基于Bootstrap-SVR-ANN算法的TBM施工速度预测

闫长斌; 汪鹤健; 周建军; 杨风威; 彭万军

doi:10.11779/CJGE202106011

基于Bootstrap-SVR-ANN算法的TBM施工速度预测 English Version

1.
郑州大学土木工程学院,河南　郑州　450001
2.
盾构及掘进技术国家重点实验室,河南　郑州　450001
3.
黄河勘测规划设计研究院有限公司,河南　郑州　450003

基金项目:

国家自然科学基金项目 41972270

国家自然科学基金项目 U1504523

河南省重点研发与推广专项 182102210014

盾构及掘进技术国家重点实验室开放课题 SKLST-2019-K06

详细信息

作者简介:
闫长斌(1979—),男,教授,博士,主要从事隧道与地下工程等方面的教学和科研工作。E-mail: yanchangbin_2001@163.com

中图分类号: U455;TV554
计量
- 文章访问数: 300
- HTML全文浏览量: 45
- PDF下载量: 167
出版历程
- 收稿日期: 2020-07-21
- 网络出版日期: 2022-12-02
- 刊出日期: 2021-05-31

Prediction of TBM advance rate based on Bootstrap method and SVR-ANN algorithm

1.
School of Civil Engineering, Zhengzhou University, Zhengzhou 450001, China
2.
State Key Laboratory of Shield Machine and Boring Technology, Zhengzhou 450001, China
3.
Yellow River Engineering Consulting Co., Ltd., Zhengzhou 450003, China

摘要

摘要: 合理评价预测施工速度关乎隧道TBM施工的成败与效益。现有的TBM施工速度预测模型多利用岩体参数和掘进参数预测瞬时/平均施工速度,对掘进过程中的不确定性和施工风险考虑不足。基于此,引入区间预测方法,提出一种基于Bootstrap-SVR-ANN算法的TBM施工速度预测模型。以兰州水源地建设工程输水隧洞双护盾TBM施工为工程依托,分析了单一性输入参数的不足,指出了选择岩体质量分级指标（RMR）、TBM工作条件等级（TWCR）两个综合性参数的合理性,并对构建的TBM施工速度区间预测模型的有效性进行了验证。研究表明,TBM施工速度区间预测模型不但具有良好的点预测效果,而且预测区间可将TBM施工速度的实测值完全包络,模型可靠性较高；模型测试集在90%,95%置信水平下的MPIW分别为9.84,11.73 m/d,随着置信水平的提高,预测区间可容纳的不确定性也不断上升；TBM掘进过程中可能的风险性与区间宽度的异常相互印证,验证了区间预测模型可以定量解释施工过程中不确定性的特点。研究成果可为TBM掘进效能预测、施工工期估算和掘进参数优化等提供科学参考。
- 隧道掘进机（TBM） /
- 施工速度 /
- 区间预测 /
- Bootstrap方法 /
- 支持向量回归 /
- 神经网络
Abstract: The reasonable prediction and evaluation of advance rate is related to the success and benefit of TBM construction. The existing prediction models for TBM advance rate mostly use parameters of rock mass and TBM tunneling to predict the instantaneous/average advance rate. To solve the problem that these models do not consider the influences of uncertainty and risk during TBM tunneling process, a prediction model for TBM advance rate based on the Bootstrap method and SVR-ANN algorithm is proposed by introducing the idea of interval prediction. Based on the project of Lanzhou water sources water conveyance tunnel constructed by double-shield TBM, the shortcomings of some single input parameters are analyzed, and the rationality of two selected comprehensive parameters, namely rock mass quality classification index (RMR) and TBM working condition class (TWCR), is pointed out. In addition, the validity of the developed interval prediction model for TBM advance rate is verified. The results show that the developed interval prediction model for TBM advance rate provides relatively accurate point prediction results and constructs a clear and reliable AR prediction interval to cover the actual TBM advance rate completely. The MPIW of model test set at confidence levels of 90% and 95% is 9.84, 11.73 m/d, respectively. With the improvement of the confidence level, the uncertainty that can be contained in the prediction interval also increases. Moreover, the possible risk of TBM tunneling process and the abnormal interval width confirm each other, which verifies that the interval prediction model can quantitatively explain the characteristics of uncertainty in the construction process. The research results may provide a new idea for the forecasting of TBM tunneling efficiency, the estimation of construction schedule as well as the optimization of tunneling parameters.
- tunnel boring machine /
- advance rate /
- interval prediction /
- Bootstrap method /
- support vector regression /
- artificial neural network

HTML全文

图 1 区间预测流程简图

Figure 1. Flow chart of interval prediction

下载: 全尺寸图片幻灯片

图 2 输水隧洞线路区工程地质剖面图

Figure 2. Engineering geological section of route area of water conveyance tunnel

下载: 全尺寸图片幻灯片

图 3 参数相关性分析

Figure 3. Correlation between parameters

下载: 全尺寸图片幻灯片

图 4 区间预测结果对比

Figure 4. Comparison of interval prediction results

下载: 全尺寸图片幻灯片

表 1 TBM主要设计参数

Table 1 Main design parameters of TBM

设计参数	开挖直径/mm	滚刀数量/把	中心滚刀		正滚刀		边滚刀		最大刀间距/mm	刀盘转速/(r·min^-1)	刀盘功率/kW	最大刀盘推力/kN	额定扭矩/(kN·m^-1)	脱困扭矩/(kN·m^-1)	最大掘进速度/(mm·min^-1)
设计参数	开挖直径/mm	滚刀数量/把	数量/把	直径/mm	数量/把	直径/mm	数量/把	直径/mm	最大刀间距/mm	刀盘转速/(r·min^-1)	刀盘功率/kW	最大刀盘推力/kN	额定扭矩/(kN·m^-1)	脱困扭矩/(kN·m^-1)	最大掘进速度/(mm·min^-1)
TBM1	5480	37	6	432	21	483	10	483	86	0~10.3	1800	22160	3458	5878	120
TBM2	5480	30	4	432	17	483	9	483	83	0~8.7	2100	11900	4210	6940	120

下载: 导出CSV

表 2 不同岩性地质单元中关键参数

Table 2 Key parameters in different lithologic geological units

岩性地质单元	里程桩号/m	UCS/MPa	CAI	TF/10³kN	RPM/(r·min^-1)	RMR平均值	TBM工作条件等级	AR/(m·d^-1)
1	T6+679.1—T6+914.9	70	3.02	8.95	6.80	71	3	26.20
2	T6+914.9—T7+122.3	70	3.02	8.57	6.70	65	3	25.93
3	T7+462.9—T8+130.3	70	3.49	8.33	6.50	69	3	22.25
4	T8+130.3—T8+441.3	70	3.02	8.71	6.00	70	3	14.81
5	T8+615.6—T9+199.4	70	3.45	8.52	6.40	77	3	26.54
6	T9+199.4—T9+259.8	15	2.93	3.50	4.50	10	11	2.88
7	T9+259.8—T9+331.8	45	2.93	4.00	5.00	20	9	5.54
8	T9+468.4—T9+979.4	70	2.93	8.68	6.70	72	2	17.03
9	T10+934.5—T11+110.7	60	3.45	8.22	6.40	51	6	29.37
10	T11+110.7—T11+555.8	45	2.93	5.48	4.90	31	8	14.36
11	T12+969.5—T13+847.5	40	0.82	4.50	5.50	54	4	28.26
12	T13+847.5—T14+662.9	40	0.82	4.30	5.40	56	4	26.30
13	T14+977.3—T15+387.5	30	0.82	4.15	5.10	39	7	20.51
14	T19+747.8—T19+462.3	15	1.82	3.25	4.50	11	11	9.21
15	T19+815.5—T19+747.8	40	1.82	3.82	5.10	27	7	22.57
16	T20+060.0—T19+815.5	50	2.51	4.85	5.50	55	5	20.37
17	T20+282.5—T20+060.0	50	2.51	5.08	6.30	60	5	22.25
18	T20+515.1—T20+282.5	65	2.51	8.35	6.10	61	2	17.89
19	T20+801.5—T20+515.1	65	2.51	8.10	6.90	62	2	17.90
20	T20+887.6—T20+801.5	40	1.82	3.75	5.70	33	8	21.53
21	T21+360.3—T21+115.8	65	1.82	6.25	6.60	71	2	14.38
22	T22+888.2—T22+821.9	65	2.51	5.85	6.80	65	2	16.58
23	T23+165.5—T22+888.2	50	1.82	4.85	5.90	57	5	27.73
24	T23+467.1—T23+290.0	50	2.51	4.57	5.40	51	5	25.30
25	T23+638.5—T23+467.1	50	2.51	5.25	5.80	49	5	24.49
26	T23+791.4—T23+638.5	50	2.51	4.68	5.10	47	5	21.84
27	T24+877.8—T24+768.1	30	0.82	3.67	5.40	28	7	21.94
28	T25+051.7—T24+880.0	40	0.82	4.85	6.20	45	4	28.62
29	T25+225.9—T25+051.7	40	0.82	5.15	5.10	47	4	24.89
30	T25+558.4—T25+370.8	30	0.82	3.76	5.50	32	7	23.45
31	T25+926.9—T25+643.2	40	0.82	3.80	5.40	50	4	20.26
32	T26+341.3—T26+025.9	30	0.82	4.33	4.80	35	7	26.28
33	T27+587.3—T27+126.3	30	0.82	3.57	4.70	26	7	21.95
34	T28+429.1—T28+274.9	40	0.75	3.30	4.90	39	4	22.03
35	T29+505.7—T29.378.8	30	0.82	3.85	5.50	27	7	21.15

