基于 Stacking 模型的采空区稳定性预测

doi:10.11872/j.issn.1005-2518.2020.06.049

Abstract

Abstract:

In the process of mining，the mining technology mainly based on the open field method will leave a large number of goaves after mining，which has great safety risks.Therefore，the prediction of the stability of the goaf is particularly important.In recent years，with the advent of the era of big data，machine learning and deep learning technology has been introduced into the mining industry.More and more scholars begin to use the prediction method of machine learning to study the stability of goaf.It is found that every single machine term model has its scope of use and scene，when facing a new goaf problem，we can not immediately give the most effective model to adapt to it.At this time，the stacking method derived from integrated learning is born.This article is mainly divided into three parts.The first part mainly introduces the Stacking method，explains the main steps of the method and the choice of the learner.Stacking is generally a two-layer structure.The first layer is composed of various elementary learners.Input the original data set to train them.The second layer uses the output of these primary learners as a new data set and new features into the secondary learner （meta-learner） for training，to complete the stack fusion of the model.The Stacking model requires that the correlation between the primary learners should be as small as possible，and the performance difference between the primary learners should not be too large.The second-level learner should choose a simpler learner.The second part is the construction of Stacking model.This paper selected eleven factors affecting the stability of mined-out area，they are the rock mass structure （X₁），geological structure（X₂），groundwater （X₃），goaf engineering layout （X₄），mining perturbation（X₅），adjacent cavity （X₆），the volume of the goaf （X₇），the exposed area of goaf roof （X₈），buried depth （X₉），goaf span mined-out area ratio （X₁₀），rock compressive strength（X₁₁）.By using the eleven dimensional data and the classification of goaf stability as training data，we put them into the stacking model with random frost，AdaBoost，extratrees and lightgbm as primary learners and logistic regression as secondary learners for training.Another part of the training data was used as test data to compare the performance of stacking model and single machine learning model.The experimental results show that the F1 and AUC of stacking model are 0.967 and 0.97 respectively.Far higher than all other single machine learning models.The stacking model is better than other single machine learning models and show stronger generalization ability.Therefore，as the third part，it is concluded that the stacking model can predict the stability of goaf better than the single machine learning model.

Key words: goaf, machine learning, ensemble learning, Stacking model, stability prediction, stability level

CLC Number:

TD76

Mufan WANG,Zhouquan LUO,Qi YU. Stability Prediction of Goaf Based on Stacking Model[J].Gold Science and Technology, 2020, 28(6): 894-901.

