Unlock instant, AI-driven research and patent intelligence for your innovation.

Deep sea sound source distance measurement method and system based on improved deep neural network

A deep neural network and ranging method technology, applied in the field of deep-sea sound source ranging methods and systems, can solve problems such as poor stability and scattered distance estimates, achieve low relative error, small fluctuation range, improve ranging performance and The effect of stability

Active Publication Date: 2021-11-16
INST OF ACOUSTICS CHINESE ACAD OF SCI
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] However, when current deep learning methods perform ranging tasks, the distance estimates are scattered within the estimated range and have poor stability.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Deep sea sound source distance measurement method and system based on improved deep neural network
  • Deep sea sound source distance measurement method and system based on improved deep neural network
  • Deep sea sound source distance measurement method and system based on improved deep neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0062] like figure 1 As shown, Embodiment 1 of the present invention proposes a deep-sea sound source ranging method based on an improved deep neural network, which is divided into three main steps: data preprocessing and labeling, generating Gaussian labels and training deep neural networks, measuring Data preprocessing and neural network output testing.

[0063] Step 1: Preprocess the simulation and experimental data. The input environmental parameters are sent to the sound field calculation program KrakenC to calculate the frequency domain simulated sound pressure field data set, which is combined with the experimental data to form a data set, and the influence of the sound source amplitude is removed by frequency normalization. For the input data (including the simulated sound pressure field data and the real received data), data preprocessing is required to reduce the influence of the sound source intensity amplitude spectrum. Use formula (1) to normalize each frequency...

Embodiment 2

[0080] Embodiment 2 of the present invention proposes a deep-sea sound source ranging system based on an improved deep neural network. The system includes: a vertical array including N array elements, a trained deep neural network, a data preprocessing module and Sound source distance estimation module;

[0081] The data preprocessing module is used to process the complex sound pressure value measured in real time by each array element of the vertical array through FFT processing to obtain discrete M frequency values, thereby forming a sequence with a data dimension of 2×M×N, Perform frequency domain normalization on the sequence;

[0082] The sound source distance estimation module is used to input the frequency domain normalized sequence into the trained deep neural network, output an L×1 array, and obtain the output layer neuron corresponding to the maximum value in the array. serial number, thereby estimating the sound source distance.

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a deep sea sound source distance measurement method and system based on an improved deep neural network, and the method and system are used for measuring the distance between a vertical array and a sound source, and the vertical array comprises N array elements. The method comprises the following steps: performing FFT (Fast Fourier Transform) processing on a complex sound pressure value measured in real time by each array element of a vertical array to obtain M discrete frequency values so as to form a sequence with a data dimension of 2 * M * N, and performing frequency domain normalization processing on the sequence; and inputting the sequence subjected to frequency domain normalization processing into a pre-trained deep neural network, outputting an L * 1 array, and obtaining a serial number of an output layer neuron corresponding to the maximum value in the array, thereby estimating the sound source distance. According to the method, the distance estimation of the underwater sound source in the typical deep sea environment can be realized, and the distance measurement precision and stability can be remarkably improved.

Description

technical field [0001] The invention relates to the field of underwater acoustic physics, in particular to a deep-sea acoustic source ranging method and system based on an improved deep neural network. Background technique [0002] As a main function of sonar system, underwater acoustic passive ranging has been a problem that underwater acoustic workers have been working on for many years. Since the ocean is a complex sound channel with time and space variation, traditional localization methods often face the problems of environment mismatch and too much computation. [0003] In recent years, as an emerging branch based on data-driven methods, deep learning has provided a new idea for underwater acoustic passive ranging with its powerful feature extraction ability and unique advantages in processing complex, high-dimensional, nonlinear and other data. The deep neural network establishes a complex nonlinear mapping between high-dimensional underwater acoustic physical quanti...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62G06N3/04G06N3/08G01S15/08
CPCG06N3/08G01S15/08G06N3/045G06F18/214Y02A90/30Y02A90/10
Inventor 王文博肖旭苏林任群言马力
Owner INST OF ACOUSTICS CHINESE ACAD OF SCI