下载: 导出CSV

表 3 不同影响参数的拟合决定系数R²

Table 3 R² of different influence parameters

影响参数	线性函数	二次函数	对数函数	幂函数	S函数	指数函数
UCS	0.024	0.188	0.058	0.111	0.193	0.058
CAI	0.069	0.157	0.090	0.087	0.093	0.075
TF	0.020	0.061	0.029	0.048	0.064	0.035
RPM	0.081	0.310	0.097	0.138	0.160	0.117

下载: 导出CSV

表 4 AR的点预测与区间预测结果

Table 4 Results of point and interval prediction of AR

序号	里程桩号/m	围岩类别	真实值/(m·d^-1)	点预测值/(m·d^-1)	置信水平90%区间/(m·d^-1)	置信水平95%区间/(m·d^-1)
1	T7+122.3—T7+322.2	石英片岩Ⅱ	22.21	23.46	[21.404,25.517]	[21.010,25.911]
2	T9+331.8—T9+468.4	花岗岩Ⅱ	17.08	22.46	[13.610,31.311]	[11.915,33.005]
3	T21+115.8—T20+887.6	变质安山岩Ⅲ	22.82	24.81	[21.535,28.091]	[20.907,28.718]
4	T21+582.3—T21+360.3	变质安山岩Ⅱ	18.50	22.67	[15.810,29.530]	[14.497,30.844]
5	T22+475.6—T22+369.5	变质安山岩Ⅱ	17.68	22.89	[14.320,31.464]	[12.677,33.106]
6	T23+949.2—T23+852.9	变质安山岩Ⅳ	24.07	19.60	[12.255,26.951]	[10.848,28.358]
7	T26+527.8—T26+341.3	砂岩Ⅲ	26.64	24.11	[19.952,28.271]	[19.156,29.067]
8	T27+821.5—T27+663.9	砂岩Ⅲ	26.26	24.99	[22.913,27.076]	[22.514,27.475]
9	T28+051.6—T27+849.9	砂岩Ⅳ	22.41	19.79	[15.486,24.098]	[14.662,24.923]
10	T29+963.9—T29+840.6	砂岩Ⅳ	20.55	19.52	[17.827,21.214]	[17.503,21.538]

下载: 导出CSV

参考文献(19)

