img

QQ群聊

img

官方微信

高级检索

黄金科学技术 ›› 2022, Vol. 30 ›› Issue (3): 392-403.doi: 10.11872/j.issn.1005-2518.2022.03.145

• 采选技术与矿山管理 • 上一篇    下一篇

基于MICE_RF的组合赋权—极限随机树岩爆预测模型

温廷新(),苏焕博()   

  1. 辽宁工程技术大学工商管理学院,辽宁 葫芦岛 125105
  • 收稿日期:2021-10-11 修回日期:2021-12-19 出版日期:2022-06-30 发布日期:2022-09-14
  • 通讯作者: 苏焕博 E-mail:wen_tx@163.com;260987894@qq.com
  • 作者简介:温廷新(1974-),男,山西太谷人,博士,教授,从事矿业系统工程、数据分析与智能决策研究工作。wen_tx@163.com
  • 基金资助:
    国家自然科学基金项目“基于数据挖掘的煤矿安全风险评价体系研究”(71371091)

Combined Weighting-Extremely Randomized Trees Rockburst Prediction Model Based on MICE_ RF

Tingxin WEN(),Huanbo SU()   

  1. School of Business Administration,Liaoning University of Engineering and Technology,Huludao 125105,Liaoning,China
  • Received:2021-10-11 Revised:2021-12-19 Online:2022-06-30 Published:2022-09-14
  • Contact: Huanbo SU E-mail:wen_tx@163.com;260987894@qq.com

摘要:

目前岩爆预测的真实训练数据量小、数据存在缺失,为了更加准确地预测岩爆等级,提出了一种基于链式随机森林多重插补(MICE_RF)算法的组合赋权—极限随机树(ET)预测模型。首先,在选取岩爆灾害主要评判指标的基础上,采用MICE_RF算法插补缺失数据;然后,由改进层次分析法(IAHP)和基于指标相关性的权重确定方法(CRITIC)确定指标主、客观权重,并引入权向量距离概念对指标组合赋权;最后,将插补和赋权后数据集采用ET算法,构建岩爆等级预测模型。利用国内外工程实例数据进行20次随机抽样试验,并与其他模型进行对比分析。结果表明:MICE_RF插补后可显著提高岩爆模型预测效果;改进AHP-CRITIC法较改进前更具优势,该模型平均预测准确率为93.10%,各比较指标结果均优于对比模型,预测结果更稳定。

关键词: 岩爆等级预测, 数据缺失, 链式随机森林的多重插补(MICE_RF)算法, 组合赋权, 权向量距离, 极限随机树(ET)算法

Abstract:

As a kind of dynamic instability geological disaster with strong abruptness and randomness,rockburst poses a great threat to the safety of personnel,equipment and buildings.Timely and accurate prediction of rockburst grade has become a hot issue in the field of underground engineering.At present,the amount of real training data of rockburst prediction is small and the data is missing.In order to predict the rockburst grade more accurately,a combined weighting-extremely randomized trees(ET) prediction model based on chain random forest multiple interpolation(MICE_RF) algorithm was proposed.According to the characteristics and causes of rockburst,six evaluation indexes including maximum shear stress,uniaxial compressive strength,uniaxial tensile strength,stress coefficient,brittleness coefficient and elastic energy index were selected to form the rockburst evaluation index,and MICE_RF algorithm was used to interpolate the missing data of rockburst data set.Then,a new combined weighting method was proposed,which is the improved analytic hierarchy process(IAHP)-weight determination method based on index correlation(CRITIC),and the weight of each index was comprehensively calculated by using the concept of weight vector distance. Finally,the ET algorithm was used to construct the rockburst prediction model after interpolation,weighting and normalization.Using the existing engineering example data at home and abroad,20 random sampling tests were carried out,and compared with other models to verify the superiority of this model in rockburst grade prediction.In this study,the interpolation effect based on MICE_RF missing value,the combined weighting effect of IAHP-CRITIC index and the comparison of the prediction effects of different models were analyzed and verified respectively.So,the ET rockburst prediction model based on MICE_RF and improved combined weighting was applied and the result of accuracy,precision,recall and RMSE were 93.10%,94.17%,93.44% and 0.2626.The results show that the MICE_RF missing data interpolation method not only increases the available rockburst data set,but also can effectively improve the prediction accuracy of three levels of no rockburst,intermediate rockburst and strong rockburst,and the average prediction accuracy of the complete data set has also been significantly improved.The improved AHP-CRITIC method has more advantages than the previous one,and the ET algorithm is significantly better than other comparison models in the results of four comparison indexes,that is,IAHP-CRITIC-ET model based on MICE_RF can significantly improve the prediction accuracy of rockburst grade,and the prediction results are more stable,which can provide effective guidance for similar projects.

