孪生神经网络在航天产品电性能测试方面的应用

doi:10.7527/S1000-6893.2022.27048

信息融合

本期目录 | 过刊浏览 | 高级检索

前一篇 | 后一篇

孪生神经网络在航天产品电性能测试方面的应用

王宇斐, 刘骁佳, 刘欢, 曹立俊, 罗志强

1. 上海航天精密机械研究所，上海 201600

收稿日期:2022-02-16 修回日期:2022-02-17 发布日期:2022-03-30
通讯作者: 刘骁佳,E-mail:lxj9039@126.com E-mail:lxj9039@126.com
基金资助:
国防基础科研项目（JCKY2019203C017）

Application of Siamese neural network in electrical performance test of aerospace products

WANG Yufei, LIU Xiaojia, LIU Huan, CAO Lijun, LUO Zhiqiang

1. Shanghai Spaceflight Precision Machinery Institute, Shanghai 201600, China

Received:2022-02-16 Revised:2022-02-17 Published:2022-03-30
Supported by:
Defense Industrial Technology Development Program （JCKY2019203C017）

摘要/Abstract

摘要： 航天产品电性能测试是在外加激励的情况下，测试航天产品不同部位的信号反馈情况，从而确定各部件的质量情况，是检验航天产品质量的重要手段之一。但是，测试得到的信号通道多、时序长、十分复杂，依赖传统信号处理方法难度较高，目前处理方式采用人工统计，效率低、易出错，亟需借助深度神经网络的方法来提高效率和准确率。本文将孪生神经网络引入航天产品电性能测试领域，将数据信息解析为图像信息，转化为图像识别问题；对几种主干网络模型进行对比分析，最终选择ResNet50作为主干网络。实验结果表明，在相同实验条件下，研究的算法能够有效地识别78个通道信号的波形信息，识别准确率达到90.385%，提高航天产品复杂测试信号的识别能力以及智能化水平。

关键词: 电性能测试, 复杂信号, 图像识别, 孪生神经网络, 主干网络

Abstract: The electrical performance test of aerospace products is to test the signal feedback of different parts of aerospace products with external excitation， so as to determine the quality of each component. It is one of the important means to test the quality of aerospace products. However， the signal obtained from the test has the characteristics of many channels， long timing and complexity， and it is difficult to rely on traditional methods to process the signal. At present， the processing method adopts manual statistics， which is inefficient and error prone. It is urgent to improve the processing efficiency and accuracy with the help of deep neural network. In this paper， the Siamese neural network is introduced in electrical performance test of aerospace products， and the data information is resolved into image information and transformed into image recognition problem. Several backbone network models are compared and analyzed， and ResNet50 is finally selected as the backbone network. The experimental results show that under the same experimental conditions， the algorithm proposed can effectively identify the waveform information of 78 channel signals and the recognition accuracy reaches 90.385%， improving the recognition ability and intelligent level of complex test signals of aerospace products.

Key words: electrical performance test, complex signal, image recognition, Siamese neural network, backbone network

中图分类号:

王宇斐, 刘骁佳, 刘欢, 曹立俊, 罗志强. 孪生神经网络在航天产品电性能测试方面的应用[J]. 航空学报, 2022, 43(S1): 727048-727048.

WANG Yufei, LIU Xiaojia, LIU Huan, CAO Lijun, LUO Zhiqiang. Application of Siamese neural network in electrical performance test of aerospace products[J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2022, 43(S1): 727048-727048.

