基于半监督深度学习的雷达收发组件故障诊断

doi:10.12305/j.issn.1001-506X.2023.10.38

摘要/Abstract

摘要：

新一代相控阵雷达针对T/R组件部署了大量传感器, 为数据驱动的组件故障诊断提供了良好基础。然而, 实际监测数据大多没有表征其故障模式的标签。结合深度置信网络(deep belief network, DBN)在特征自学习方面的优势和自编码器(auto-encoder, AE)重构输入数据的能力, 提出一种基于DBN-AE半监督学习模型的故障特征提取及智能诊断方法, 并应用烟花算法优化模型结构。该方法利用原始无标签状态数据训练DBN-AE模型, 提取深层特征, 再通过有监督再训练建立深层特征与故障模式之间的关系模型。所提方法在某型相控阵雷达T/R模块上得到了实验验证, 有效提升了故障识别准确率和智能水准。

关键词: 相控阵雷达, T/R组件, 故障诊断, 深度置信网络, 深度自编码器, 烟花算法

Abstract:

A large number of sensors are deployed in the new generation phased array radar for T/R components, which provides prerequisites for their data-driven fault diagnosis. However, actual detection data lack labels representing specific fault mode. Combining the advantages of deep belief network (DBN) in feature self-learning and the ability of auto-encoder (AE) to reconstruct input data, a fault feature extraction and intelligent diagnosis method based on DBN-AE semi-supervised learning model is proposed, and its structure is optimized by the firework algorithm (FWA). The original unlabeled data is directly used to train the DBN-AE model to extract deep feature, then the relation model between deep features and fault modes is built through supervised retraining. The proposed method is verified to enhance the fault diagnosis accuracy and intelligence of the T/R module on a certain phased array radar.

Key words: phased array radar, T/R module, fault diagnosis, deep belief network (DBN), deep auto-encoder (AE), firework algorithm (FWA)

中图分类号:

TP277

陈毓坤, 于晖, 陆宁云. 基于半监督深度学习的雷达收发组件故障诊断[J]. 系统工程与电子技术, 2023, 45(10): 3329-3337.

Yukun CHEN, Hui YU, Ningyun LU. Fault diagnosis of radar T/R module based on semi-supervised deep learning[J]. Systems Engineering and Electronics, 2023, 45(10): 3329-3337.