Figures/Tables 9

Fig.1

Fig.2

Table 1

Table 2

Table 3

Table 4

Table 5

Fig.3

Table 6

References 22

1	冯远建，李子龙.矿山采空区稳定性分析及安全治理方法研究［J］.煤炭与化工，2014，37（6）：28-30.
	Feng Yuanjian，Li Zilong.Mine goaf stability analysis and security management method research［J］.Coal and Che-mical Industry，2014，37（6）：28-30.
2	晁军，雷醒民，樊育豪.三维有限元数值模拟在采空区稳定性评价中的应用［J］.工程建设与设计，2018（23）：15-18.
	Chao Jun，Lei Xingmin，Fan Yuhao.Application of three-dimensional finite element numerical simulation in stability evaluation of mined-out area［J］.Construction and Design for Engineering，2018（23）：15-18.
3	杨宁，尹贤刚，钟勇，等.基于3DMine-FLAC3D的复杂空区稳定性分析［J］.化工矿物与加工，2019（4）：22-24，28.
	Yang Ning，Yin Xiangang，Zhong Yong，et al.Stability analysis of complex goaf based on 3DMine-FLAC^3D［J］.Industrial Minerals and Processing，2019（4）：22-24，28.
4	邓岁伟，冯锐，赵铭久.基于ANSYS的采空区稳定性数值模拟分析［J］.采矿技术，2018，18（1）：34-37.
	Deng Suiwei，Feng Rui，Zhao Mingjiu.Numerical simulation analysis of goaf stability based on ANSYS［J］.Mining Technology，2018，18（1）：34-37.
5	季惠龙，侯克鹏，张成良，等.大型采空区顶板冒落危害预测［J］.采矿技术，2010，10（2）：50-52，76.
	Ji Huilong，Hou Kepeng，Zhang Chengliang，et al.Hazard prediction of roof caving in large goaf ［J］.Mining Technology，2010，10（2）：50-52，76.
6	曾令义，潘冬.崩落法转充填法隔离层安全厚度的确定［J］.矿冶工程，2015，35（4）：17-20.
	Zeng Lingyi，Pan Dong.Determination of safety thickness of isolation layer during transition from caving to cut-and-fill stoping［J］.Mining and Metallurgical Engineering，2015，35（4）：17-20.
7	张慧生，顾伟.典型曲线法在山东龙口软岩煤矿区开采沉陷预测中的应用［J］.地质找矿论丛，2014，29（1）：84-90.
	Zhang Huisheng，Gu Wei.Establishment and application of typical curve method for Longkou soft rock coal-mining area in Shandong Province［J］.Contributions to Geology and Mineral Resources Research，2014，29（1）：84-90.
8	许章平，栾元重，相涛，等.基于概率积分法和Kalman滤波模型的矿区变形预测［J］.煤炭技术，2018，37（12）：111-114.
	Xu Zhangping，Luan Yuanzhong，Xiang Tao.Mining deformation prediction based on probability integral method and Kalman filter model［J］.Coal Technology，2018，37（12）：111-114.
9	孟东芳，唐皓，童锋.基于突变级数法的岩体质量评价［J］.中国煤炭地质，2018，30（10）：64-69.
	Meng Dongfang，Tang Hao，Tong Feng.Rock mass quality assessment based on catastrophe progression method［J］.Coal Geology of China，2018，30（10）：64-69.
10	王海峰，李夕兵，董陇军，等.基于支持向量机的采空区稳定性分级［J］.中国安全生产科学技术，2014，10（10）：154-159.
	Wang Haifeng，Li Xibing，Dong Longjun，et al.Classification of goaf stability based on support vector machine［J］.Journal of Safety Science and Technology，2014，10（10）：154-159.
11	黄新典，褚夫蛟.基于逻辑回归-聚类算法的采空区危险等级评价模型［J］.金属矿山，2019，48（8）：179-184.
	Huang Xindian，Chu Fujiao.Risk grade evaluation model of goaf based on logical regression and clustering algorithm［J］.Metal Mine，2019，48（8）：179-184.
12	李孜军，林武清，陈阳.基于AGA-BP神经网络的采空区危险性评价［J］.中国安全生产科学技术，2015，11（7）：135- 141.
	Li Zijun，Lin Wuqing，Chen Yang.Evaluation on risk of goaf based on AGA-BP neural network［J］.Journal of Safety Science and Technology，2015，11（7）：135- 141.
13	Qin Y G，Luo Z Q，Ma S W，et al.Evaluation of goaf stability based on transfer learning theory of artificial intelligence［J］.IEEE Access，2019，7：96912-96925.
14	李岩，赵建文.基于PCA-SVM 的采空区危险性评价［J］.河北冶金，2014（11）：72-77.
	Li Yan，Zhao Jianwen.Dangerous assessment based on PCA-SVM for stoped-out area［J］.Hebei Metallurgy，2014（11）：72-77.
15	Wolpert D H.Stacked generalization［J］.Neural Networks，1992，5（2）：241-259.
16	Breiman L.Random forests［J］.Machine Learning，2001，45（1）：5-32.
17	Freund Y，Schapire R E.A decision-theoretic generalization of on-line learning and an application to boosting［C］// Proceeding of the Second European Conference on Computational Learning Theory.Berlin：Springer-Verlag，1995：23-27.
18	Geurts P， Ernst D， Wehenkel L.Extremely randomized trees［J］.Machine Learning， 2006， 63（1）：3-42.
19	Ke G L，Meng Q，Finley T，et al.LightGBM：A highly efficient gradient boosting decision tree［C］//Proceedings of the 31st International Conference on Neural Information Processing Systems（NIPS' 17）.New York：Curran Associates Inc.，Red Hook，2017：3149-3157.
20	陈怀利.基于盲数理论的采空区危险性评价模型［J］.采矿技术，2019，19（3）：43-46.
	Chen Huaili.Risk evaluation model of goaf based on blind number theory［J］.Mining Technology，2019，19（3）：43-46.
21	赵超，马俊超，万黎明，等.基于层次可拓分析法的采空区塌陷危险性评价研究［J］.安全与环境工程，2016，23（6）：35-40.
	Zhao Chao，Ma Junchao，Wan Liming，et al.Risk evaluation of underground goaf collapse based on the hierarchy extension analysis［J］.Safety and Environmental Engineering，2016，23（6）：35-40.
22	汪朝，郭进平，王李管.采空区危险性的支持向量机识别［J］.重庆大学学报（自然科学版），2015，38（4）：85-90.
	Wang Zhao，Guo Jinping，Wang Liguan.Recognition of goaf risk based on support vector machines method［J］.Journal of Chongqing University（Natural Science Edition），2015，38（4）：85-90.

