CN117150223B - Cross spectrum turbulence denoising method and system based on AUV multi-scale fixed frequency decomposition - Google Patents

Cross spectrum turbulence denoising method and system based on AUV multi-scale fixed frequency decomposition Download PDF

Info

Publication number
CN117150223B
CN117150223B CN202311411491.4A CN202311411491A CN117150223B CN 117150223 B CN117150223 B CN 117150223B CN 202311411491 A CN202311411491 A CN 202311411491A CN 117150223 B CN117150223 B CN 117150223B
Authority
CN
China
Prior art keywords
shearing
component
modal
vibration acceleration
signal
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202311411491.4A
Other languages
Chinese (zh)
Other versions
CN117150223A (en
Inventor
杨华
毛蓓蓓
张馨睿
郑雨轩
朱小宇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ocean University of China
Original Assignee
Ocean University of China
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Ocean University of China filed Critical Ocean University of China
Priority to CN202311411491.4A priority Critical patent/CN117150223B/en
Publication of CN117150223A publication Critical patent/CN117150223A/en
Application granted granted Critical
Publication of CN117150223B publication Critical patent/CN117150223B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/10Pre-processing; Data cleansing
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/21Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/02Preprocessing
    • G06F2218/04Denoising
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02TCLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T90/00Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)

Abstract

The invention belongs to the technical field of turbulence denoising, and discloses a cross spectrum turbulence denoising method and system based on AUV multi-scale fixed frequency decomposition. According to the method, a multiband decomposition denoising strategy is adopted, an AUV multi-source vibration frequency band priori recognition platform is built to realize multi-characteristic signal division, and noise disturbance in a focusing frequency band is removed by a combined cross spectrum method. The method can well realize the characteristic segmentation of the original shear signal, complete the decomposition of the multisource mixed signal to the specific vibration frequency band signal, perform noise attenuation work according to the specific characteristics of the focusing frequency band, improve the accuracy of acquiring the shear data by the AUV carrying the turbulizer, reduce the influence of the observation platform on the shear observation data, and play a role in promoting the research accuracy of the turbulence evolution mechanism.

