A method and system for underwater sound source localization based on domain adaptive network

Abstract

The invention discloses a method and system for locating underwater sound sources based on a domain self-adaptive network. The method includes: generating simulated sound pressure data using environmental information in the experimental sea area, collecting actual sound pressure data in the experimental sea area; Noise and normalize the pressure data and the collected actual sound pressure data, and calculate the covariance matrix; generate labels; build a domain adaptive network, which includes a feature extraction network, a sound pressure sample classification network and a domain classification network ; Use the trained domain adaptive network to calculate the classification probability of the actual sound pressure data, and use the position corresponding to the category with the highest probability as the predicted sound source position. The present invention introduces unsupervised actual data into the training process by introducing domain self-adaptive learning, which can effectively improve the accuracy of the network's cross-domain positioning of underwater sound sources, and alleviate the lack of actual data to a certain extent. insufficient to train the model.

Description

technical field [0001] The present invention relates to the technical field of acoustic signal processing and the field of artificial intelligence, in particular to an underwater sound source localization method and system based on a domain adaptive network Background technique [0002] In recent years, more and more attention has been paid to the strategic position of the ocean. Target positioning is closely related to the detection and development of marine resources. Therefore, the passive positioning of target sound sources is an important issue in the field of underwater acoustics. At present, the matching field processing method based on physical acoustic field modeling is the mainstream algorithm in the field of underwater acoustic passive positioning. However, the traditional matching field takes a long time to calculate when the mesh is finer, and it is easily affected by the accuracy of environmental prior information. Therefore, in recent years, the underwater so...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a method and system for underwater sound source localization based on a domain adaptive network for the purpose of locating the position of the sound source in view of the shortcomings of existing underwater sound source localization algorithms.

Images