Related Articles 15

[1]	Rui TIAN,Haidong MENG,Shijiang CHEN,Chuangye WANG,Dening SUN,Lei SHI. Comparative Study on Three Rockburst Prediction Models of Intensity Classi-fication Based on Machine Learning [J]. Gold Science and Technology, 2020, 28(6): 920-929.
[2]	Zhiqin LIAO, Liguan WANG, Zhengxiang HE. Feature Extraction and Classification of Mine Microseismic Signals Based on EEMD and Correlation Dimension [J]. Gold Science and Technology, 2020, 28(4): 585-594.
[3]	Mingzhi DANG,Jun ZHANG,Mingtao JIA. Application and Research of Microseismic Monitoring Technology and Disaster Early Warning Methods in Huangtupo Copper and Zinc Mine [J]. Gold Science and Technology, 2020, 28(2): 246-254.
[4]	Hongwei DENG,Weiyou ZHANG,Songtao YU,Yuxu GAO. Two Dimensional Evaluation Model of Goaf Stability Based on Variable Weight Contact Cloud [J]. Gold Science and Technology, 2020, 28(1): 32-41.
[5]	Li CHENG,Huanxin LIU,Mingde ZHU,Qinzheng WU. Current Situation and Prospect of Research on Underground Goaf in Metal Mines [J]. Gold Science and Technology, 2020, 28(1): 70-81.
[6]	Yuan GAO,Qingfa CHEN,Tenglong JIANG. Research on Three-Dimensional Numerical Model Construction Method and Cemented Filling Treatment of Complex Cavities in Daxin Manganese Mine [J]. Gold Science and Technology, 2019, 27(6): 851-861.
[7]	Shijiao YANG,Zhihui WANG. Numerical Simulation on Stability of Shallow-buried Goaf Group with Overlying Highway [J]. Gold Science and Technology, 2019, 27(4): 505-512.
[8]	Xuebin XIE,Huchen XIONG,Herong XIE,Jiankun LI,Tingyu TIAN. Catastrophic Instability Model of Goaf Group System Considering Horizontal Loads [J]. Gold Science and Technology, 2019, 27(3): 368-377.
[9]	Nao LV,Haibo WANG. Numerical Simulation of Collapse Disaster Before and After Weakening of Thick and Hard Roof [J]. Gold Science and Technology, 2019, 27(2): 257-264.
[10]	Zilong ZHOU,Yifan WANG,Changtao KE. Reliability Evaluation of Pillars System Based on Domino Failure Effect [J]. Gold Science and Technology, 2018, 26(6): 729-735.
[11]	Zhongyuan GU,Keping ZHOU. Study on Time-Variant Mechanics Properties of Shallow Goaf Group Based on Creep Experiment [J]. Gold Science and Technology, 2018, 26(4): 511-519.
[12]	BAI Zhaoyang，WANG Guowei，ZHANG Peng，LIU Shuanping. Research on Pillaring Blasting Location Under Level Gob Group Conditions [J]. Gold Science and Technology, 2017, 25(4): 81-86.
[13]	DENG Gao，YANG Shan. Stability Evaluation of Goafs Based on Combined Forecasting and Variable Precision Rough Fuzzy Set [J]. Gold Science and Technology, 2017, 25(3): 98-107.
[14]	CAO Zuhua，LUO Zhouquan，QIN Yaguang，WANG Yule. #br# Local Search Simulated Annealing for Filtering of Point Cloud [J]. Gold Science and Technology, 2017, 25(1): 106-111.
[15]	LUO Zhouquan，QIN Yaguang，ZHOU Jiming . Development of CAD Data Interface in 3D Modeling Visualized Integrated System of Goaf [J]. Gold Science and Technology, 2015, 23(5): 41-46.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

