Underwater sound target identification method and system based on improved DBN

Abstract

The invention provides an underwater acoustic target identification method and system based on an improved DBN. The method comprises the following steps: acquiring an actual radiation noise signal of an underwater acoustic target, and obtaining frequency spectrum data after fast Fourier transform; on the basis of the frequency spectrum data, performing calculation through a three-layer stacked RBM and a full connection layer of the improved DBN model with the stored parameters to obtain feature values of all classes, and converting the feature values into probabilities of all classes through a softmax classifier; and according to the label value corresponding to the category, outputting the category label with the maximum probability value, and outputting the category name. According to the underwater acoustic target identification method, cross entropy of hidden layer output and label information is introduced, a target function of DBN network training is improved, the problem of classification information fuzziness in the unsupervised pre-training process is solved, the influence of marine environment noise is weakened, and the identification accuracy is improved under the condition that the identification speed is enough.

Description

technical field [0001] The invention belongs to the field of underwater acoustic target detection and identification, and in particular relates to an improved DBN-based underwater acoustic target identification method and system. Background technique [0002] In recent years, due to the improvement of computing power and the improvement of neural network training and optimization algorithms, deep learning methods have been widely used in underwater target recognition. Deep learning has great potential in underwater acoustic target detection and recognition: Compared with shallow models, deep learning models can describe the rich internal information of underwater acoustic target data by simulating the learning process of the brain, and ultimately improve the recognition accuracy. Therefore, in order to improve national defense strength, choose the best countermeasure to quickly and effectively detect and defend against the attack of underwater acoustic targets, and seize the...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to solve the problem of fuzzy classification information in the process of unsupervised pre-training by proposing an improved DBN-based underwater acoustic target recognition method and system, which can reduce the impact of marine environmental noise and improve the recognition speed. Improve recognition accuracy under sufficient conditions

Images