Ship underwater noise deep learning identification method based on eigen probability density function

A technology of probability density function and deep learning, applied in the field of marine information, can solve the problems that the modulation spectrum method is difficult to obtain stable features, the time-varying influence of underwater ship signals, and the poor recognition effect, so as to reduce the degree of interference and meet the The effect of processing requirements and accurate and stable identification

Active Publication Date: 2022-05-10
THE FIRST INST OF OCEANOGRAPHY SOA
View PDF11 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, in the modulation spectrum method, a near-stationary approximation is used, which makes it difficult to obtain stable features in the modulation spectrum method, which is seriously affected by the time-varying signal of the underwater ship, and the recognition effect is not good

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
  • Ship underwater noise deep learning identification method based on eigen probability density function
  • Ship underwater noise deep learning identification method based on eigen probability density function
  • Ship underwater noise deep learning identification method based on eigen probability density function

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0032] Here we take the underwater noise signals of 5 ships as an example. The names of the ships, their length and sampling rate information are shown in the table below. Divide these signals into frames, the length of each frame is 10 seconds, and the overlap between frames is 9 seconds, so for these 5 ships, we get 123, 170, 215, 402 and 675 frames respectively. For these data, the processing steps of the inventive method (see figure 1 )as follows:

[0033]

[0034] The first step is to standardize operations. Perform a normalization operation on a frame of noise signal s with a total length of 10 seconds to obtain a standardized noise signal s t . Since the sampling rate is already consistent, it is no longer necessary to work with a unified sampling rate. Remove the DC component of the signal: , where mean() represents the mean value operation; the power of the signal is normalized: , where std() means to take the standard deviation operation. After normalizat...

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 relates to a ship underwater noise deep learning identification method based on an intrinsic probability density function, and belongs to the technical field of ocean information, the method comprises the following steps: carrying out deep learning identification on ship underwater noise based on the intrinsic probability density function, obtaining an intrinsic mode function of a signal through mode decomposition, solving corresponding probability density, and carrying out deep learning identification on the ship underwater noise based on the intrinsic probability density function; and a deep learning classifier is adopted to realize automatic identification of the ship type. The method is more stable in a complex and changeable marine environment, compared with the prior art, the recognition accuracy is improved to 99.6% from 90.8%, the operation speed is improved, and the online processing requirement can be met.

Description

technical field [0001] The invention belongs to the technical field of marine information, and relates to a deep learning recognition method of ship underwater noise based on an intrinsic probability density function. Background technique [0002] Accurate identification of maritime ship targets has important practical significance for maritime ship monitoring and route planning, protection of maritime rights and interests, and enhancement of maritime military power. At present, underwater radiation noise is one of the commonly used signals in ship identification. [0003] The noise signal generated during the ship's navigation will spread around in the form of sound waves underwater. Acoustic wave propagation will be affected by the ocean water environment and the boundaries of the sea surface and the seabed. It is a complex physical process, which is manifested in two aspects: one is that the acoustic channel will modulate the signal during propagation, which has nonlinea...

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
IPC IPC(8): G06K9/00G06N3/04G06N3/08
CPCG06N3/08G06N3/045G06F2218/12Y02T90/00
Inventor 姜莹刘宗伟杨春梅吕连港段德鑫张远凌
Owner THE FIRST INST OF OCEANOGRAPHY SOA
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap