基于LSTM深度学习网络的充填管道运行状态检测

肖柏林; 苗胜军; 夏道洪; 黄华桃; 张靖宇

doi:10.1007/s12613-022-2560-y

Volume 30 Issue 8

Aug. 2023

Turn off MathJax

图(11) / 表(5)

数据统计

计量

文章访问数: 746
HTML全文浏览量: 266
PDF下载量: 44
被引次数: 0

文章导航 > 矿物冶金与材料学报（英文） > 2023 > 30(8): 1573-1583

Bolin Xiao, Shengjun Miao, Daohong Xia, Huatao Huang, and Jingyu Zhang, Detecting the backfill pipeline blockage and leakage through an LSTM-based deep learning model, Int. J. Miner. Metall. Mater., 30(2023), No. 8, pp. 1573-1583. https://doi.org/10.1007/s12613-022-2560-y

Cite this article as:

Bolin Xiao, Shengjun Miao, Daohong Xia, Huatao Huang, and Jingyu Zhang, Detecting the backfill pipeline blockage and leakage through an LSTM-based deep learning model, Int. J. Miner. Metall. Mater., 30(2023), No. 8, pp. 1573-1583. https://doi.org/10.1007/s12613-022-2560-y

Citation:

PDF( 1576 KB)

引用本文 PDF XML SpringerLink

研究论文

基于LSTM深度学习网络的充填管道运行状态检测

1)
北京科技大学金属矿山高效开采与安全教育部重点实验室, 北京 100083, 中国
2)
北京科技大学高等工程师学院, 北京 100083, 中国
3)
北京科技大学土木与资源工程学院, 北京 100083, 中国
4)
云南金诚信矿业管理有限公司, 昆明 650217, 中国
5)
新疆大明矿业集团股份有限公司, 哈密 839121, 中国

通讯作者:
苗胜军 E-mail: miaoshengjun@ustb.edu.cn

文章亮点

(1) 系统研究了高浓度充填料将自流管输工艺中的压力和流量变化规律。

(2) 建立了充填管道运行时泄漏与堵管异常状态的深度学习自动检测模型。

(3) 结合管道堵泄工业试验验证了模型的准确性与先进性。

中文摘要

管道的运行状态，尤其是堵塞和泄漏的检测，对保障充填的连续性和安全性至关重要；粗骨料高浓度自流充填的管网系统具有高度复杂性，鲜有一种专用、高效、精准的管道异常检测方法。本文以金川高浓度自流充填为例，提出了一种基于长短期记忆深度学习模型的管道堵塞和泄漏诊断方法。首先，通过压力和流量传感器构建工业管道的实时监测系统；针对负样本匮乏问题，设计开展堵管、泄漏工业实验；分析不同工作状态下管道的压力流量统计特征；接着，基于长段时记忆网络，建立了管道运行状态检测的深度学习模型，并对模型进行对比检验评价。研究结果表明：充填工艺具有间歇性、周期性、多段性等特征；正常工况内的压力和流量信号整体波动稳定，但不同周期差异较大。堵塞时的区间信号特征突变大，泄漏的区间特征变化持续时间长、波动范围广。将本文提出的LSTM-DL模型与其他4种常用先进方法对比，LSTM-DL模型在所有状态下具有98.31%的最高预测准确率，且在堵塞和泄漏状态下具有3.21％的最低误判或虚警率。本文的LSTM-DL模型可以准确识别复杂高浓度自流充填管道系统中的各种运行状态。

关键词:

Research Article

Detecting the backfill pipeline blockage and leakage through an LSTM-based deep learning model

+ Author Affiliations

1)
State Key Laboratory of High-Efficiency Mining and Metal Mine Safety, Ministry of Education, University of Science and Technology Beijing, Beijing 100083, China
2)
School of Advanced Engineering, University of Science and Technology Beijing, Beijing 100083, China
3)
School of Civil and Resource Engineering, University of Science and Technology Beijing, Beijing 100083, China
4)
Yunnan Jinchengxin Mining Management Co., Ltd., Kunming 650217, China
5)
Xinjiang Daming Mining Group Co., Ltd., Hami 839121, China

The authors declare no potential conflict of interest.

Corresponding author:
Shengjun Miao E-mail: miaoshengjun@ustb.edu.cn
Received: 7 August 2022; Revised: 19 October 2022; Accepted: 20 October 2022; Available online: 25 October 2022