Comments

Recommended 10

[1]	YAN Jie, QIN Ze-Li, XIE Wen-Bing, CA Bang-Yong. [J]. J4, 2010, 18(4): 22 -26 .
[2]	SONG He-Min, FENG Chi-Li, DING Xian-Hua. Geological-Geochemical Charicteristics and Prospecting Direction of Jiaojiekou Mining Area, North Part of Taihang Mountain[J]. J4, 2010, 18(3): 54 -58 .
[3]	LI Chu-Fang, XU Yong-An, CHAO Yin-Yin, WANG Mei-Juan, ZHANG Dai, LIU Jun, SUN Liang-Liang. Looking for Strata Bound Type Gold Deposit in Liaodong Metallogenic Belt[J]. J4, 2010, 18(3): 59 -62 .
[4]	HU Qin-Xia, LI Jian-Zhong, YU Guang-Meng, XIE Yan-Fang, ZHANG Ku-Xiao. Discussion on Gold Ore Point of Bailongjiang Metallogenic Belt[J]. J4, 2010, 18(3): 51 -53 .
[5]	CHEN Hua-Dun. [J]. J4, 2010, 18(4): 50 -53 .
[6]	CUI Ting-Jun, DAI Ke-Sai, PENG Yong, FU Xing. Discussion on the Geological Characteristics and Metallogenic Regularity of the Southern Edge of Qaidam Basin Gold Metallogenic Belt in Qinghai Province[J]. J4, 2010, 18(3): 63 -67 .
[7]	YANG Meng-Rong, MAO Chang-Xian. Uncertainty Evaluation of Arsenic and Antimony in Chemical Prospecting Sa-mple by Atomic Fluorescence Spectrometry[J]. J4, 2010, 18(3): 68 -71 .
[8]	SU Jian-Hua, LIU Shu-Lin. Study of Gold Extraction from Tail Solution with High Acid Content and Low Concentration[J]. J4, 2010, 18(3): 72 -75 .
[9]	WANG Da-Beng, SONG Bing-Jian, HUI Ku-Meng. The Application of High Power IP Method for Searching Concealed Metallic Mine in Beishuiquan，Liaoning Province[J]. J4, 2010, 18(3): 76 -78 .
[10]	LIU Qing-Guang, GAO Hai-Feng, HUANG Suo-Yang. The Application of PDA in Geology Department of Jiaojia Gold Mine[J]. J4, 2010, 18(3): 79 -82 .

稳定性等级	稳定性状况	失稳风险
1	采空区稳定性很好	无失稳风险或失稳风险很小
2	采空区基本稳定或稳定性较好	失稳风险较低
3	采空区稳定性较差	失稳风险较大
4	采空区不稳定或稳定性极差	失稳风险很大

因素影响程度等级	地质因素		水文因素	采空区因素	环境因素		评分范围
因素影响程度等级	岩体结构（X₁）	地质构造（X₂）	地下水（X₃）	采空区工程布置（X₄）	采动扰动情况（X₅）	相邻空区情况（X₆）	评分范围
1	完整块状结构	无断层、褶皱	无淋水痕迹	合理	开采范围无爆破作业影响	影响范围内无其他采空区，为孤立空区或为6R之外（R为孤立空区半径）	0.75~1.00
2	层状结构	褶皱影响小	围岩可见水迹	比较合理	开采范围爆破作业影响一般	影响范围内采空区面积一般，数量不多且相邻较近	0.50~0.75
3	碎裂结构	断层部分切割或褶皱影响大	雨季有淋水	部分合理	开采范围采场作业影响较大	影响范围内采空区面积大，数量多，但分布较为分散	0.25~050
4	松散结构	断层贯穿围岩	长期有淋水	布置不合理	开采范围采场作业影响大	影响范围内采空区面积较大，数量较多，相邻较近且比较密集，为采空区群	0~0.25