参考文献

[1] CHEN G, TENG Q Z. Robust infrared target tracking algorithm based on Siamese network[J]. Computer Applications and Software, 2022, 39(1): 195-199 (in Chinese). 陈果, 滕奇志. 基于孪生网络的鲁棒红外目标跟踪算法[J]. 计算机应用与软件, 2022, 39(1): 195-199.
[2] ZHANG X N, GUO Q L, LIU S Y. Analysis and prospect of deep learning technology and its fault diagnosis application[J]. Journal of Xi'an Jiaotong University, 2020, 54(12): 1-13 (in Chinese). 张西宁, 郭清林, 刘书语. 深度学习技术及其故障诊断应用分析与展望[J]. 西安交通大学学报, 2020, 54(12): 1-13.
[3] WEN L, LI X Y, GAO L, et al. A new convolutional neural network-based data-driven fault diagnosis method[J]. IEEE Transactions on Industrial Electronics, 2018, 65(7): 5990-5998.
[4] SHAO S Y, MCALEER S, YAN R Q, et al. Highly accurate machine fault diagnosis using deep transfer learning[J]. IEEE Transactions on Industrial Informatics, 2019, 15(4): 2446-2455.
[5] AKERET J, CHANG C, LUCCHI A, et al. Radio frequency interference mitigation using deep convolutional neural networks[J]. Astronomy and Computing, 2017, 18: 35-39.
[6] ZHOU X, HE X X, ZHENG C W. Radio signal recognition based on image deep learning[J]. Journal on Communications, 2019, 40(7): 114-125 (in Chinese). 周鑫, 何晓新, 郑昌文. 基于图像深度学习的无线电信号识别[J]. 通信学报, 2019, 40(7): 114-125.
[7] LIU H, ZHOU J Z, ZHENG Y, et al. Fault diagnosis of rolling bearings with recurrent neural network-based autoencoders[J]. ISA Transactions, 2018, 77: 167-178.
[8]
[9] CHI Y W, YANG S X, JIAO W D. A multi-label fault classification method for rolling bearing based on LSTM-RNN[J]. Journal of Vibration, Measurement & Diagnosis, 2020, 40(3): 563-571, 629 (in Chinese). 池永为, 杨世锡, 焦卫东. 基于LSTM-RNN的滚动轴承故障多标签分类方法[J]. 振动测试与诊断, 2020, 40(3): 563-571, 629.
[10]
[11] LIU X N, ZHOU Y, ZHAO J Q, et al. Siamese convolutional neural networks for remote sensing scene classification[J]. IEEE Geoscience and Remote Sensing Letters, 2019, 16(8): 1200-1204.
[12]
[13] DANG Y, ZHANG J X, DENG K Z, et al. Study on the evaluation of land cover classification using remote sensing images based on AlexNet[J]. Journal of Geo-Information Science, 2017, 19(11): 1530-1537 (in Chinese). 党宇, 张继贤, 邓喀中, 等. 基于深度学习AlexNet的遥感影像地表覆盖分类评价研究[J]. 地球信息科学学报, 2017, 19(11): 1530-1537.
[14] ZHU B C, ZHANG M F, YU S, et al. Research on vehicle classification method based on improved AlexNet[J]. Journal of Physics: Conference Series, 2021, 1955(1): 012060.
[15] LIAO L M, ZHANG W. Batch mixed training traffic sign recognition based on improved VGG16 network[J]. Electronic Science and Technology, 2021, 34(8): 8-13 (in Chinese). 廖璐明, 张伟. 基于改进VGG16网络的混合批量训练交通标志识别[J]. 电子科技, 2021, 34(8): 8-13.
[16] PARDEDE J, SITOHANG B, AKBAR S, et al. Implementation of transfer learning using VGG16 on fruit ripeness detection[J]. International Journal of Intelligent Systems and Applications, 2021, 13(2): 52-61.
[17] SONG F Y, WU L M, ZHENG G Z, et al. Optimization of structural pruning based on MobileNetV3[J]. Automation & Information Engineering, 2019, 40(6): 20-25 (in Chinese). 宋非洋, 吴黎明, 郑耿哲, 等. 基于MobileNetV3的结构性剪枝优化[J]. 自动化与信息工程, 2019, 40(6): 20-25.
[18] BAI T B, GAO J L, YANG J W, et al. A study on railway surface defects detection based on machine vision[J]. Entropy, 2021, 23(11): 1437.
[19] CHANG J, GUAN S Q, SHI H Y, et al. Strip defect classification based on improved generative adversarial networks and MobileNetV3[J]. Laser & Optoelectronics Progress, 2021, 58(4): 221-226 (in Chinese). 常江, 管声启, 师红宇, 等. 基于改进生成对抗网络和MobileNetV3的带钢缺陷分类[J]. 激光与光电子学进展, 2021, 58(4): 221-226.
[20] LI Z M. Practice of gesture recognition based on Resnet50[J]. Journal of Physics: Conference Series, 2020, 1574(1): 012154.

E-mail：hkxb@buaa.edu.cn

关于我们

期刊社服务

专业学科

封面文章

友情链接

主管单位：中国科学技术协会主办单位：中国航空学会北京航空航天大学

孪生神经网络在航天产品电性能测试方面的应用

Application of Siamese neural network in electrical performance test of aerospace products

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 4

编辑推荐

Metrics

本文评价

[1]	徐贵力, 徐静, 王彪, 田裕鹏, 郭瑞鹏, 吕温. 基于光照模糊相似融合不变矩的航天器目标识别[J]. 航空学报, 2014, 35(3): 857-867.
[2]	徐贵力, 钟志伟, 王彪, 田裕鹏, 郭瑞鹏, 李开宇. 具有几何和光照不变性的不变矩构建研究[J]. 航空学报, 2013, 34(7): 1698-1705.
[3]	程咏梅;潘泉;张洪才;王刚. 信息融合图像识别算法及其在三维飞机图像识别中的应用研究[J]. 航空学报, 2004, 25(2): 176-179.
[4]	石友成. 一个新的实时光学图像识别系统[J]. 航空学报, 1987, 8(5): 251-258.