Abstract

Abstract

Detecting a pipeline’s abnormal status, which is typically a blockage and leakage accident, is important for the continuity and safety of mine backfill. The pipeline system for gravity-transport high-density backfill (GHB) is complex. Specifically designed, efficient, and accurate abnormal pipeline detection methods for GHB are rare. This work presents a long short-term memory-based deep learning (LSTM-DL) model for GHB pipeline blockage and leakage diagnosis. First, an industrial pipeline monitoring system was introduced using pressure and flow sensors. Second, blockage and leakage field experiments were designed to solve the problem of negative sample deficiency. The pipeline’s statistical characteristics with different working statuses were analyzed to show their complexity. Third, the architecture of the LSTM-DL model was elaborated on and evaluated. Finally, the LSTM-DL model was compared with state-of-the-art (SOTA) learning algorithms. The results show that the backfilling cycle comprises multiple working phases and is intermittent. Although pressure and flow signals fluctuate stably in a normal cycle, their values are diverse in different cycles. Plugging causes a sudden change in interval signal features; leakage results in long variation duration and a wide fluctuation range. Among the SOTA models, the LSTM-DL model has the highest detection accuracy of 98.31% for all states and the lowest misjudgment or false positive rate of 3.21% for blockage and leakage states. The proposed model can accurately recognize various pipeline statuses of complex GHB systems.
- mine backfill;
- blockage and leakage;
- pipeline detection;
- long short-term memory networks;
- deep learning

FullText(HTML)

Supplements(0)

References(34)

