Multi-Static Target FusionREcognition method Based. Bi-Spectrum Feature Prediction

Hongwen Wang, Leyuan Jin

Abstract


Bi-Spectrum Feature of target are abstract and the Method Based on Feature indication is used to conduct multi-Static tar-get recognition.The target recognition conditional probability is obtained through the Bayes rule.And the formula is simplified.The prediction feature of the last node is obtained by BP(Back Propagation)Neural Network,And the prediction error probabilistic is calcu-related using the Gaussian Mixture Model.Then the result multiple by the recognition probability of each former receiver to obtain the conditional probability of the target type.The procedure is rated for every target,And the target type response to the maximum probability is the type of target to be identified.Multi-Static simulation experiment is conducted in Alibaba,And the above method is hired to identify the targets.Compare with the Mono-Static system,The recognition rate of the improved method innovation-ses by above 10% under the condition of certain signal-To-Reverberation Ratio.


Keywords


Multi-Static Sonar System,Target Recognition,Features Prediction,Error Mapping,Bi-Spectrum Feature

Full Text:

PDF

References


AI Xiaofeng.,Li Yongzhen,Et al.Some Results on characteristics of Bistatic high-Range Resolution profiles for target classifica-tion[J].IetRAdar sonar navig.,2012,6.(5.):379-388.

AI Xiaofeng,Zou Xiaohai,Yang Jianhua,Et al.Micro-Motion Feature ExtractionROtational target in bistaticRAdar[C]//IEEE CIERAdar Conference,2011:609-612.

AZIMI-Sadjadi MahmoodR.,Yao de,Huang Qiang,Et al.Underwater Target classification using wavelet packets and Neural Networks[J].IEEE Transactions on Neural Networks,2000,11(3.):784-794.

SalazaRJaime,RObinson Marc,AZIMI-Sadjadi MahmoodR.Multi-Aspect Pattern Classification Using Predictive Net-works, Error Mapping[C]/Neural Networks2002.Ijcnn '02.Proceedings 2002 International Joint Conference.Honolulu2002:1842-1847.

Bed Xiaofeng.Double/More base sonar positioning and Target Characteristics Research [D].Northwest Industrial University.2003:1-2.

Ling National.More address sonar system an overview [J].Acoustic and Electronic Engineering.1995,(2):12-18.

Edge starting regards Zhang Xue-work and.Pattern recognition [M].Tsinghua University Press.2000.

Zhang Lin Huang Min.Based onEMDAnd slice bi-spectrum of Bearing Fault Diagnosis Methods [J]., Beijing, aviation Space University Journal,201036(3):287-290.

Liu Wu bed zhi jun beans Ren-fu.Based on Time-Domain double spectrum characteristics of radar target recognition [J].Data collection of and Processing,200924(6):709-713.

Liu Yin.Based on Active Acoustic methods of fish Recognition Technology Research [D]., China Academy of Sciences Graduate School.2012.

Fresh Xiaodong fan Yucheng guarantee Investment Limited Company.Based onFisherThan of Mel Cepstrum coefficient mixed characteristics extraction methods [J].Computer Application,201434(2):558-561.




DOI: https://doi.org/10.18686/mcs.v1i1.1136

Refbacks

  • There are currently no refbacks.