Key words: rockburst grade prediction, missing data, multiple interpolation algorithm of chain random forest (MICE_RF), combination weighting, weight vector distance, extremely randomized trees(ET) algorithm

中图分类号: 

  • X936

图1

MICE_RF算法插补过程"

图2

基于ET算法的岩爆等级预测流程"

图3

岩爆等级预测模型的运行步骤和流程"

表1

原始岩爆数据集描述统计"

评判指标个案数缺失个数最小值最大值
X1/MPa161272.600167.200
X2/MPa1612718.320306.580
X3137510.38022.600
X418800.0525.263
X5164244.47980.000
X618800.85010.570
有效个案数137---

表2

不同插补方法的效果对比"

评估指标MeanSVMKNNRFMCMCMICE_RF
PACr0.81610.81980.82950.86660.87970.9114
RMSE15.744116.102316.791913.522012.778910.9850
DACK-S statistic0.10950.09850.05840.05840.03640.0329
K-S p-value0.45200.52840.86180.91000.99860.9999

表3

部分岩爆样本数据集"

序号岩爆指标实际类别
X1X2X3X4X5X6
14.6020.003.000.2306.6671.391
27.5052.003.700.14414.0541.301
????????
29143.40136.507.200.31818.9585.603
29291.30225.6017.200.40513.1167.303

表4

最佳主观权重计算过程"

判断矩阵C1C2C3C4
X1权重0.30620.27150.29490.3604
X2权重0.03780.03850.04120.0440
X3权重0.02790.02850.03850.0342
X4权重0.42840.47790.39760.3685
X5权重0.06660.05630.06240.0626
X6权重0.13300.12730.16540.1303
X1权重区间[0.2715,0.2949,0.3062,0.3604]
X2权重区间[0.0378,0.0385,0.0412,0.0440]
X3权重区间[0.0279,0.0285,0.0342,0.0385]
X4权重区间[0.3685,0.3976,0.4284,0.4779]
X5权重区间[0.0563,0.0624,0.0626,0.0666]
X6权重区间[0.1273,0.1303,0.1330,0.1654]
最优指标[0.3083,0.0404,0.0323,0.4181,0.0620,0.1390]

表5

岩爆评判指标权重系数"

权重向量X1X2X3X4X5X6
w10.30820.04040.03230.41810.06200.1390
w20.10970.10460.15140.35300.17690.1044
w00.18910.07890.10370.37900.13110.1182

表6

插补前后岩爆等级预测结果"

数据集各岩爆等级预测准确率平均准确率均方误差
1234
插补前0.75000.80000.86670.66670.79550.4523
插补后1.00000.80000.93751.00000.93100.2626

表7

不同指标组合赋权效果对比"

模型平均预测准确率均方误差
ET0.89660.3216
AHP-CRITIC-ET0.91380.2936
IAHP-CRITIC-ET0.93100.2626

图4

各模型预测等级与实际等级对比"

图5

各模型预测效果比较"

