基于Faster R-CNN的海底管道智能检测方法

doi:10.16265/j.cnki.issn1003-3033.2023.06.2360

中国安全科学学报 ›› 2023, Vol. 33 ›› Issue (6): 80-87.doi: 10.16265/j.cnki.issn1003-3033.2023.06.2360

基于Faster R-CNN的海底管道智能检测方法

俞进¹(), 唐建华², 神祥凯³, 刘金海³

1 中国海洋石油集团有限公司,北京 100010
2 中海油能源发展装备技术有限公司,天津 300452
3 东北大学信息科学与工程学院,辽宁沈阳 110819

收稿日期:2023-01-29 修回日期:2023-04-15 出版日期:2023-08-07

作者简介:

俞进(1970—),男,江苏泗阳人,博士,高级工程师,主要从事海洋、陆地、海外、常规及非常规油气勘探开发生产管理、特大型央企规划计划管理工作。E-mail: oc.com.cn" yujin@cnooc.com.cn。

基金资助:
国家自然科学基金资助(61973071); 辽宁省自然科学基金资助(2019-KF-03-04)

Intelligent detection method for submarine pipelines based on Faster R-CNN

YU Jin¹(), TANG Jianhua², SHEN Xiangkai³, LIU Jinhai³

1 China National Offshore Oil Corporation, Beijing 100010, China
2 Energy Development Equipment Technology Co., Ltd., China National Offshore Oil Corporation, Tianjin 300452, China
3 School of Information Science and Engineering, Northeastern University, Shenyang Liaoning 110819, China

Received:2023-01-29 Revised:2023-04-15 Published:2023-08-07

摘要/Abstract

摘要：

为提高海底管道缺陷及组件的检测精度并实现智能化海底管道安全检测,提出一种基于快速区域卷积神经网络(Faster R-CNN)的海底管道智能检测方法。首先,通过基值校正和分段映射-伪彩色化方法,将漏磁检测信号转化为伪彩色图,以增强漏磁信号的关键特征;其次,基于多模态数据增强来提升检测模型的泛化能力;然后,基于多模态数据增强后的样本训练改进的Faster R-CNN网络,建立最优的智能检测模型;最后,以试验场和渤海在役管道为例,验证所提方法的有效性。结果表明:所提方法的平均检测精度可达93.8%,相较原始的Faster R-CNN算法提高8%,且平均交并比达到0.75,能够精准地实现海底油气管道多目标检测,保障海底管道的安全运行。

关键词: 快速区域卷积神经网络(Faster R-CNN), 海底管道, 智能检测, 多目标检测

Abstract:

In order to improve the detection accuracy of submarine pipeline defects and components and to realize intelligent submarine pipeline safety detection, a Faster R-CNN-based intelligent detection method for submarine pipelines was proposed. Firstly, the key features of the signals were enhanced by converting the signals into a pseudo-color map through basis value correction and segmentation mapping-pseudo-colorization methods. Secondly, the generalization capability of the detection model was improved based on multimodal data enhancement. Then, the improved Faster R-CNN network was trained based on the samples after multimodal data enhancement to establish the optimal intelligent detection model. Finally, the effectiveness of the proposed method was verified by using the test site and the Bohai Sea in service as examples. The results show that the average detection accuracy of the proposed method can reach 93.8%, which is 8% better than the original Faster R-CNN algorithm, and the average intersection over union reaches 0.75, which can accurately achieve the multi-target detection of submarine oil and gas pipelines and ensure the safe operation of submarine pipelines.

Key words: Faster region-convolutional neural network(Faster R-CNN), submarine pipelines, intelligent detection, magnetic flux leakage internal inspection

俞进, 唐建华, 神祥凯, 刘金海. 基于Faster R-CNN的海底管道智能检测方法[J]. 中国安全科学学报, 2023, 33(6): 80-87.

YU Jin, TANG Jianhua, SHEN Xiangkai, LIU Jinhai. Intelligent detection method for submarine pipelines based on Faster R-CNN[J]. China Safety Science Journal, 2023, 33(6): 80-87.

图/表 15

图1

图2

图3

图4

图5

图6

图7

图8

图9

图10

图11

表1

续表1

表2

图12

参考文献 13

[1]	张新生, 王旭业, 张莹莹, 等. 海底腐蚀管道剩余寿命预测与维修策略研究[J]. 中国安全科学学报, 2022, 32(3):41-47. doi: 10.16265/j.cnki.issn1003-3033.2022.03.006
	ZHANG Xinsheng, WANG Xuye, ZHANG Yingying, et al. Research on remaining life prediction and maintenance strategy of corroded submarine pipelines[J]. China Safety Science Journal, 2022, 32(3):41-47. doi: 10.16265/j.cnki.issn1003-3033.2022.03.006
[2]	刘金海, 赵贺, 神祥凯, 等. 基于漏磁内检测的自监督缺陷检测方法[J]. 仪器仪表学报, 2020, 41(9):180-187.
	LIU Jinhai, ZHAO He, SHEN Xiangkai, et al. Self-supervised defect detection method based on magnetic flux leakage internal detection[J]. Chinese Journal of Scientific Instrument, 2020, 41(9):180-187.
[3]	毕傲睿, 骆正山, 宋莹莹, 等. 内腐蚀海底管道剩余强度的FOA-GRNN模型[J]. 中国安全科学学报, 2020, 30(6):78-83. doi: 10.16265/j.cnki.issn1003-3033.2020.06.012
	BI Aorui, LUO Zhengshan, SONG Yingying, et al. Residual strength analysis of internally corroded submarine pipeline based on FOA-GRNN model[J]. China Safety Science Journal, 2020, 30(6):78-83. doi: 10.16265/j.cnki.issn1003-3033.2020.06.012
[4]	曲杰, 孙玉江, 苑世宁, 等. 某海底管道内腐蚀原因分析及防护[J]. 焊管, 2022, 45(12):42-45.
	QU Jie, SUN Yujiang, YUAN Shining, et al. Cause analysis and protection of internal corrosion in subsea pipeline[J]. Welded Pipe and Tube, 2022, 45(12):42-45.
[5]	卢森骧, 神祥凯, 张俊楠, 等. 基于三轴融合的漏磁内检测数据缺陷反演方法研究[J]. 仪器仪表学报, 2021, 42(12):245-253.
	LU Senxiang, SHEN Xiangkai, ZHANG Junnan, et al. Research on defect inversion method of magnetic flux leakage internal inspection data based on triaxial fusion[J]. Chinese Journal of Scientific Instrument, 2021, 42(12):245-253.
[6]	郝永梅, 杜璋昊, 杨文斌, 等. 基于改进ELMD和多尺度熵的管道泄漏信号识别[J]. 中国安全科学学报, 2019, 29(8):105-111. doi: 10.16265/j.cnki.issn1003-3033.2019.08.017
	HAO Yongmei, DU Zhanghao, YANG Wenbin, et al. Pipeline leakage signal identification based on improved ELMD and multi-scale entropy[J]. China Safety Science Journal, 2019, 29(8):105-111. doi: 10.16265/j.cnki.issn1003-3033.2019.08.017
[7]	王超群, 梁伟, 梁晓斌. CEEMD-FCM模型下的管道缺陷识别方法[J]. 中国安全科学学报, 2020, 30(1):87-93. doi: 10.16265/j.cnki.issn1003-3033.2020.01.014
	WANG Chaoqun, LIANG Wei, LIANG Xiaobin. Pipeline defect recognition method based on CEEMD-FCM[J]. China Safety Science Journal, 2020, 30(1):87-93. doi: 10.16265/j.cnki.issn1003-3033.2020.01.014
[8]	蹇清平, 艾志久, 张勇, 等. 基于支持向量机的油管内外表面缺陷识别方法[J]. 机械科学与技术, 2015, 34(1):118-123.
	JIAN Qingping, AI Zhijiu, ZHANG Yong, et al. Identification of oil tube defects on internal and external surfaces based on SVM[J]. Mechanical Science and Technology for Aerospace Engineering, 2015, 34(1):118-123.
[9]	赵翰学, 张咪, 郭岩宝, 等. 基于机器学习的管道金属损失缺陷识别方法[J]. 石油机械, 2020, 48(12):138-145.
	ZHAO Hanxue, ZHANG Mi, GUO Yanbao, et al. Recognition method of pipeline metal loss defects based on machine learning[J]. China Petroleum Machinery, 2020, 48(12):138-145.
[10]	刘金海, 付明芮, 唐建华. 基于漏磁内检测的缺陷识别方法[J]. 仪器仪表学报, 2016, 37(11):2572-2581.
	LIU Jinhai, FU Mingrui, TANG Jianhua. MFL inner detection based defect recognition method[J]. Chinese Journal of Scientific Instrument, 2016, 37(11):2572-2581.
[11]	王宏安, 陈国明. 基于深度学习的漏磁检测缺陷识别方法[J]. 石油机械, 2020, 48(5):127-132.
	WANG Hongan, CHEN Guoming. Magnetic flux leakage defect detection based on deep learning[J]. China Petroleum Machinery, 2020, 48(5):127-132.
[12]	杨理践, 曹辉. 基于深度学习的管道焊缝法兰组件识别方法[J]. 仪器仪表学报, 2018, 39(2):193-202.
	YANG Lijian, CAO Hui. Deep learning based weld and flange identification in pipeline[J]. Chinese Journal of Scientific Instrument, 2018, 39(2):193-202.
[13]	REN Shaoqing, HE Kaiming, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2017, 39(6):1137-1149.

