苗作华,杨凡,王永琦,严蔚涵,谭志兵,朱叶风

• 1:武汉科技大学资源与环境工程学院
• 2:冶金工业安全风险数智防控与应急联动重点实验室
• 3:易普力(新疆)矿山工程有限公司
• 4:宁波钢铁有限公司安全保卫部

摘要(Abstract)：

露天矿山台阶爆破后爆堆的松散系数是衡量爆破质量的重要指标，在明确了设计方量后，准确预测爆堆体积对于优化爆破参数、控制爆破效果及估算后续装运工程量具有重要意义。针对露天矿山爆破后爆堆体积测量效率低、小样本预测精度不足等问题，提出了一种基于机载激光雷达(Li DAR)与机器学习的智能化预测模型。首先采用轻量化无人机Li DAR系统，实现了爆堆点云数据的免像控快速采集;其次特征选择部分使用Deep Seek接口构建了融合Pearson相关性分析、ANOVA方差检验、随机森林特征重要性评估及递归特征消除的特征选择模型，从8项初始参数中筛选出6个关键影响因素。最后GS-KCV优化贝叶斯岭回归预测模型在测试集上R~2达到0.76，较传统方法提升显著，MAE(19 949.67 m~3)和RMSE(23 020.80 m~3)分别降低26%和28%，解决了现有模型在小样本场景下易陷局部最优的难题。通过大疆M350无人机与禅思L_2激光雷达的工程化应用验证，基于机载激光雷达与机器学习的智能化预测模型实现了爆破参数获取-体积预测-效果评估的全流程优化。

关键词(KeyWords)： 矿山爆破;贝叶斯岭回归;爆堆体积预测;机载激光雷达;特征选择;网格搜索

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金资助项目(41071242、41971237);; 教育部产学合作协同育人项目~~

