A method of underwater acoustic signal enhancement based on autoencoder
An autoencoder and underwater acoustic signal technology, applied in neural learning methods, instruments, biological neural network models, etc., can solve the problem of reduced filtering effect, achieve high noise reduction level, strong robustness, and save manpower
Active Publication Date: 2022-08-05
HARBIN ENG UNIV
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, for the noise generated by the nonlinear system, since the signal and the noise are both broadband continuum on the spectrum, the filtering effect of the traditional method is greatly reduced, which requires exploring a new noise reduction method suitable for nonlinear signals.
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0039] Below in conjunction with specific embodiments, the present invention will be further illustrated, and it should be understood that these embodiments are only used to illustrate the present invention and not to limit the scope of the present invention. The modifications all fall within the scope defined by the appended claims of this application.
[0040] An automatic encoder-based underwater acoustic signal enhancement method disclosed in an embodiment of the present invention mainly includes the following steps:
[0041] (1) Construct a regression autoencoder neural network model with the same number of input and output neurons. The frame diagram of the network is as follows figure 1 As shown, the training network adopts the joint auto-encoder (DAE+CDAE) method combining the denoising auto-encoder and the convolutional denoising automatic-encoder (CDAE). In the pre-training stage, adding noise to the training set (train_clean) is called a noise-added signal (train_no...
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
Login to View More Abstract
The invention discloses an underwater acoustic signal enhancement method based on an automatic encoder, which belongs to the field of underwater acoustic signal processing. Aiming at the problem that the feature extraction of echo signal in active sonar is difficult, the present invention designs an automatic encoder combining a noise reduction automatic encoder and a convolution noise reduction automatic encoder. Firstly, the noise-containing signal is preprocessed by using the noise reduction advantages of the noise reduction auto-encoder on the whole signal; then combined with the optimization of the local features of the signal by the convolutional noise reduction auto-encoder, the signal is denoised locally, so as to realize the signal enhanced. The method of the invention can directly use the time domain waveform of the received signal as the characteristic input, and retain the amplitude and phase characteristics of the signal. The experimental results show that the present invention not only effectively reduces the noise component in the signal, but also achieves a better recovery effect in both the time domain and the frequency domain.
Description
technical field [0001] The invention relates to an underwater acoustic signal enhancement method, in particular to an underwater acoustic signal enhancement method based on a deep learning technology, and belongs to the field of underwater acoustic signal processing. Background technique [0002] In the process of signal processing, due to the influence of noise, long-distance detection and weak signal processing become very difficult. Noise reduction has become a topic that has troubled researchers for a long time. Noise reduction is an essential process for effective signal analysis. . In the traditional signal processing method, noise reduction is done by filtering. When using the linear filtering method, according to the distribution characteristics of the signal in the frequency domain, as long as the time series is long enough, the noise in periodic and quasi-periodic signals is can be completely eliminated. However, for the noise generated by the nonlinear system, t...
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
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G01S7/539G06N3/04G06N3/08
CPCG01S7/539G06N3/084G06N3/045
Inventor 李理罗五雄殷敬伟郭龙祥于雪松顾师嘉韩笑
Owner HARBIN ENG UNIV