Research on HRRP Sequences Recognition Based on Discriminative Infinite Fuzzy Restricted Boltzmann Machine Model

doi:10.12382/bgxb.2024.0552

Abstract

Abstract:

Aiming at the problem of poor target recognition of radar high-resolution range profiles (HRRP) sequences in the presence of interference or severe background clutter, a discriminative infinite fuzzy restricted Boltzmann machine (Dis-iFRBM) model is proposed. The proposed model employs a combination of two distinct machine learning techniques: discriminative restricted Boltzmann machine (Dis-RBM) classification and infinite restricted Boltzmann machine (iRBM) model complexity adaptive features. The model parameters are extended from real numbers to fuzzy parameters, drawing on the advantages of the fuzzy neural network in extracting the features more stably in a low signal-to-noise ratio environment. This approach achieves a more stable extraction of the original features of HRRP sequences as well as more robust recognition of radar targets. The recognition stability and robustness of Dis-iFRBM are validated through recognition experiments on several HRRP sequences, and the comparison experiments the proposed model and other models demonstrate the effectiveness and superiority of the proposed model in the "noisy" environment.

Key words: high-resolution range profile, target recognition, discriminative infinite fuzzy restricted Boltzmann machine, random permutation, noisy data

CLC Number:

TJ765

CHEN Shichao, WEI Jingbiao, FAN Jun, WEI Xizhang, WANG Zechao, SUN Qian, LIU Ming. Research on HRRP Sequences Recognition Based on Discriminative Infinite Fuzzy Restricted Boltzmann Machine Model[J]. Acta Armamentarii, 2024, 45(S1): 43-50.

Figures/Tables 12

Fig.1 RBM with n visible and m hidden neurons

Fig.2 Structure of Dis-iFRBM model

Fig.3 Training process of Dis-iFRBM

Fig.4 HRRP sequence data inversion process

Fig.5 Recognition rates of two models under different sequence lengths for 3 types of targets

Fig.6 Recognition rates of two models under different sequence lengths for 10 types of targets

Fig.7 Recognition rates of two models with sequence length of 5 and SNR of 20db

Fig.8 Recognition rates of multiple models with sequence length of 5 and SNR of 20db

Fig.9 Mean and variance of the recognition rates of multiple models after 50 rounds of training

Table 1 Recognition rates of different models

模型	Dis-iFRBM	Dis-iRBM	ClassRBM	SVM
平均识别率	86.01%	81.63%	77.43%	73.81%

Table 2 The variance of average accuracy of different sequence lengths

模型	L_s
模型	5	10	15	20	30
Dis-iRBM	0.0016	0.0016	0.0016	0.0016	0.0016
Dis-iFRBM	0.0003	0.0008	0.0002	0.0003	0.0002

