首页|时域声波障碍反散射问题的神经网络方法

时域声波障碍反散射问题的神经网络方法

扫码查看
研究了一种求解时域声波移动障碍物反散射问题的神经网络方法.该方法由一维卷积模块和多头自注意力机制模块构成,其中一维卷积模块的特征提取能力有效捕捉散射数据的局部特征;多头自注意力机制模块的全局信息捕捉能力综合分析散射数据的全局特征,采用误差的反向传播进行训练,反演障碍物的运动轨迹.实验结果表明,该方法能有效反演移动障碍物的运动轨迹.
Neural Network Methods for the Inverse Scattering Problem of Acoustic Obstacles in the Time Domain
In this paper,a neural network method for solving the time-domain inverse scattering problem of moving obstacles is studied.The method is composed of one-dimensional convolution module and multi-head self-attention mechanism module,in which the ability of feature extraction of one-dimensional convolution module can effectively capture the local features of scattering data;The global information capture ability of the multi-head self-attention mechanism module synthetically analyzes the global characteristics of the scattered data,and uses the back propagation of the error to train,and retrieves the movement track of the obstacle.The experimental results show that this method can effectively retrieve the moving obstacle trajectory.

time-domain acoustic inverse scattering problemsmulti-head self-attentionone-dimensional convolutionfeed-forward neural network

刘一雄、孟品超

展开 >

长春理工大学 数学与统计学院,长春 130022

时域声波反散射问题 多头自注意力 一维卷积 前馈神经网络

国家自然科学基金项目吉林省自然科学基金项目吉林省科技计划项目吉林省工业技术研发项目

1227120720220101040JCYDZJ202201ZYTS5852022C047-2

2024

长春理工大学学报(自然科学版)
长春理工大学

长春理工大学学报(自然科学版)

CSTPCD
影响因子:0.432
ISSN:1672-9870
年,卷(期):2024.47(5)