作者：陈志恒1, 罗文斌1,2, 常俊杰1,3,4, 钟海鹰1, 吴中权1, 郑阳2
作者单位：1. 南昌航空大学 无损检测技术教育部重点实验室，江西 南昌 330063;
2. 中国特种设备检测研究院，北京 100029;
3. 日本探头株式会社，日本 横滨 232-0033;
4. 浙江清华长三角研究院，浙江 嘉兴 314000
关键词：储油罐;液位检测;兰姆波;EMD;神经网络
摘要：
由于仅从信号时域幅值的大小信息虽然能够判断储油罐中不同介质的液位，但是获得的特征信息非常有限，为获得更多储油罐中不同介质信号的特征信息来提高液面识别率，针对储油罐罐壁厚度为5 mm的钢制储油罐为对象，采用空气耦合超声兰姆波同侧相向检测法，并使用A0模态对储油罐进行检测。利用经验模态分解（EMD）对采集储油罐中的不同介质信号进行EMD处理，求得各阶本征模函数（IMF）。通过分析各阶IMF分量的时域、频域信号与原始信号的相关性，并且以各IMF分量的时域、频域信号为特征值输入到BP神经网络进行决策。实验结果表明，使用该方法能准确地对储油罐不同介质液位在10 mm范围进行识别与分类，识别率可达99%，其检测范围满足实际检测需求。
Neural network air coupled ultrasonic oil tank liquid level detection based on EMD
CHEN Zhiheng1, LUO Wenbin1,2, CHANG Junjie1,3,4, ZHONG Haiying1, WU Zhongquan1, ZHENG Yang2
1. Key Laboratory of Nondestructive Testing Technology, Ministry of Education, Nanchang Hangkong University, Nanchang 330063, China;
2. China Special Equipment Inspection and Research Institute, Beijing 100029, China;
3. Nippon Probe Co., Ltd., Yokohama 232-0033, Japan;
4. Yangtze Delta Region Institute of Tsinghua University, Jiaxing 314000, China
Abstract: Since only the signal amplitude information in the time domain can determine the liquid level of different media in the storage tank, the obtained characteristic information is very limited. In order to obtain more characteristic information of signals of different media in the storage tank to improve the liquid level recognition rate. As for the steel oil tanks with wall thickness of 5 mm, the ipsilateral detection method of air-coupled ultrasonic Lamb wave was adopted, and the A0 mode was used to detect the oil tanks. Empirical mode decomposition (EMD) was used to process signals from different media in the collected storage tank, and IMF was obtained. By analyzing the IMF component of each order time domain and frequency domain signal and original signal correlation, and with each IMF component characteristics of time domain and frequency domain signal input to the BP neural network to make decisions, the experimental results show that using this method can accurate liquid level in storage tanks of different media in the range of 10 mm for identification and classification, recognition rate can reach 99%, the detection range in the results meet the practical needs.
Keywords: oil tank;liquid level detection;Lamb wave;empirical mode decomposition;neural network
2021, 47(1):9-14 收稿日期: 2020-06-29;收到修改稿日期: 2020-07-30
基金项目: 国家自然科学基金 (11464030)；南昌航空大学研究生创新专项资金项目 (YC2019050)
作者简介: 陈志恒(1997-),男,江西抚州市人,硕士研究生,专业方向无损检测
参考文献
[1] 童鑫, 韩成成, 卢翌. 汽车油箱液位测量技术研究[J]. 电子科技, 2012, 25(8): 144
[2] ZHANG L, SHI Z, WU J. A hierarchical oil tank detector with deep surrounding features for high-resolution optical satellite imagery[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2015, 8(10): 1-15
[3] 张建杭. 简析汽车油箱液位测量技术[J]. 电子世界, 2014(9): 26-27
[4] ZHAO W, YANG H, SHEN Z, et al. Oil tanks extraction from high resolution imagery using a directional and weighted hough voting method[J]. Journal of the Indian Society of Remote Sensing, 2015, 43(3): 539-549
[5] WANG L, CHEN G. Dynamic distribution of electrostatic charge and potential in oil tanks during filling process[J]. IEEE Transactions on Dielectrics and Electrical Insulation, 2017, 24(5): 2791-2799
[6] 葛亮, 苗瑞, 葛秋原, 等. 窄搭接焊缝涡流信号的特征提取与缺陷识别[J]. 中国机械工程, 2019, 30(2): 225-229
[7] 李茂, 杨录, 张艳花. 基于EMD及主成分分析的缺陷超声信号特征提取研究[J]. 中国测试, 2018, 44(2): 118-121
[8] 戴波, 赵晶, 周炎. 超声波管道内检测腐蚀缺陷分类识别研究[J]. 机床与液压, 2008, 36(7): 194-198
[9] 郑善朴, 陆铭慧, 王俊东, 等. 多层结构中脱粘缺陷的超声检测方法[J]. 应用声学, 2019, 38(1): 134-141
[10] 李娟, 杨录, 张艳花. 基于EMD与欧氏距离的缺陷信号识别方法[J]. 山西电子技术, 2011(2): 35-36
[11] 郭天太, 陆屹立,王 雷,等. 二维投影式光学传感器轴类零件测量研究[J]. 中国测试, 2020, 46(1): 24-28
[12] 刘贵栋, 沈毅. 基于Hilbert-Huang变换的医学超声信号去噪[J]. 中国医学物理学杂志, 2007, 24(5): 377-380
[13] 孙凌逸, 黄先祥, 蔡伟, 等. 基于经验模态分解的超声波管外测压信号去噪[J]. 声学技术, 2011, 30(5): 427-431
[14] COHEN L. Time-frequency analysis: theory and applications[M]. Prentice-Hall, Englewood Cliffs, New Jersey, 1994.
[15] HUANG N E, SHEN Z, LONG S R, et al. The empirical mode decom-position and the Hilbert spectrum for nonlinear and non-stationary time series analysis[J]. Proc. R. Soc. Lond (A), 1998, 454(1): 903-955