作者(Author): 苗作华,杨凡,王永琦,严蔚涵,谭志兵,朱叶风

参考文献(References)：

• [1]于鑫鑫，张曌，马赛赛．基于无人机航测技术的露天矿爆堆特征感知方法应用[J]．露天采矿技术，2025,40(3):45-48.[1]YU Xin-xin,ZHANG Zhao,MA Sai-sai. Application of explosive pile characteristics perception method in open-pit mines based on UAV survey technology[J]. Opencast Mining Technology,2025,40(3):45-48.(in Chinese)
• [2]苗作华，谢媛，任磊，等．基于三维可视化的爆堆形态灰色敏感度分析[J]．金属矿山，2022(2):83-89.[2]MIAO Zuo-hua,XIE Yuan,REN lei,et al. Grey sensitivity analysis of detonation heap morphology based on 3D visualization[J]. Metal Mine,2022(2):83-89.(in Chinese)
• [3]吴江昊，姜涛．基于机载激光雷达的高陡边坡孤立危岩体识别方法[J]．经纬天地，2024(6):19-23,55.[3]WU Jiang-hao,JIANG Tao. Identification method on isolated dangerous rock mass in high-steep slope based on airborne Li DAR[J]. Survey World,2024(6):19-23,55.(in Chinese)
• [4]潘大伟，周雷，王世平，等．基于机载Li DAR和LANDIS PRO模型的矿山生态复垦区地上植被固碳能力及其提升路径研究[J/OL]．生态学杂志，2025:1-13.[2025-10-21]. https:∥link. cnki. net/urlid/21. 1148. Q.20241127. 1442. 004.[4]PAN Da-wei,ZHOU Lei,WANG Shi-ping,et al. Carbon sequestration capacity of aboveground vegetation in mine ecological reclamation area based on airborne Li DAR and LANDIS PRO model and its promotion path[J/OL]. Chinese Journal of Ecology,2025:1-13.[2025-10-21]. https:∥link. cnki. net/urlid/21. 1148. Q. 20241127. 1442.004.(in Chinese)
• [5]杨英杰．机载激光雷达在矿山治理中土石方量计算的应用[J]．世界有色金属，2024(22):207-209.[5]YANG Ying-jie. Application of airborne Li DAR in calculating earthwork volume in mine management[J]. World Nonferrous Metals,2024(22):207-209.(in Chinese)
• [6]夏淑媛，董永峰，王利琴．基于特征工程的XGBoost爆破块度预测研究[J]．爆破，2023,40(2):97-101,131.[6]XIA Shu-yuan,DONG Yong-feng,Wang Li-qin. Research on XGBoost burst block size prediction based on feature engineering[J]. Blasting,2023,40(2):97-101,131.(in Chinese)
• [7]刘欢，潘城．岩石爆破块度预测方法研究现状及展望[J]．建井技术，2025,46(3):73-81.[7]LIU Huan,PAN Cheng. Research status and prospects of rock blasting fragmentation prediction methods[J]. Mine Construction Technolgoy. 2025,46(3):73-81.(in Chinese)
• [8]张文涛，汪海波，高朋飞，等．基于PCA-WOA-XGBoost的露天矿山爆破振动峰值振速预测[J]．工程爆破，2024,30(6):155-167,177.[8]ZHANG Wen-tao,WANG Hai-bo,GAO Peng-fei,et al.Prediction of PPV of blasting vibration in open-pit mine based on PCA-WOA-XGBoost model[J]. Engineering Blasting,2024,30(6):155-167,177.(in Chinese)
• [9]刘骎，王文通，张千俊，等．基于WOA-SVM模型的爆破块度预测研究[J]．采矿技术，2024,24(2):103-107.[9]LIU Jun,WANG Wen-tong,ZHANG Qian-jun,et al. Research on blast fragmentation prediction based on WOASVM model[J]. Mining Technology,2024,24(2):103-107.(in Chinese)
• [10]黄晶柱，钟依禄，黄裘俊，等．基于高斯过程回归矿山爆破飞石距离预测模型[J]．工程爆破，2023,29(2):73-79,108.[10]HUANG Jin-zhu,ZHONG Yi-lu,HUANG Qiu-jun,et al.Prediction model of blasting flyrock distance in mine based on Gaussian process regression[J]. Engineering Blasting,2023,29(2):73-79,108.(in Chinese)
• [11]王纯杰，戚顺欣，张洪阳．Logistic回归模型参数的贝叶斯估计及应用[J]．统计与决策，2020,36(22):14-18.[11]WANG Chun-jie,QI Shun-xin,ZHANG Hong yang.Bayesian estimation and application for parameters in Logistic regression model[J]. Statistics&Decision,2020,36(22):14-18.(in Chinese)
• [12]费鸿禄，左壮壮，蒋安俊，等．基于KS-GS-SVR的峰值爆破振速预测[J]．工程爆破，2023,29(2):120-128.[12]FEI Hong-lu,ZUO Zhuang-zhuang,JIANG An-jun,et al.Peak blasting vibration velocity prediction based on KSGS-SVR[J]. Engineering Blasting,2023,29(2):120-128.(in Chinese)
• [13]赵颖，岳中文，薛克军，等．基于特征选择的GSKCV-XGBoost露天金属矿爆破块度预测模型[J]．工程爆破，2024,30(6):168-177.[13]ZHAO Yin,YUE Zhong-wen,XUE Ke-jun,et al. GSKCV-XGBoost open-pit metal mine blasting block size prediction model based on feature selection[J]. Engineering Blasting,2024,30(6):168-177.(in Chinese)
• [14]王麟．基于矿山地质测绘的无人机机载激光雷达技术的分析[J]．世界有色金属，2024(12):149-151.[14]WANG Lin. Analysis of unmanned aerial vehicle airborne Li DAR technology based on mining geological surveying and mapping[J]. World Nonferrous Metals,2024(12):149-151.(in Chinese)
• [15]刘新跃，胡科，杨文．机载激光雷达在地形测绘中的运用[J]．科技视界，2025,15(2):6-8.[15]LIU Xin-yue,HU Ke,YANG Wen. The application of airborne lidar in topographic mapping[J]. Technology Perspective,2025,15(2):6-8.(in Chinese)
• [16]闫魏力，张驰，王洛锋．无人机载三维激光扫描技术在露天矿山测量中的应用[J]．黄金，2022,43(8):41-44.[16]YAN Wei-li,ZHANG Chi,WANG Luo-feng. Application of UAV-based 3D laser scanning technology in open-pit mine surveying[J]. Gold,2022,43(8):41-44.(in Chinese)