基于PSO-SVM算法联合数值模拟的红岭铅锌矿地表移动带圈定.doc

基金项目：国家重点研发计划资助项目(2016YFC0600706); 国家自然科学基金重点项目 This work was supported by the State Key Research Development Program of China((2016YFC0600706) and the Key Projects of National Natural Science Foundation of China. 作者简介：彭定潇（1994—），男，江西上饶人，采矿工程专业，硕士研究生，从事采矿工程与数值模拟研究。pdxfamily@163.com 基于PSO-SVM算法联合数值模拟的红岭铅锌矿地表移动带圈定 彭定潇，李夕兵 （中南大学 资源与安全工程学院，湖南 长沙 410083） 摘要：为了更加准确圈定红岭铅锌矿地表移动带，依据PSO-SVM算法与数值模拟联合确定红岭铅锌矿各矿体移动角。统计国内外30个崩落法矿山的地质特征数据，选取矿体上下盘围岩性质、上下盘围岩稳固程度、矿体倾角、矿体厚度以及开采深度共五种因素作为模型的输入参数，矿体上下盘移动角作为输出参数，基于PSO优化算法建立崩落法矿山移动角预测模型。采用前处理软件 HyperMesh对矿区进行高精度建模，并导入有限差分法软件FLAC3D之中，利用数值模拟分析了地表移动变形。结果表明：利用PSO-SVM算法与数值模拟得出的移动角基本吻合，相互验证。最终确定红岭铅锌矿1、1-1、2号矿体的上下盘移动角，由此得出红岭铅锌矿地表移动带范围。为圈定金属矿山地表移动带提供了一种可靠的方法。 关键词：地下金属矿山；地表移动带；支持向量机；数值模拟 中图分类号：TD325 文献标志码：A Delineation of Surface Displacement Belt for Hongling Lead-Zinc Mine Based on PSO-SVM Algorithm Combine with Numerical Simulation PENG Ding-xiao, LI Xi-bing (School of Resources and Safety Engineering，Central South University，Changsha，Hunan 410083，China) Abstract: In order to obtain more accurate delineation of surface displacement belt for Hongling lead-zinc mine, the motion angles were calculated by the PSO-SVM algorithm combine with the results of numerical simulation. We collected the mining surface movement data of 30 caving mine in china and abroad, and selected a total of five factors as the input variable which was Pu’s coefficient of the hanging wall and foot wall of ore body, stability of surrounding rock, deposit angle, mining thickness, mining depth, and the motion angel of terrane were be selected as the output variable to build the SVM model for the prediction of the motion angel in caving mine based on the particle swarm optimization method. The 3D geological model was established accurately by the pre-processing software HyperMesh, and then the model was imported to a commercial code (FLAC3D), surface movement and deformation were numerically analyzed. The results shows that the motion angles obtained by PSO-