基于t分布擴(kuò)展概率主成分分析模型的一維距離像識(shí)別方法

李彬; 李輝; 郭淞云

doi:10.11999/JEIT161220

基于t分布擴(kuò)展概率主成分分析模型的一維距離像識(shí)別方法

doi: 10.11999/JEIT161220 cstr: 32379.14.JEIT161220

基金項(xiàng)目:

國(guó)家自然科學(xué)基金(61571364)，西北工業(yè)大學(xué)研究生創(chuàng)意創(chuàng)新種子基金(Z2017022)

計(jì)量
- 文章訪問(wèn)數(shù): 1315
- HTML全文瀏覽量: 184
- PDF下載量: 271
- 被引次數(shù): 0
出版歷程
- 收稿日期: 2016-11-10
- 修回日期: 2017-04-06
- 刊出日期: 2017-08-19

Using t-distribution Based Probabilistic Principal Component Analysis Model for High Resolution Range Profile Recognition

Funds:

The National Natural Science Foundation of China (61571364), The Seed Foundation of Innovation and Creation for Graduate Students in Northwestern Poly-technical University (Z2017022)

摘要

摘要: 該文針對(duì)概率主成分分析(PPCA)模型用于1維高分辨距離像(HRRP)識(shí)別對(duì)噪聲敏感的問(wèn)題，對(duì)經(jīng)典PPCA模型進(jìn)行修正。該方法將基于高斯分布的PPCA模型擴(kuò)展為基于t分布的PPCA模型，能夠綜合利用t分布對(duì)噪聲穩(wěn)健和PPCA模型自由參數(shù)少的特性。同時(shí)為了減少目標(biāo)方位敏感性對(duì)HRRP統(tǒng)計(jì)建模的影響，進(jìn)一步將t分布模型擴(kuò)展為混合概率t分布模型，能夠以分布趨同的原則將不同方位幀內(nèi)具有相同統(tǒng)計(jì)特性的HRRP數(shù)據(jù)進(jìn)行聚類，減少模型的失配，改善識(shí)別性能。模型參數(shù)通過(guò)期望最大值(EM)算法估計(jì)，可提高計(jì)算效率。最后，通過(guò)貝葉斯規(guī)則，以獲取的統(tǒng)計(jì)特征識(shí)別測(cè)試數(shù)據(jù)，仿真結(jié)果表明該方法能夠提高低信噪比條件下PPCA模型的穩(wěn)健性。
- 雷達(dá)目標(biāo)識(shí)別 /
- 高分辨距離像 /
- 概率主成分分析 /
- t分布 /
- 特征提取
Abstract: In order to improve the sensitivity problem of using Probabilistic Principal Component Analysis (PPCA) model for HRRP recognition, a modified method is proposed. T-distribution is adopted as the basis of PPCA model rather than Gaussian distribution in this method, which utilizes not only the t-distributions robustness, but also less free parameters of PPCA characteristic. Further, to eliminate the targets azimuth sensitivity, the mixture t-distribution is substituted for single t-distribution. This modification offers a potential to model the similar density of HRRP in different azimuth range adequately for clustering and reduces the mismatch between models, thus improves the recognition performance. Estimation of parameters is achieved by EM algorithm to avoid the drawbacks of maximum-likelihood estimation and improve the estimation efficiency. Finally, in the simulation experiment Bayesian rule and the estimation statistical features are adopted together to test new HRRPs, the results show this method can improve the robustness of PPCA model in low SNR conditions.
- Radar target recognition /
- High Resolution Range Profile (HRRP) /
- Probabilistic Principal Component Analysis (PPCA) /
- t-distribution /
- Feature extraction

HTML全文

參考文獻(xiàn)(28)

LIU Hongwei, CHEN Bo, FENG B, et al. Radar high- resolution range profiles target recognition based on stable dictionary learning[J]. IET Radar, Sonar Navigation, 2016, 10(2): 228-237. doi: 10.1049/iet-rsn.2015.0007.

ZHOU Daiying. Radar target HRRP recognition based on reconstructive and discriminative dictionary learning[J]. Signal Processing, 2016, 126: 52-64. doi: 10.1016/j.sigpro. 2015.12.006.

PAN Xiaoyi, WANG Wei, FENG D, et al. Signature extraction from rotating targets based on a fraction of HRRPs[J]. IEEE Transactions on Antennas and Propagation, 2015, 63(2): 585-592. doi: 10.1109/TAP.2014.2379955.

PAN Mian, JIANG Jie, LI Zhu, et al. Radar HRRP recognition based on discriminant deep auto-encoders with small training data size[J]. Electronics Letters, 2016, 52(20): 1725-1727. doi: 10.1049/el.2016.3060.

DU L, HE H, ZHAO L, et al. Noise robust radar HRRP target recognition based on scatter matching algorithm[J]. IEEE Sensors Journal, 2016, 16(6): 1743-1753. doi: 10.1109/JSEN. 2015.2501850.

LUNDAN J and KOIVUNEN V. Deep learning for HRRP- based target recognition in multi-static radar systems[C]. IEEE Radar Conference, Philadelphia, PA, 2016. doi: 10.1109 /RADAR.2016.7485271.