[1]	王梦恕. 岩石隧道掘进机(TBM)施工及工程实例[M]. 北京: 中国铁道出版社, 2004. WANG Meng-shu. Rock Tunnel Boring Machine (TBM) and Engineering Projects[M]. Beijing: China Railway Publishing House, 2004. (in Chinese)
[2]	BRULAND A. Hard Rock Tunnel Boring[D]. Trondheim: Norwegian University of Science and Technology, 1998.
[3]	BARTON N. TBM Tunnelling in Jointed and Faulted Rock[M]. Boca Raton: CRC Press, 2000: 170-175.
[4]	BENARDOS A G, KALIAMPAKOS D C. Modelling TBM performance with artificial neural networks[J]. Tunnelling and Underground Space Technology, 2004, 19(6): 597-605. doi: 10.1016/j.tust.2004.02.128
[5]	GRIMA M A, BRUINES P A, VERHOEF P N W. Modeling tunnel boring machine performance by neuro-fuzzy methods[J]. Tunnelling and Underground Space Technology, 2000, 15(4): 501. doi: 10.1016/S0886-7798(01)00022-0
[6]	ZHOU J, BEJARBANEH B Y, ARMAGHANI D J, et al. Forecasting of TBM advance rate in hard rock condition based on artificial neural network and genetic programming echniques[J]. Bulletin of Engineering Geology and the Environment, 2020, 79(4): 2069-2084. doi: 10.1007/s10064-019-01626-8
[7]	蒋朝辉, 董梦林, 桂卫华, 等. 基于Bootstrap 的高炉铁水硅含量二维预报[J]. 自动化学报, 2016, 42(5): 715-723. https://www.cnki.com.cn/Article/CJFDTOTAL-MOTO201605011.htm JIANG Zhao-hui, DONG Meng-lin, GUI Wei-hua, et al. Two-dimensional prediction for silicon content of hot metal of blast furnace based on bootstrap[J]. Acta Automatica Sinica, 2016, 42(5): 715-723. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-MOTO201605011.htm
[8]	EFRON B. Bootstrap methods: another look at the jackknife[J]. Annals of Statistics, 1979, 7(1): 1-26.
[9]	LI D, TANG X, PHOON K. Bootstrap method for characterizing the effect of uncertainty in shear strength parameters on slope reliability[J]. Reliability Engineering and System Safety, 2015, 140: 99-106. doi: 10.1016/j.ress.2015.03.034
[10]	WAN C, XU Z, WANG Y, et al. A hybrid approach for probabilistic forecasting of electricity price[J]. IEEE Transactions on Smart Grid, 2014, 5(1): 463-470. doi: 10.1109/TSG.2013.2274465
[11]	MA J, TANG H, LIU X, et al. Probabilistic forecasting of landslide displacement accounting for epistemic uncertainty: a case study in the Three Gorges Reservoir area, China[J]. Landslides, 2018, 15(6): 1145-1153. doi: 10.1007/s10346-017-0941-5
[12]	李麟玮, 吴益平, 苗发盛, 等. 基于不同Bootstrap 方法和KELM-BPNN 模型的滑坡位移区间预测[J]. 岩石力学与工程学报, 2019, 38(5): 912-926. https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX201905006.htm LI Lin-wei, WU Yi-ping, MIAO Fa-sheng, et al. Landslide displacement interval prediction based on different Bootstrap methods and KELM-BPNN model[J]. Chinese Journal of Rock Mechanics and Engineering, 2019, 38(5): 912-926. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX201905006.htm
[13]	孙东磊, 王艳, 于一潇, 等. 基于BP 神经网络的短期光伏集群功率区间预测[J]. 山东大学学报(工学版), 2020, 50(4): 1-7. https://www.cnki.com.cn/Article/CJFDTOTAL-SDGY202005011.htm SUN Dong-lei, WANG Yan, YU Yi-xiao. Interval Prediction of short-term regional photovoltaic power based on BP neural network[J]. Journal of Shandong University (English Science), 2020, 50(4): 1-7. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-SDGY202005011.htm
[14]	龙志和, 欧变玲. Bootstrap 方法在经济计量领域的应用[J]. 工业技术经济, 2008, 27(7): 132-135. https://www.cnki.com.cn/Article/CJFDTOTAL-GHZJ200807038.htm LONG Zhi-he, OU Bian-ling. Bootstrap method in econometric applications[J]. Industrial Technology and Economy, 2008, 27(7): 132-135. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-GHZJ200807038.htm
[15]	PRADHAN B. A comparative study on the predictive ability of the decision tree, support vector machine and neuro-fuzzy models in landslide susceptibility mapping using GIS[J]. Computers & Geosciences, 2013, 51: 350-365. doi: 10.3969/j.issn.1001-3695.2013.02.007
[16]	郭卫新, 杨继华, 齐三红, 等. 花岗岩地层双护盾TBM 卡机原因分析及处理措施[J]. 资源环境与工程, 2017, 31(5): 610-613. https://www.cnki.com.cn/Article/CJFDTOTAL-HBDK201705022.htm GUO Wei-xin, YANG Ji-hua, QI San-hong. Cause analysis and treatment measures of double shield TBM blocked in granite stratum[J]. Resources Environment & Engineering, 2017, 31(5): 610-613. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-HBDK201705022.htm
[17]	刘泉声, 刘建平, 潘玉丛, 等. 硬岩隧道掘进机性能预测模型研究进展[J]. 岩石力学与工程学报, 2016, 35(增刊1): 2767-2786. https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX2016S1021.htm LIU Quan-sheng, LIU Jian-ping, PAN Yu-cong, et al. Research advances of tunnel boring machine performance prediction models for hard rock[J]. Chinese Journal of Rock Mechanics and Engineering, 2016, 35(S1): 2767-2786. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX2016S1021.htm
[18]	杨媛媛, 黄宏伟. 围岩分类在TBM滚刀寿命预测中的应用[J]. 地下空间与工程学报, 2005, 1(5): 721-724. https://www.cnki.com.cn/Article/CJFDTOTAL-BASE200505015.htm YANG Yuan-yuan, HUANG Hong-wei. Application of rock mass classification in cutter life prediction of TBM[J]. Chinese Journal of Underground Space and Engineering, 2005, 1(5): 721-724. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-BASE200505015.htm
[19]	ZIO E. A study of the bootstrap method for estimating the accuracy of artificial neural networks in predicting nuclear transient processes[J]. IEEE Transactions on Nuclear Science, 2006, 53(3): 1460-1478. doi: 10.1109/TNS.2006.871662

