应用ELM建立预测模型,并应用PSO对ELM预测模型的输入层与隐含层的连接权值、隐含层阈值进行优化,以提高其预测精度。研究表明,经过PSO的优化,将预测模型的最大相对误差(4.705×10^-8)、均方误差(6.243×10^-5)及均方根误差(0.008)等预测误差参数分别降低到1.516×10^-8,1.158×10^-5和0.003,说明PSO-ELM预测模型具有更高的预测精度,该预测模型可在后续研究中进一步应用于矿山边坡地表变形预测中,以期提升矿山生产安全。In order to improve the model accuracy of slope surface deformation prediction data,the influence factors of surface deformation of mine slope was considered and the prediction model of the limit learning machine was established based on particle swarm optimization.Firstly,the mine slope surface deformation monitoring data and influencing factors data were used to establish the prediction model utilizing the classical particle swarm optimization algorithm and the limit learning machine method.Secondly,the surface deformation of the mine slope and its influencing factors were collected in Anjialing open-pit mining area.Particle swarm optimization(PSO) was applied to optimize the connection weight and threshold of the input layer and the hidden layer to improve the prediction accuracy of the model.Finally,through the optimization application of PSO,the maximum relative error(4.705×10^-8),mean square error(6.243×10^-5) and root-mean-square error(0.008) of the prediction model were reduced to 1.516×10^-8,1.158×10^-5 and 0.003 respectively.The experimental results showed that the proposed prediction model had higher prediction accuracy than other models,and it could be applied to the prediction of surface deformation of mine slope in the follow-up study,so as to improve the safety level of mine.

关 键 词：矿山边坡 变形监测 粒子群算法 极限学习机 预测模型 

分 类 号：TD325[矿业工程—矿井建设]