序号	岩体结构（X₁）	地质构造（X₂）	地下水（X₃）	采空区工程布置（X₄）	采动扰动情况（X₅）	相邻空区情况（X₆）	采空区体积（X₇）/m³	采空区顶板暴露面积（X₈）/m²	采空区埋藏深度（X₉）/m	采空区跨高比（X₁₀）	岩石抗压强度（X₁₁）/MPa	等级
1	3	1	2	2	4	4	46 680	4 330	370	4	39	4
2	2	2	2	1	4	4	11 700	1 300	250	2	49	2
3	3	3	2	2	4	4	32 240	2 480	300	3	37	3
4	3	2	2	3	4	4	33 600	1 680	270	4	50	3
5	3	1	3	2	4	4	179 400	5 100	332	3	28	4
6	3	3	2	3	4	4	137 760	5 740	295	4	52	3
7	2	1	1	1	4	4	46 560	2 910	305	1	53	1
8	2	1	2	1	1	1	43 600	2 180	320	2	59	2
9	1	1	2	1	4	4	12 000	1 200	335	1	66	1
10	1	2	3	2	4	4	232 580	4 010	380	2	50	3
11	1	2	1	1	4	4	33 900	2 260	305	2	59	1
12	1	1	2	1	1	1	18 850	1 450	290	1	61	1
13	1	1	2	1	3	3	18 130	2 590	201	1	52	1
14	1	1	1	1	3	3	29 160	2 430	208	1	55	1
15	3	3	2	2	3	3	10 575	2 350	208	1	56	2
16	1	1	1	1	3	3	18 000	1 800	208	2	54	1
17	1	2	2	2	4	4	25 600	1 600	411	3	57	2
18	1	1	2	1	1	1	16 000	2 000	300	2	54	2
19	1	1	1	1	3	3	44 850	1 950	201	2	55	2
20	1	1	1	1	3	3	37 500	5 000	195	1	52	1
21	1	1	1	1	3	3	42 510	3 270	180	2	51	2
22	1	2	1	2	4	3	9 750	1 300	180	2	54	2
23	1	2	1	1	4	4	10 726	1 730	180	2	55	2
24	1	1	1	1	3	3	11 220	1 870	230	1	53	1
25	1	1	1	1	3	3	16 380	1 170	230	2	53	2
26	1	1	1	1	1	1	39 840	2 490	230	2	54	1
27	1	1	1	1	1	1	31 720	2 440	230	2	53	1
28	1	1	2	3	4	4	80 040	3 480	230	3	43	3
29	1	1	1	1	1	1	11 280	1 410	230	2	55	1

序号	岩体结构（X₁）	地质构造（X₂）	地下水（X₃）	采空区工程布置（X₄）	采动扰动情况（X₅）	相邻空区情况（X₆）	采空区体积（X₇）/m³	采空区顶板暴露面积（X₈）/m²	采空区埋藏深度（X₉）/m	采空区跨高比（X₁₀）	岩石抗压强度（X₁₁）/MPa	等级
1	1	1	1	1	3	3	39 843	4 942	190	1	52	1
2	2	2	3	1	1	1	70 477	2 540	202	1	49	3
3	2	4	2	3	3	4	26 360	1 191	261	3	73	3
4	2	1	1	1	1	1	70 815	1 595	192	2	46	2
5	1	1	2	2	3	4	117 333	1 882	404	2	45	2
6	3	3	2	2	1	1	104 661	2 498	165	3	59	3
7	1	3	3	2	3	2	6 128	3 498	232	1	50	2
8	2	2	4	2	4	4	86 811	1 326	278	3	44	4
9	2	1	1	3	1	3	125 955	3 888	415	1	44	2
10	2	4	2	3	3	4	26 360	1 191	261	2	73	3
11	1	1	1	1	3	3	11 220	1 986	225	1	54	1
12	1	1	1	1	1	1	74 880	2 186	191	2	52	2
13	1	2	1	2	4	3	9 750	1 359	184	2	54	2
14	1	1	1	1	1	1	39 840	2 547	225	2	54	1
15	3	3	2	2	1	1	104 661	2 498	165	3	59	3

模型	准确率	召回率	F1 值
Random Frost	0.911	0.866	0.868
Adaboost	0.946	0.933	0.934
ExtraTrees	0.927	0.900	0.901
LightGBM	0.927	0.900	0.901
Stacking	0.970	0.966	0.967

Stability Prediction of Goaf Based on Stacking Model

RichHTML

PDF (PC)