Underwater Explosion Signal Feature Analysis Method

Based on the background and peculiarity of the underwater explode signal, this paper introduces some theoretical and practical methods to analyze the underwater explode signals. Also these methods performed effective by the computer simulation and practical computation with the actual signal data. This paper concentrates on the detection and feature extraction of the underwater exploding signal. The first section of this paper introduces how to use the wavelet transform filter to pertreat the signals under the circumstance of the ocean noise. The second section aimed at the detection of the instantaneous signal, which is transient, unstable and difficult to predict its waveform type and occurring time mixed in the noise. According to these features of the exploding signal, this paper provided some method to detect the signal in the underwater noise circumstance and test these methods with the actual signal data. In the third section this paper emphasize the work of feature extraction of the underwater exploding signal. We applied the wavelet transform packet method in this field and tested it good and useful through classification experiment.