Marine noise signal identification method based on deep neural network (DNN)

Abstract

The invention discloses a marine noise signal identification method based on a deep neural network (DNN). According to the method, a deep-neural-network model is established, continuous training and updating of forward operating and back propagation are carried out on weight values of each layer of neurons in the model to obtain classification weight values capable of distinguishing different types of ocean noise signals, and thus identification on the different types of ocean noise signals is realized. According to the identification method of the invention, a deep belief network is utilizedto carry out initial weight value training of the deep neural network, obtained weight values are used as initial weight values of training of the deep neural network, then training is carried out ondata, thus identification on the different types of ocean noise signals is realized. According to the method, the deep neural network and the initial values trained by the deep belief network are utilized to enable accuracy of test results to be high, and high-precision identification requirements can be met.

Description

technical field [0001] The invention belongs to the field of weak signal detection, and in particular relates to an ocean noise signal recognition method based on a deep neural network. Background technique [0002] In the field of practical engineering applications, ocean noise signals exist in a large number in the ocean environment, but there are not many studies on them. In the research on its classification, classic methods such as SVM and stochastic resonance were used in the past, and many methods choose to filter out most of the noise and then extract useful specific signals. This kind of complete signal does not necessarily have value in reality, and its detailed classification has high value in some fields. [0003] Based on the deep learning algorithm, different ocean noise signals are classified and learned, and a five-layer DNN deep neural network is established to train different types of ocean noise signals. Through continuous training and updating of the we...

AI Technical Summary

Problems solved by technology

[0004] However, the previous deep learning algorithms are all based on the background of random initial value selection. In actual training, this initial value often results in a large error, which makes the final result and the actual result have a low similarity problem.

Images