图/表 17

图1

图2

表1

图3

图4

图5

图6

表2

表3

表4

图7

表5

图8

图9

图10

图11

表6

参考文献 34

1	CHEN Y K, HU X K, YU H, et al. A health assessment method for radar antenna array system[C]//Proc. of the 41st Chinese Control Conference, 2022: 4028-4033.
2	ZHAO R , YAN R Q , CHEN Z H , et al. Deep learning and its applications to machine health monitoring[J]. Mechanical Systems and Signal Processing, 2019, 115, 213- 237. doi: 10.1016/j.ymssp.2018.05.050
3	KHAN S , YAIRI T . A review on the application of deep learning in system health management[J]. Mechanical Systems and Signal Processing, 2018, 107 (1): 241- 265.
4	文成林, 吕菲亚. 基于深度学习的故障诊断方法综述[J]. 电子与信息学报, 2020, 42 (1): 234- 248.
	WEN C L , LV F Y . Review on deep learning based fault diagnosis[J]. Journal of Electronics & Information Technology, 2020, 42 (1): 234- 248.
5	BERRY M W , MOHAMED A H , YAP B W . Supervised and unsupervised learning for data science[M]. Cham, Switzerland: Springer, 2010: 3- 21.
6	殷瑞刚, 魏帅, 李晗, 等. 深度学习中的无监督学习方法综述[J]. 计算机系统应用, 2016, 25 (8): 1- 7. doi: 10.15888/j.cnki.csa.005283
	YIN R G , WEI S , LI H , et al. Introduction of unsupervised learning methods in deep learning[J]. Computer Science and Applications, 2016, 25 (8): 1- 7. doi: 10.15888/j.cnki.csa.005283
7	VANE J E , HOOS H H . A survey on semi-supervised learning[J]. Machine Learning, 2020, 109 (2): 373- 440. doi: 10.1007/s10994-019-05855-6
8	ZHAO M H , MYEONGSU K , TANG B P , et al. Multiple wavelet coefficients fusion in deep residual networks for fault diagnosis[J]. IEEE Trans.on Industrial Electronics, 2019, 66 (6): 4696- 4706. doi: 10.1109/TIE.2018.2866050
9	ZHENG J D , CAO S J , PAN H Y , et al. Spectral envelope-based adaptive empirical Fourier decomposition method and its application to rolling bearing fault diagnosis[J]. ISA Transactions, 2022, 129 (Part B): 476- 492.
10	FU Q , JING B , HE P J , et al. Fault feature selection and diagnosis of rolling bearings based on EEMD and optimized Elman_AdaBoost algorithm[J]. IEEE Sensors Journal, 2018, 18 (12): 5024- 5034. doi: 10.1109/JSEN.2018.2830109
11	HAN D Y , GUO X C , SHI P M . An intelligent fault diagnosis method of variable condition gearbox based on improved DBN combined with WPEE and MPE[J]. IEEE Access, 2020, 8, 131299- 131309. doi: 10.1109/ACCESS.2020.3008208
12	单外平, 曾雪琼. 基于深度信念网络的信号重构与轴承故障识别[J]. 电子设计工程, 2016, 24 (4): 67- 71. doi: 10.3969/j.issn.1674-6236.2016.04.023
	SHAN W P , ZENG X Q . Signal reconstruction and bearing fault identification based on deep belief network[J]. Electronic Design Engineering, 2016, 24 (4): 67- 71. doi: 10.3969/j.issn.1674-6236.2016.04.023
13	GAJJAR S , KULAHCI M , PALAZOGLU A . Real-time fault detection and diagnosis using sparse principal component analysis[J]. Journal of Process Control, 2018, 67, 112- 128. doi: 10.1016/j.jprocont.2017.03.005
14	LIU S Y , DONG L , LIAO X Z , et al. Photovoltaic array fault diagnosis based on Gaussian kernel fuzzy c-means clustering algorithm[J]. Sensors, 2019, 19 (7): 1520. doi: 10.3390/s19071520
15	YANG J , YANG Y X , XIE G . Diagnosis of incipient fault based on sliding-scale resampling strategy and improved deep autoencoder[J]. IEEE Sensors Journal, 2020, 20 (15): 8336- 8348. doi: 10.1109/JSEN.2020.2976523
16	WU X J , XU M D , LI C D , et al. Research on image reconstruction algorithms based on autoencoder neural network of restricted Boltzmann machine[J]. Flow Measurement and Instrumentation, 2021, 80, 102009. doi: 10.1016/j.flowmeasinst.2021.102009
17	PENG K X , JIAO R H , DONG J , et al. A deep belief network based health indicator construction and remaining useful life prediction using improved particle filter[J]. Neurocomputing, 2019, 361 (C): 19- 28.
18	郭方洪, 易新伟, 徐博文, 等. 基于深度信念网络和迁移学习的隐匿FDI攻击入侵检测[J]. 控制与决策, 2022, 37 (4): 913- 921.
	GUO F H , YI X W , XU B W , et al. Stealthy FDI attack detection based on deep belief network and transfer learning[J]. Control and Decision, 2022, 37 (4): 913- 921.
19	WU X Y , ZHANG Y , CHENG C M , et al. A hybrid classification autoencoder for semi-supervised fault diagnosis in rotating machinery[J]. Mechanical Systems and Signal Processing, 2021, 149, 107327. doi: 10.1016/j.ymssp.2020.107327
20	ZHONG S S , FU S , LIN L . A novel gas turbine fault diagnosis method based on transfer learning with CNN[J]. Measurement, 2019, 137, 435- 453. doi: 10.1016/j.measurement.2019.01.022
21	ELLEFSEN A L , BJORLYKHAUG E , AESOY V , et al. Remaining useful life predictions for turbofan engine degradation using semi-supervised deep architecture[J]. Reliability Engineering & System Safety, 2019, 183, 240- 251.
22	LIAO Y X , HUANG R Y , LI J P , et al. Deep semi-supervised domain generalization network for rotary machinery fault diagnosis under variable speed[J]. IEEE Trans.on Instrumentation and Measurement, 2020, 69 (10): 8064- 8075.
23	SCHVARTZMAN D , TORRES S M , YU T Y . Distributed beams: concept of operations for polarimetric rotating phased array radar[J]. IEEE Trans.on Geoscience and Remote Sensing, 2021, 59 (11): 9173- 9191. doi: 10.1109/TGRS.2020.3047090
24	葛建军, 张春城. 数字阵列雷达[M]. 北京: 国防工业出版社, 2017.
	GE J J , ZHANG C C . Digital array radar[M]. Beijing: National Defense Industry Press, 2017.
25	DING L F , GENG F L , CHEN J C . Radar principles[M]. Beijing: Publishing House of Electronics, 2014: 273- 285.
26	HINTON G E , SALAKHUTDINOV R R . Reducing the dimensionality of data with neural networks[J]. Science, 2006, 313 (5786): 504- 507. doi: 10.1126/science.1127647
27	FISCHER A , IGEL C . An introduction to restricted boltzmann machines[M]. Heidelberg: Springer Berlin, 2012: 14- 36.
28	HINTON G E , OSINDERO S , TEH Y W . A fast learning algorithm for deep belief nets[J]. Neural Computation, 2006, 18 (7): 1527- 1554. doi: 10.1162/neco.2006.18.7.1527
29	CHEN Y S , LIN Z H , XING Z , et al. Deep learning-based classification of hyperspectral data[J]. IEEE Journal of Selected Topics in Applied Earth Observations & Remote Sensing, 2014, 7 (6): 2094- 2107.
30	BENGIO Y, LAMBLIN P, POPOVICI D, et al. Greedy layer-wise training of deep networks[C]//Proc. of the 19th International Conference on Neural Information Processing Systems, 2006: 153-160.
31	TAN Y . Fireworks algorithm[M]. Heidelberg: Springer eBook, 2015: 25- 35.
32	MOUNA G M , SADOK B . A new multi-objective firework algorithm to solve the multimodal planning network problem[J]. International Journal of Applied Metaheuristic Computing, 2020, 11 (4): 91- 113.
33	WANG W D , LIU K J , YANG C , et al. Cyber physical energy optimization control design for PHEVs based on enhanced firework algorithm[J]. IEEE Trans.on Vehicular Technology, 2021, 70 (1): 282- 291.
34	LUO H , HE C , ZHOU J N , et al. Rolling bearing sub-health recognition via extreme learning machine based on deep belief network optimized by improved fireworks[J]. IEEE Access, 2021, 9, 42013- 42026.