References

[1]	M. Helinski, M. Fahey, and A. Fourie, Behavior of cemented paste backfill in two mine stopes: Measurements and modeling, J. Geotech. Geoenviron. Eng., 137(2011), No. 2, p. 171. doi: 10.1061/(ASCE)GT.1943-5606.0000418
[2]	X. Zhao, A. Fourie, and C.C. Qi, Mechanics and safety issues in tailing-based backfill: A review, Int. J. Miner. Metall. Mater., 27(2020), No. 9, p. 1165. doi: 10.1007/s12613-020-2004-5
[3]	L. Liu, J. Xin, C. Huan, et al., Effect of curing time on the mesoscopic parameters of cemented paste backfill simulated using the particle flow code technique, Int. J. Miner. Metall. Mater., 28(2021), No. 4, p. 590. doi: 10.1007/s12613-020-2007-2
[4]	K. Been, E.T. Brown, and N. Hepworth, Liquefaction potential of paste fill at Neves Corvo Mine, Portugal, Min. Technol., 111(2002), No. 1, p. 47. doi: 10.1179/mnt.2002.111.1.47
[5]	R. Cooke, Design procedure for hydraulic backfill distribution systems, J. South. Afr. Inst. Min. Metall., 101(2001), No. 2, p. 97.
[6]	A.X. Wu, Z.E. Ruan, and J.D. Wang, Rheological behavior of paste in metal mines, Int. J. Miner. Metall. Mater., 29(2022), No. 4, p. 717. doi: 10.1007/s12613-022-2423-6
[7]	H.F. Lu, T. Iseley, S. Behbahani, and L.D. Fu, Leakage detection techniques for oil and gas pipelines: State-of-the-art, Tunnelling Underground Space Technol., 98(2020), art. No. 103249. doi: 10.1016/j.tust.2019.103249
[8]	A. Mostafapour and S. Davoudi, Analysis of leakage in high pressure pipe using acoustic emission method, Appl. Acoust., 74(2013), No. 3, p. 335. doi: 10.1016/j.apacoust.2012.07.012
[9]	S. El-Zahab, E. Mohammed Abdelkader, and T. Zayed, An accelerometer-based leak detection system, Mech. Syst. Signal Process., 108(2018), p. 276. doi: 10.1016/j.ymssp.2018.02.030
[10]	P. Stajanca, S. Chruscicki, T. Homann, S. Seifert, D. Schmidt, and A. Habib, Detection of leak-induced pipeline vibrations using fiber-optic distributed acoustic sensing, Sensors (Basel), 18(2018), No. 9, art. No. 2841.
[11]	D. Zaman, M.K. Tiwari, A.K. Gupta, and D. Sen, A review of leakage detection strategies for pressurised pipeline in steady-state, Eng. Fail. Anal., 109(2020), art. No. 104264. doi: 10.1016/j.engfailanal.2019.104264
[12]	Y.M. Yan, Y.T. Liang, H.R. Zhang, et al., A two-stage optimization method for unmanned aerial vehicle inspection of an oil and gas pipeline network, Pet. Sci., 16(2019), No. 2, p. 458. doi: 10.1007/s12182-019-0305-y
[13]	V.G. Bondur, Aerospace methods and technologies for monitoring oil and gas areas and facilities, Izv. Atmos. Ocean. Phys., 47(2011), No. 9, p. 1007. doi: 10.1134/S0001433811090039
[14]	M.A. Adegboye, W.K. Fung, and A. Karnik, Recent advances in pipeline monitoring and oil leakage detection technologies: Principles and approaches, Sensors (Basel), 19(2019), No. 11, art. No. 2548.
[15]	N. Aziz, S.A.K. Tanoli, and F. Nawaz, A programmable logic controller based remote pipeline monitoring system, Process. Saf. Environ. Prot., 149(2021), p. 894. doi: 10.1016/j.psep.2021.03.045
[16]	L. Yang and Q. Zhao, A novel PPA method for fluid pipeline leak detection based on OPELM and bidirectional LSTM, IEEE Access, 8(2020), p. 107185. doi: 10.1109/ACCESS.2020.3000960
[17]	C.C. Sun, B. Parellada, V. Puig, and G. Cembrano, Leak localization in water distribution networks using pressure and data-driven classifier approach, Water, 12(2019), No. 1, art. No. 54. doi: 10.3390/w12010054
[18]	B. Wang, Y.B. Guo, D.G. Wang, Y.S. Zhang, R.Y. He, and J.Z. Chen, Prediction model of natural gas pipeline crack evolution based on optimized DCNN-LSTM, Mech. Syst. Signal Process., 181(2022), art. No. 109557. doi: 10.1016/j.ymssp.2022.109557
[19]	X. Li, M. Guo, R. Zhang, and G. Chen, A data-driven prediction model for maximum pitting corrosion depth of subsea oil pipelines using SSA-LSTM approach, Ocean. Eng., 261(2022), art. No. 112062. doi: 10.1016/j.oceaneng.2022.112062
[20]	C. Spandonidis, P. Theodoropoulos, F. Giannopoulos, N. Galiatsatos, and A. Petsa, Evaluation of deep learning approaches for oil & gas pipeline leak detection using wireless sensor networks, Eng. Appl. Artif. Intell., 113(2022), art. No. 104890. doi: 10.1016/j.engappai.2022.104890
[21]	H.Y. Cheng, S.C. Wu, X.Q. Zhang, and A.X. Wu, Effect of particle gradation characteristics on yield stress of cemented paste backfill, Int. J. Miner. Metall. Mater., 27(2020), No. 1, p. 10. doi: 10.1007/s12613-019-1865-y
[22]	R.C. Silva, Experimental characterization techniques for solid-liquid slurry flows in pipelines: A review, Processes, 10(2022), No. 3, art. No. 597. doi: 10.3390/pr10030597
[23]	P.A. Gorshkalev, M.D. Chernosvitov, and D.S. Nikitina, Comparing pipelines made of different materials for replacement of old on-site pipelines of Kinel pumping-filtration plant, IOP Conf. Ser.: Mater. Sci. Eng., 687(2019), No. 4, art. No. 044021. doi: 10.1088/1757-899X/687/4/044021
[24]	A. Sherstinsky, Fundamentals of recurrent neural network (RNN) and long short-term memory (LSTM) network, Physica D: Nonlinear Phenom., 404(2020), art. No. 132306. doi: 10.1016/j.physd.2019.132306
[25]	F.A. Gers, J. Schmidhuber, and F. Cummins, Learning to forget: Continual prediction with LSTM, Neural Comput., 12(2000), No. 10, p. 2451. doi: 10.1162/089976600300015015
[26]	C. Garbin, X.Q. Zhu, and O. Marques, Dropout vs. batch normalization: An empirical study of their impact to deep learning, Multimed. Tools Appl., 79(2020), No. 19-20, p. 12777. doi: 10.1007/s11042-019-08453-9
[27]	P. Rodríguez, M.A. Bautista, J. Gonzàlez, and S. Escalera, Beyond one-hot encoding: Lower dimensional target embedding, Image Vis. Comput., 75(2018), p. 21. doi: 10.1016/j.imavis.2018.04.004
[28]	S.M. LaValle, M.S. Branicky, and S.R. Lindemann, On the relationship between classical grid search and probabilistic roadmaps, Int. J. Robotics Res., 23(2004), No. 7-8, p. 673. doi: 10.1177/0278364904045481
[29]	Y. Yao, L. Rosasco, and A. Caponnetto, On early stopping in gradient descent learning, Constr Approx, 26(2007), No. 2, p. 289. doi: 10.1007/s00365-006-0663-2
[30]	B. Xiao, Z. Wen, F. Wu, L. Li, Z. Yang, and Q. Gao, A simple L-shape pipe flow test for practical rheological properties of backfill slurry: A case study, Powder Technol., 356(2019), p. 1008. doi: 10.1016/j.powtec.2019.09.012
[31]	P. Li, Y.B. Hou, and M.F. Cai, Factors influencing the pumpability of unclassified tailings slurry and its interval division, Int. J. Miner. Metall. Mater., 26(2019), No. 4, p. 417. doi: 10.1007/s12613-019-1750-8
[32]	A.X. Wu, H. Jiao, H.J. Wang, et al., Status and development trends of paste disposal technology with ultra-fine unclassified tailings in China, [in] Paste 2011: Proceedings of the 14th International Seminar on Paste and Thickened Tailings, Perth, 2011, p. 477.
[33]	M.M. Saritas and A. Yasar, Performance analysis of ANN and naive Bayes classification algorithm for data classification, Int. J. Intell. Syst. Appl. Eng., 7(2019), No. 2, p. 88. doi: 10.18201/ijisae.2019252786
[34]	C.J. Mantas and J. Abellán, Credal-C4.5: Decision tree based on imprecise probabilities to classify noisy data, Expert Syst. Appl., 41(2014), No. 10, p. 4625. doi: 10.1016/j.eswa.2014.01.017