施引文献(24)

期刊类型引用(13)

1.	满轲，柳宗旭，商艳，宋志飞，刘晓丽，苏宝. 基于灰色关联分析下深度学习盾构姿态预测模型. 工程科学与技术. 2025(02): 203-213 . 百度学术
2.	满轲，曹子祥，刘晓丽，宋志飞，柳宗旭，刘汭琳，武立文. 地质突变条件下基于组合模型的围岩等级和TBM掘进参数预测. 河海大学学报(自然科学版). 2024(01): 55-62 . 百度学术
3.	邓志兴，谢康，李泰灃，苏谦，韩征，肖宪普. 基于机器学习的高铁边坡位移预测不确定性度量与应用. 中国铁道科学. 2024(01): 56-67 . 百度学术
4.	姜浩，郑亚强，金治军，马庆，董强，郑德焰，林峰，赵炳武. 基于ANN算法的钢结构安装工程质量状态评价. 化工管理. 2024(06): 102-109 . 百度学术
5.	郝晶晶，段鹏鑫，陈雨欣，段晓晨. 基于IGWO-SVR的地铁车站投资预测. 铁道学报. 2024(05): 179-188 . 百度学术
6.	张洁. 新疆超长隧洞TBM智能掘进施工方法研究. 黑龙江水利科技. 2024(08): 123-125 . 百度学术
7.	林平，李有鹏，谭彬. 基于Bootstrap方法和LSSVM模型的滑坡位移区间预测. 测绘与空间地理信息. 2024(09): 48-51+56 . 百度学术
8.	张琦，甘超，曹卫华. 大洋钻探过程钻速在线区间预测方法——以微型钻探船室内模拟实验为例. 钻探工程. 2024(05): 45-52 . 百度学术
9.	赵高峰，姜宝元，芮福鑫，马洪素，李洁勇，赵晓豹，龚秋明. 基于数值仿真的复杂岩体TBM掘进性能评估模型. 中南大学学报(自然科学版). 2023(03): 984-997 . 百度学术
10.	常建涛，乔子萱，孔宪光，杨胜康，罗才文. 多维非线性特征重构与融合的复杂产品工期预测方法. 机械工程学报. 2023(06): 294-308 . 百度学术
11.	禹海涛，朱晨阳. 基于BP神经网络的圆形隧道地震响应预测方法及参数分析. 隧道与地下工程灾害防治. 2023(03): 19-26 . 百度学术
12.	彭继慎，郝茗，宋立业，李希桐. 基于TSSA-SVR算法的TBM掘进速度预测. 辽宁工程技术大学学报(自然科学版). 2023(05): 634-640 . 百度学术
13.	闫静，张雪英，李凤莲，陈桂军，黄丽霞. 结合栈式监督AE与可变加权ELM的回归预测模型. 计算机工程. 2022(08): 62-69+76 . 百度学术