Fig.10 Recognition rates of two models with different SNRs

References 23

[1]	陈士超, 卢福刚, 马菲, 等. 基于观测匹配的毫米波导引头目标选择方法[J]. 战术导弹技术, 2022(3):40-46.
	CHEN S C, LU F G, MA F, et al. A target selection method for the MMW seeker based on tracking preserving[J]. Tactical Missile Technology, 2022(3):40-46. (in Chinese)
[2]	邹汝平, 陈士超, 陈韵, 等. 激光寻的空地导弹制导控制方法[J]. 兵工学报, 2021, 42(1):1-10.
	ZOU R P, CHEN S C, CHEN Y, et al. A guidance control method for air-to-ground missiles with laser seekers[J]. Acta Armamentarii, 2021, 42(1):1-10. (in Chinese) doi: 10.3969/j.issn.1000-1093.2021.01.001
[3]	邹汝平, 陈士超, 陈韵, 等. 一种图像寻的空地导弹的制导控制方法[J]. 战术导弹技术, 2021(1):93-98,114.
	ZOU R P, CHEN S C, CHEN Y, et al. A guidance control method for imaging air-to-ground missiles[J]. Tactical Missile Technology, 2021(1):93-98, 114. (in Chinese)
[4]	宋益恒, 王彦华, 李阳, 等. 基于深度生成网络的雷达HRRP生成技术[J]. 信号处理, 2019, 35(6):1118-1122.
	SONG Y H, WANG Y H, LI Y, et al. Radar HRRP generation using deep generative networks[J]. Journal of Signal Processing, 2019, 35(6):1118-1122. (in Chinese)
[5]	李阳, 刘艺辰, 张亮, 等. 基于语义引导层次化分类的雷达地面目标HRRP识别方法[J]. 信号处理, 2024, 40(1):126-137.
	LI Y, LIU Y C, ZHANG L, et al. Semantic knowledge guided hierarchical classification for radar ground target HRRP recognition[J]. Journal of Signal Processing, 2024, 40(1):126-137. (in Chinese)
[6]	LAN D, WANG P H, LIU H W, et al. Bayesian spatiotemporal multitask learning for radar HRRP target recognition[J]. IEEE Transactions on Signal Processing, 2011, 59(7): 3182-3196.
[7]	陈健, 杜兰, 廖磊瑶. 基于参数化统计模型的雷达HRRP目标识别方法综述[J]. 雷达学报, 2022, 11(6):1020-1047.
	CHEN J, DU L, LIAO L Y. Survey of radar HRRP target recognition based on parametric statistical model[J]. Journal of Radars, 2022, 11(6): 1020-1047. (in Chinese)
[8]	赵东波, 李辉. 粒子群优化的SVM在雷达目标识别中的应用[J]. 信息技术, 2017(10):141-145.
	ZHAO D B, LI H. Application of support vector machine based on particle swarm optimization in radar target recognition[J]. Information Technology, 2017(10):141-145. (in Chinese)
[9]	DU L, WANG P H, LIU H W, et al. Radar HRRP target recognition based on dynamic multi-task hidden Markov model[C]// Proceedings of the IEEE National Radar Conference Proceedings. Kansas City, MO, US: IEEE, 2011:253-255.
[10]	云涛, 潘泉, 郝宇航, 等. 基于HRRP时频特征和多尺度非对称卷积神经网络的目标识别算法[J]. 西北工业大学学报, 2023, 41(3):537-545.
	YUN T, PAN Q, HAO Y H, et al. Target recognition algorithm base on HRRP time-spectrogram feature and multiscale asymmetric convolutional neural network[J]. Journal of Northwestern Polytechnical University, 2023, 41(3):537-545. (in Chinese)
[11]	XING Z J, WEI G F, BO D, et al. Ship HRRP target recognition method based on transformer and convolutional attention network[C]// Proceedings of the fifteenth International Conference on Signal Processing Systems (ICSPS 2023). Xi’an, China: SPIE, 2024: 558-567.
[12]	MA M Y, LIU K, LUO X, et al. Multi-view polarization HRRP target recognition based on convolutional neural network[J]. IoT as a Service, 2021, 346(1):710-720.
[13]	LIU J Q, CHEN B, CHEN W C, et al. Radar HRRP target recognition with target aware two-dimensional recurrent neural network[C]// Proceedings of the 2019 IEEE International Conference on Signal Processing. Communications and Computing Dalian, China: IEEE, 2019: 1-6.
[14]	徐彬, 陈渤, 刘家麒, 等. 采用双向 LSTM 模型的雷达HRRP目标识别[J]. 西安电子科技大学学报, 2019, 46(2): 29-34.
	XU B, CHEN B, LIU J L, et al. Radar HRRP target recognition by the bidirectional LSTM model[J]. Journal of Xidian University, 2019, 46(2): 29-34. (in Chinese)
[15]	CHEN C L P, ZHANG C, CHEN L, et al. Fuzzy restricted Boltzmann machine for the enhancement of deep learning[J]. IEEE Transactions on Fuzzy Systems, 2015, 23(6):2163-2173.
[16]	FENG S, CHEN C L P. A fuzzy restricted Boltzmann machine: novel learning algorithms based on the crisp possibilistic mean value of fuzzy numbers[J]. IEEE Transactions on Fuzzy Systems, 2018, 26(1):117-130.
[17]	CHEN C L P, FENG S. Generative and discriminative fuzzy restricted Boltzmann machine learning for text and image classification[J]. IEEE Transactions on Cybernetics, 2020, 50(5):2237-2248. doi: 10.1109/TCYB.2018.2869902 pmid: 30295638
[18]	RUSLAN S, ANDRIY M, GEOFFREY H. Restricted Boltzmann machines for collaborative filtering[C]// Proceedings of the International Conference on Machine Learning: ACM, 2007.Corvalis, Oregon, US: ACM, 2007: 791-798.
[19]	LE ROUX N, BENGIO Y. Representational power of restricted Boltzmann machines and deep belief networks[J]. Neural computation, 2008, 20(6): 1631-1649. doi: 10.1162/neco.2008.04-07-510 pmid: 18254699
[20]	CÔTÉ M A, LAROCHELLE H. An infinite restricted Boltzmann machine[J]. Neural Computation, 2016, 28(7): 1265-1288.
[21]	LAROCHELLE H, MANDEL M, PASCANU R, et al. Learning algorithms for the classification restricted Boltzmann machine[J]. Journal of Machine Learning Research, 2012, 13(1):643-669.
[22]	PENG X, GAO X Z, LI X. On better training the infinite restricted Boltzmann machines[J]. Machine Learning, 2018, 107(6):943-968.
[23]	GUDNASON J, CUI J J, BROOKES M. HRR automatic target recognition from superresolution scattering center features[J]. IEEE Transactions on Aerospace and Electronic Systems, 2009, 45(4):1512-1524.