JACOB S P and OSULLIVAN J A. Automatic target recognition using sequences of high resolution radar range- profiles[J]. IEEE Transactions on Aerospace and Electronic Systems, 2000, 36(2): 364-381. doi: 10.1109/7.845214.

WEBB A R. Gamma mixture models for target recognition[J]. Pattern Recognition, 2000, 33(12): 2045-2054. doi: 10.1016/ S0031-3203(99)00195-8.

COPSEY K D and WEBB A R. Bayesian gamma mixture model approach to radar target recognition[J]. IEEE Transactions on Aerospace and Electronic Systems, 2003, 39(4): 1201-1217. doi: 10.1109/TAES.2003.1261122.

DU L, LIU H, BAO Z, et al. A two-distribution compounded statistical model for Radar HRRP target recognition[J]. IEEE Transactions on Signal Processing, 2006, 54(6): 2226-2238. doi: 10.1109/TSP.2006.873534.

WANG Caiyun and XIE Jiling. Radar high resolution range profile target recognition based on t-mixture model[C]. IEEE Radar Conference, Kansas, 2011: 762-767. doi: 10.1109/ RADAR.2011.5960640.

DU L, LIU H, BAO Z, et al. Radar HRRP statistical recognition: Parametric model and model selection[J]. IEEE Transactions on Signal Processing, 2008, 56(5): 1931-1944. doi: 10.1109/TSP.2007.912283.

LIU Hongwei, DU Lan, WANG Penghui, et al. Radar HRRP automatic target recognition: Algorithms and applications[C]. Proceedings of IEEE CIE International Conference on Radar, Chengdu, 2011: 14-17. doi: 10.1109/CIE-Radar.2011. 6159709.

王鵬輝, 杜蘭, 劉宏偉, 等. 雷達(dá)高分辨距離像分幀新方法[J].西安電子科技大學(xué)學(xué)報(bào), 2011, 38(6): 22-29. doi: 10.3969 /j.issn.1001-2400.2011.06.004.

WANG P H, DU L, LIU H W, et al. New frame segmentation method for radar HRRPs[J]. Journal of Xidian University, 2011, 38(6): 22-29. doi: 10.3969/j.issn.1001-2400.2011.06. 004.

王鵬輝, 杜蘭, 劉宏偉. 基于復(fù)高斯模型的雷達(dá)高分辨距離像目標(biāo)識(shí)別新方法[J]. 光學(xué)學(xué)報(bào), 2014, 34(2): 1-10. doi: 10.3788 /AOS201434.0228004.

WANG P H, DU L, and LIU H W. A new method based on complex Gaussian models for radar high resolution range profile target recognition[J]. Acta Optica Sinica, 2014, 34(2): 1-10. doi: 10.3788/AOS201434.0228004.

CHEN T, MARTIN E B, MONTAGUE G A, et al. Robust probabilistic PCA with missing data and contribution analysis for outlier detection[J]. Computational Statistics Data Analysis, 2009, 53(10): 3706-3716. doi: 10.1016/j.csda. 2009.03.014.

LANGE K L, LITTLE R J, TAYLOR J M, et al. Robust statistical modeling using the t distribution[J]. Journal of the American Statistical Association, 2012, 84(408): 881-896. doi: 10.1080/01621459.1989.10478852.

TIPPING M E and BISHOP C M. Probabilistic principal component analysis[J]. Journal of The Royal Statistical Society Series B Statistical Methodology, 1999, 61(3): 611-622. doi: 10.1111/1467-9868.00196.

ZHOU X and LIU X. The EM algorithm for the extended finite mixture of the factor analyzers model[J]. Computational Statistics Data Analysis, 2008, 52(8): 3939-3953. doi: 10.1016/j.csda.2008.01.023.

PEEL D and MCLACHLAN G J. Robust mixture modelling using the t distribution[J]. Statistics and Computing, 2000, 10(4): 339-348. doi: 10.1023/A:1008981510081.

TIPPING M E and BISHOP C M. Mixtures of probabilistic principal component analyzers[J]. Neural Computation, 1999, 11(2): 443-482. doi: 10.1162/089976699300016728.

ZHU D, LIU Y, HUO K, et al. A novel high-precision phase- derived-range method for direct sampling LFM radar[J]. IEEE Transactions on Geoscience and Remote Sensing, 2016, 54(2): 1131-1141. doi: 10.1109/TGRS.2015.2474144.

但波, 姜永華, 李敬軍, 等. 雷達(dá)高分辨距離像自適應(yīng)角域劃分方法[J]. 系統(tǒng)工程與電子技術(shù), 2014, 36(11): 2178-2185. doi: 10.3969/j.issn.1001-506X.2014.11.11.

DAN B, JIANG Y H, LI J J, et al. Adaptive angular-sector segmentation method for radar HRRP[J]. Systems Engineering and Electronics, 2014, 36(11): 2178-2185. doi: 10.3969/j.issn.1001-506X.2014.11.11.

黃得雙. 高分辨雷達(dá)智能信號(hào)處理技術(shù)[M]. 北京: 機(jī)械工業(yè)出版社, 2001: 19-31.

HUANG D S. Intelligent Signal Processing Technique for High Resolution Radars[M]. Beijing: China Machine Press, 2001: 19-31.

相關(guān)文章

施引文獻(xiàn)

資源附件(0)

訪問(wèn)統(tǒng)計(jì)