缺陷	人工标注坐标	Faster R-CNN算法	本文算法	IoU1	IoU2
2	[108,33,155,61]	[112,41,158,64]	[110,37,155,59]	0.57	0.75
3	[178,31,249,64]	[184,36,245,63]	[181,34,251,62]	0.71	0.79

缺陷	人工标注坐标	Faster R-CNN算法	本文算法	IoU1	IoU2
1	[25,32,63,46]	[23,32,64,50]	[25,31,63,48]	0.72	0.82
2	[58,47,97,61]	[61,50,94,64]	[56,47,99,63]	0.56	0.79
3	[83,31,112,43]	[82,33,113,45]	[82,32,114,46]	0.67	0.67
4	[113,30,129,42]	[114,30,126,46]	[117,31,128,42]	0.60	0.63
5	[130,29,145,42]	[127,31,142,43]	[128,31,147,44]	0.54	0.60

基于Faster R-CNN的海底管道智能检测方法

Intelligent detection method for submarine pipelines based on Faster R-CNN

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 15

参考文献 13

相关文章 7

编辑推荐

Metrics

本文评价

[1]	刘富鹏, 杨九, 吴世博, 徐立新. 基于FDHHFLTS-BN的海底管道泄漏失效风险定量分析[J]. 中国安全科学学报, 2024, 34(1): 166-170.
[2]	张新生, 蔡宝泉. 基于改进随机森林模型的海底管道腐蚀预测^*[J]. 中国安全科学学报, 2021, 31(8): 69-74.
[3]	毕傲睿, 骆正山, 宋莹莹, 张新生. 内腐蚀海底管道剩余强度的FOA-GRNN模型[J]. 中国安全科学学报, 2020, 30(6): 78-83.
[4]	雷云, 余建星, 吴朝晖, 陈海成, 刘俊雄. 基于模糊ANP的海底管道失效风险综合评价[J]. 中国安全科学学报, 2019, 29(5): 178-184.
[5]	张新生, 赵梦旭, 王小完. 尾段残差修正GM(1,1)模型在管道腐蚀预测中的应用[J]. 中国安全科学学报, 2017, 27(1): 65-70.
[6]	胡显伟，段梦兰，官耀华. 基于模糊Bow-tie模型的深水海底管道定量风险评价研究[J]. 中国安全科学学报, 2012, 22(3): 128-.
[7]	李毅佳，余建星. 基于FMEA和模糊理论的海底管道建造期质量风险分析[J]. 中国安全科学学报, 2012, 22(1): 112-.