The Technology Research on Aircraft Passive Sonic Detect and Identification
|School||Changchun University of Science and Technology|
|Course||Pattern Recognition and Intelligent Systems|
|Keywords||Aircraft Passive audio detection Wavelet Analysis Feature Extraction Neural Networks|
With the development of electronic jamming , anti-radiation missiles , stealth technology in modern warfare , as the main part of the airborne early warning system radar increasingly vulnerable to attack, passive audio detection technology is effective supplement of radar detection technology , it issued audio signal received battlefield target motion as the object of study , they do not signal , which has a strong hidden . Locate and identify aircraft passive audio detection system is divided into two sub-systems , this paper focuses on the identification subsystem . Firstly, the application of modern signal processing means to target audio signal wavelet transform feature extraction methods . Energy scale detail signals based on wavelet theory wavelet decomposition time domain based on the spectral characteristics of two kinds of battlefield targets audio feature extraction algorithm , using this algorithm to obtain a lower-dimensional feature vectors . And design the BP neural network and SOM neural network two classifications is . Target recognition experiments using helicopter II battlefield targets measured acoustic signal data , application of the wavelet analysis feature extraction techniques and classifiers identify classification , to examine their performance .