Acurna E, Rodriguez C,2004.The treatment of missing values and its effect in the classifier accuracy[C]//Proceedings of the Meeting of the International Federation of Classification Societies (IFCS).Chicago:International Federation Classification Societies: 639-647.
Afraei S, Shahriar K, Madani S H,2019.Developing intelligent classification models for rockburst prediction after recognizing significant predictor variables,Section 1:Literature review and data preprocessing procedure[J].Tunnelling and Underground Space Technology,83:324-353.
Boshuizen H C, Knook D L,1999.Multiple imputation of missing blood pressure covariates in survival analysis[J].Statistics in Medicine,(7):681-694.
Chen Juan, Wang Xianyu, Luo Lingling,et al,2020.Missing value filling effect:A comparison between machine learning and statistical learning[J].Statistics and Decision Making,36 (17):28-32.
Diakoulaki D, Mavrotas G, Papayannakis L,1995.Determining objective weights in multiple criteria problems:The CRITIC method[J].Computers and Operations Research,22(7):763-770.
Huang Jian, Xia Yuanyou, Lin Manqing,2019.Study on multi-dimensional cloud model prediction of rockburst based on improved combination weighting[J].Chinese Journal of Safety Science,29 (7):26-32.
Li Mingliang, Li Kegang, Qin Qingci,et al,2021.Discussion and selection of machine learning algorithm model for rockburst intensity grade prediction [J].Chinese Journal of Rock Mechanics and Engineering,40 (Supp.1):2806-2816.
Li Renhao, Gu Helong, Li Xibing,et al,2020.A PSO-RBF neural network model for rockburst tendency prediction [J].Gold Science and Technology,28(1):134-141.
Liu Fei,2020.Study on the Evolution and Warning of Rockbursts in Deep-buried Tunnels of the Hanjiang-to-Weihe River Diversion Project by Microseismic Monitoring [D].Dalian:Dalian University of Technology.
Liu Fengqin,2009.Multiple imputation of missing values of income variables based on chain equation[J].Statistical Research,26 (1):71-77.
Long Yanfang,2017.Research on Short-term Traffic Flow Prediction Model Based on Ensembles of Extremely Randomized Trees[D].Changsha:Hunan University.
Lu Furan, Chen Jianhong,2018.Rockburst prediction method based on AHP and entropy weight TOPSIS model [J].Gold Science and Technology,26(3):365-371.
Nugroho H, Utama N P, Surendro K,2021.Class center-based firefly algorithm for handling missing data[J].Journal of Big Data,8(1):1-14.
Qian Chao, Chen Jianxun, Luo Yanbin,et al,2016.Missing data interpolation method for highway tunnel operation based on random forest [J].Transportation System Engineering and Information,16 (3):81-87.