故障模式	故障表现
电源欠压	电源电压偏低, 发射功放输出功率低
电源过压	电源电压偏高, 发射功放输出功率高
发射通道故障	发射功放输出功率下降或为零
接收通道故障	接收信号功率下降或无接收信号
发射和接收通道故障	发射和接收通道监测幅度均异常

序号	状态量名称	变量描述	单位
1~8	TxMODULE_OUTPOWER1~8	发射功放输出功率	dB
9~16	DAM_CHANNEL_REC1~8	接收通道接收信号	dB
17~24	DAM_CHANNEL_NOISE1~8	接收通道噪声信号	dB
25	RT_PS_V	收发电源电压	V
26	RT_PS_C	收发电源电流	A

参数	数值	参数	数值
学习率	0.1	预训练迭代次数	20
初始权重	随机值	编码部分隐藏层数	4
初始偏置	0	解码部分隐藏层数	4
初始动量	0.5	输入层单元数	26
稳态动量	0.9	输出层单元数	26
权重衰减	0.000 2	反向微调迭代次数	500

参数	数值	参数	数值
初始烟花数	10	迭代次数	50
爆炸半径调节常数	2	变量维数	4
爆炸数目调节常数	6	变量上界	50
变异火花数	5	变量下界	1

方法	训练误差	测试误差	时间/s
本文算法	0.053 8	0.059 2	176.39
PSO-DBN-AE	0.055 6	0.061 6	180.55
未优化DBN-AE	0.059 5	0.066 2	146.08
单一DAE	0.071 2	0.076 3	133.45