基于改进型卷积神经网络的水下声源定位：深海实验应用。

Localization of Immersed Sources by Modified Convolutional Neural Network: Application to a Deep-Sea Experiment.

机构信息

Institute of Acoustics, Chinese Academy of Sciences, Beijing 100190, China.

Key Laboratory of Underwater Acoustics Environment, Chinese Academy of Sciences, Beijing 100190, China.

出版信息

Sensors (Basel). 2021 Apr 29;21(9):3109. doi: 10.3390/s21093109.

DOI:10.3390/s21093109

PMID:33946971

原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC8124261/

Abstract

A modified convolutional neural network (CNN) is proposed to enhance the reliability of source ranging based on acoustic field data received by a vertical array. Compared to the traditional method, the output layer is modified by outputting Gauss regression sequences, expressed using a Gaussian probability distribution form centered on the actual distance. The processed results of deep-sea experimental data confirmed that the ranging performance of the CNN with a Gauss regression output was better than that using single regression and classification outputs. The mean relative error between the predicted distance and the actual value was ~2.77%, and the positioning accuracy with 10% and 5% error was 99.56% and 90.14%, respectively.

摘要

提出了一种改进的卷积神经网络（CNN），用于增强基于垂直阵接收的声场数据的源测距可靠性。与传统方法相比，通过输出高斯回归序列来修改输出层，该序列采用以实际距离为中心的高斯概率分布形式表示。深海实验数据的处理结果证实，具有高斯回归输出的 CNN 的测距性能优于使用单一回归和分类输出的 CNN。预测距离与实际值之间的平均相对误差约为 2.77%，误差为 10%和 5%时的定位精度分别为 99.56%和 90.14%。