Saaty T L,1994.How to make a decision:The analytic hierarchy process[J].Interfaces,24(6):19-43.
Shang Huandi, Wang Ping, Pei Mingsong,et al,2017.Rockburst prediction based on rough set and weighted grey correlation analysis[J].Industrial Safety and Environmental Protection,43 (6):47-51.
Song Liang, Wan Jianzhou,2020.Comparative study on missing data interpolation methods[J].Statistics and Decision Ma-king,36(18):10-14.
Tan Wenkan, Ye Yicheng, Hu Nanyan,et al,2021.Strong rockburst prediction based on LOF and improved SMOTE algorithm[J].Chinese Journal of Rock Mechanics and Engine-ering,40(6):1186-1194.
Tang Zhili, Xu Qianjun,2020.Research on rockburst prediction based on nine machine learning algorithms[J].Chinese Journal of Rock Mechanics and Engineering,39(4):773-781.
Tian Rui, Meng Haidong, Chen Shijiang,et al,2020a.Prediction of intensity classification of rockburst based on deep neural network [J].Journal of China Coal Society,45(Supp.1):191-201.
Tian Rui, Meng Haidong, Chen Shijiang,et al,2020b.Prediction model of rockburst intensity classification based on RF-AHP-Cloud model[J].Chinese Journal of Safety Science,30(7):166-172.
Wang Junxia, Zhang Yu, Yan Zheming,et al,2013.Research on performance evaluation of rural public goods supply based on combination weighting method[J].Journal of Northwest University(Philosophy and Social Sciences Edition),43(2):117-121.
Wang Xianlong, Feng Zao, Zhao Yanfeng,2021.An active learning method for unbalanced sample set of pipeline blockage[J].Chemical Automation and Instrumentation,48(3):222-231.
Wang Yuanhan, Li Wodong, Li Qiguang,et al,1998 .Fuzzy mathematics comprehensive evaluation method for rockburst prediction[J].Chinese Journal of Rock Mechanics and Engineering,(5):15-23.
Wu H W, Zhen J, Zhang J,2020.Urban rail transit operation safety evaluation based on an improved CRITIC method and cloud model[J].Journal of Rail Transport Planning & Management,16:100206..
Wu Tongyu, Wu Shaoxiong,2018.Missing value interpolation of statistical data based on kernel principal component analysis and particle swarm optimization support vector machine[J].Statistics and Decision Making,34(8):21-24.
Xie X, Jiang W, Guo J,2021.Research on rockburst prediction classification based on GA-XGB model[J].IEEE Access,9:83993-84020.
Xie Xuebin, Li Dexuan, Kong Lingyan,et al,2020.Prediction model of rockburst tendency grade based on CRITIC-XGB algorithm[J].Chinese Journal of Rock Mechanics and Engineering,39(10):1975-1982.
Yin Xin, Liu Quansheng, Wang Xinyu,et al,2020.Prediction model of rockburst intensity classification based on combined weighting and attribute interval recognition theory [J]. Journal of China Coal Society,45(11):3772-3780.
Zhang Xiangyu,2021.Study on Rock Burst Mechanism and Comprehensive Prediction Method of Rock Mass with Structural Plane [D].Jinan:Shandong University.
Zheng Y, Zhong H, Fang Y,et al,2019.Rockburst prediction model based on entropy weight integrated with grey relational BP neural network[J].Advances in Civil Engineering,(4):1-8..
陈娟,王献雨,罗玲玲,等,2020.缺失值填补效果:机器学习与统计学习的比较[J].统计与决策,36(17):28-32.
黄建,夏元友,吝曼卿,2019.基于改进组合赋权的岩爆多维云模型预测研究[J].中国安全科学学报,29(7):26-32.
李明亮,李克钢,秦庆词,等,2021.岩爆烈度等级预测的机器学习算法模型探讨及选择[J].岩石力学与工程学报,40(增1):2806-2816.
李任豪,顾合龙,李夕兵,等,2020.基于PSO-RBF神经网络模型的岩爆倾向性预测[J].黄金科学技术,28(1):134-141.
刘飞,2020.引汉济渭深埋隧洞岩爆孕育特征与微震监测预警研究[D].大连:大连理工大学.
刘凤芹,2009.基于链式方程的收入变量缺失值的多重插补[J].统计研究,26(1):71-77.
龙艳芳,2017.基于极限随机树集成的短时交通流预测模型研究[D].长沙:湖南大学.
卢富然,陈建宏,2018.基于AHP和熵权TOPSIS模型的岩爆预测方法[J].黄金科学技术,26(3):365-371.
钱超,陈建勋,罗彦斌,等,2016.基于随机森林的公路隧道运营缺失数据插补方法[J].交通运输系统工程与信息,16(3):81-87.
商欢迪,王平,裴明松,等,2017.基于粗糙集和加权灰色关联分析的岩爆预测[J].工业安全与环保,43(6):47-51.
宋亮,万建洲,2020.缺失数据插补方法的比较研究[J].统计与决策,36(18):10-14.
谭文侃,叶义成,胡南燕,等,2021.LOF与改进SMOTE算法组合的强烈岩爆预测[J].岩石力学与工程学报,40(6):1186-1194.
汤志立,徐千军,2020.基于9种机器学习算法的岩爆预测研究[J].岩石力学与工程学报,39(4):773-781.
田睿,孟海东,陈世江,等,2020a.基于深度神经网络的岩爆烈度分级预测[J].煤炭学报,45(增1):191-201.
田睿,孟海东,陈世江,等,2020b.RF-AHP-云模型下岩爆烈度分级预测模型[J].中国安全科学学报,30(7):166-172.
王俊霞,张玉,鄢哲明,等,2013.基于组合赋权方法的农村公共产品供给绩效评价研究[J].西北大学学报(哲学社会科学版),43(2):117-121.
王显龙,冯早,赵燕锋,2021.一种面向管道堵塞不均衡样本集的主动学习方法[J].化工自动化及仪表,48(3):222-231.
王元汉,李卧东,李启光,等,1998.岩爆预测的模糊数学综合评判方法[J].岩石力学与工程学报,(5):15-23.
吴桐雨,吴少雄,2018.基于核主成分分析和粒子群优化支持向量机的统计数据缺失值插补[J].统计与决策,34(8):21-24.
谢学斌,李德玄,孔令燕,等,2020.基于CRITIC-XGB算法的岩爆倾向等级预测模型[J].岩石力学与工程学报,39(10):1975-1982.
殷欣,刘泉声,王心语,等,2020.基于组合赋权和属性区间识别理论的岩爆烈度分级预测模型[J].煤炭学报,45(11):3772-3780.
张翔宇,2021.含结构面岩体岩爆发生机理及综合预测方法研究[D].济南:山东大学.
[1] 李筱, 许钧, 张成旭, 隋来伦, 王在勇. 基于CWM-TOPSIS模型的金属矿山企业安全管控能力评价[J]. 黄金科学技术, 2024, 32(1): 100-108.
[2] 杨玮, 邓博, 龙涛, 邓莎, 薛梦鸽, 方楠. 基于效果—效率的金矿绿色矿山建设综合评价研究[J]. 黄金科学技术, 2023, 31(6): 919-929.
[3] 杨玮, 薛梦鸽, 龙涛, 邓莎, 邓博, 方楠. 基于DPSIR模型的黄金行业绿色矿山建设综合评价研究[J]. 黄金科学技术, 2023, 31(4): 635-645.
[4] 徐先锋,邢鹏飞,王岁红,汪泳. 基于博弈论G1-EW-TOPSIS法的岩体质量评价和应用[J]. 黄金科学技术, 2022, 30(5): 704-712.
[5] 柯愈贤,王成,方立发,廖宝泉. 基于组合权重和物元分析的矿山安全生产状况研究[J]. 黄金科学技术, 2020, 28(6): 910-919.
[6] 李彤彤, 王玺, 刘焕新, 侯奎奎, 李夕兵. 基于组合赋权的T-FME岩爆倾向性预测模型研究及应用[J]. 黄金科学技术, 2020, 28(4): 565-574.
[7] 赵国彦,吴攀,朱幸福,赵源,李洋,邱菊. 基于灰色关联分析的三山岛金矿绿色开采技术优先级评价[J]. 黄金科学技术, 2019, 27(6): 835-843.
[8] 李欢,明俊桦,石晓凤. 基于博弈论和模糊综合评判的锌冶炼企业清洁生产评价[J]. 黄金科学技术, 2018, 26(5): 635-646.
[9] 周海林,杨珊,陈建宏*. 黄金矿山清洁生产评价指标体系研究与应用[J]. 黄金科学技术, 2017, 25(5): 93-100.
[10] 万孝衡,王新民,朱阳亚,姜志良,陈秋松. 基于组合赋权TOPSIS法的采场结构参数优选[J]. 黄金科学技术, 2014, 22(5): 69-73.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!