Description

Cross spectrum turbulence denoising method and system based on AUV multi-scale fixed frequency decomposition
Technical Field
The invention belongs to the technical field of turbulence denoising, and relates to a cross spectrum turbulence denoising method and system based on AUV multi-scale fixed frequency decomposition.
Background
Turbulence is the research foundation for ocean energy and water exchange, and has important regulation effects on ocean circulation, global climate change and ocean circulation. Limited to dynamic unstructured marine environments, the study of turbulent mixing properties relies primarily on physical quantities acquired by actual marine observations. Autonomous underwater vehicles (Autonomous underwater vehicle, AUV) are efficient observation means for acquiring space-time multi-scale ocean turbulence information due to the observation advantages that the autonomous underwater vehicles are not limited by the space-time range of the sea area and are flexible. The true and reliable turbulence data is the core of AUV mobile ocean observation, and accurate observation information plays a vital role in explaining an ocean turbulence space-time evolution mechanism. However, during observation, noise generated by mechanical vibrations inside the AUV inevitably contaminates the turbulent shear signal. Therefore, the method for detecting and denoising the noise of the observation platform has important significance for improving the accuracy of turbulent flow observation signals under the complex ocean background. The AUV platform has frequency band diversity on noise pollution of turbulent shear signals, and in the traditional decomposition method, the denoising work of multi-component signals is generally started from frequency band division work, and multi-component mixed signals are converted into single components for denoising processing, such as motor fault detection or feature recognition. Therefore, reasonably and effectively dividing the frequency band is a precondition for efficient signal decomposition and denoising.
In the practical application process, most of signal decomposition algorithms are difficult to reasonably decompose an original signal into signal component modes in characteristic information sets, and vibration signals generated by weak motors are extremely easy to mask by strong vibration signals, so that the aim of comprehensive elimination is difficult to achieve. For example, conventional Empirical Mode Decomposition (EMD) algorithms have very strong adaptivity in the process of signal decomposition and are widely used in nonlinear system signal decomposition, such as atmospheric signals, marine signals, acoustic signals, but EMD algorithms are accompanied by very strong modal aliasing during application, contaminating and interfering with subsequent analysis of partial signal characteristics; the wavelet decomposition algorithm can effectively finish the excavation and decomposition of the non-stationary abrupt change signals in the application process, but the decomposition result is influenced by wavelet base selection to a certain extent, and meanwhile, the forced zero filling measures of the edge signals in the process of convoluting with the wavelet cause signal distortion, edge effect and influence on the data decomposition quality. Similar to the EMD algorithm, the Variational Modal Decomposition (VMD) can realize the decomposition of the signal self-adaption to a certain extent, but the number of modes obtained by the self-adaption decomposition is regulated and controlled by the parameter scale alpha, and the decomposed modes cannot be completely matched with the frequency and the characteristic information of the required fixed vibration.
Therefore, a decomposition denoising method more suitable for AUV multi-source noise is needed, which has important significance for accurately acquiring ocean signals in complex environments and promoting the improvement of the research accuracy of the subsequent ocean evolution mechanism.
Disclosure of Invention
The invention aims to provide a cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition, which is based on a cross spectrum turbulence denoising mode of multi-scale fixed frequency decomposition so as to eliminate the influence of AUV multi-source vibration noise and realize accurate and effective observation of ocean turbulence data, thereby meeting the scientific requirements of efficient ocean information observation under the background of complex ocean environment.
In order to achieve the above purpose, the invention adopts the following technical scheme:
an AUV multi-scale fixed-frequency decomposition-based cross spectrum turbulence denoising method comprises the following steps:
step 1, acquiring turbulence information in a marine environment through an AUV carrying turbulence meter, and further obtaining an original vibration acceleration signal and an original shearing signal which are observed by the AUV instrument during the period of submergence, horizontal navigation and floating;
and 2, carrying out multi-scale fixed frequency decomposition on the measured original shearing signal and the original vibration acceleration signal according to the prior inherent noise frequency sequence to obtain a narrow-band shearing mode component to be denoised, a narrow-band vibration acceleration mode component, and a residual broadband shearing mode component and a residual broadband vibration acceleration mode component which do not need to be denoised.
Step 3, cross spectrum denoising is carried out on vibration acceleration modal components and shearing modal components which are in one-to-one correspondence with the inherent vibration noise characteristic frequency in a single frequency band, and the shearing modal components after noise rejection are obtained;
and 4, reconstructing the denoised shearing mode component and residual broadband shearing mode component data.
On the basis of the cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition, the invention also provides a corresponding cross spectrum turbulence denoising system based on AUV multi-scale fixed frequency decomposition, which adopts the following technical scheme:
an AUV multi-scale fixed frequency decomposition-based cross spectrum turbulence denoising system comprises the following steps:
the multi-scale fixed frequency decomposition module carries out multi-scale fixed frequency decomposition on the measured original shearing signal and the original vibration acceleration signal according to the prior inherent noise frequency sequence to obtain a narrow-band shearing mode component and a narrow-band vibration acceleration mode component to be denoised, and a residual broadband shearing mode component and a residual broadband vibration acceleration mode component which do not need denoising;
the cross spectrum denoising module is used for denoising the cross spectrum in a single frequency band of the vibration acceleration modal component and the shearing modal component which are in one-to-one correspondence with the characteristic frequency of the inherent vibration noise to obtain a shearing modal component after noise rejection;
and the reconstruction module is used for reconstructing the denoised shearing mode component and the residual broadband shearing mode component data.
In addition, on the basis of the cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition, the invention also provides computer equipment for realizing the cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition.
The computer device comprises a memory and a processor, wherein executable codes are stored in the memory, and the processor is used for realizing the steps of the cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition when executing the executable codes.
In addition, on the basis of the cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition, the invention also provides a computer readable storage medium for realizing the cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition.
The computer readable storage medium has stored thereon a program for implementing the steps of the above-mentioned cross-spectral turbulence denoising method based on AUV multi-scale fixed frequency decomposition, when the program is executed by a processor.
The invention has the following advantages:
as described above, the invention relates to a cross spectrum turbulence denoising method and system based on AUV multi-scale fixed frequency decomposition. According to the method, a multiband decomposition denoising strategy is adopted, an AUV multisource vibration frequency band priori recognition platform is built to realize multi-characteristic signal division, noise disturbance in a focusing frequency band is removed by a combined cross spectrum method, the capability limitation of the existing denoising technology is overcome, and full cognition on ocean turbulence evolution and support on optimization development of an ocean observation instrument are promoted. The method can well realize characteristic segmentation of the original signals, complete decomposition of the multi-source mixed signals to the signals of the specific vibration frequency bands, perform noise attenuation work according to the specific characteristics of the focusing frequency bands, improve the accuracy of acquiring data by the AUV carried turbulizer, reduce the influence of an observation platform on the turbulence observation data, and play a role in promoting further improvement of the accuracy of turbulence evolution mechanism research.
Drawings
FIG. 1 is a flow chart of a cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition in an embodiment of the invention.
Fig. 2 is a schematic diagram of an experimental diagram of an AUV-mounted turbulator in an embodiment of the present invention.
Detailed Description
The invention is described in further detail below with reference to the attached drawings and detailed description:
an experimental schematic diagram of an AUV-mounted turbulator is shown in fig. 2. As can be seen from fig. 2, the whole observation operation flow is divided into an AUV submerging stage, an AUV horizontal sailing state and an AUV floating stage. Wherein the vibration response of the AUV platform comes from the characteristic superposition of the opening frequency of each motor, and the difference of the running postures under different motion states corresponds to different vibration characteristics. The vibration source of the AUV includes a propulsion motor, a roll motor, a pitch motor, and a buoyancy driven pump. The opening of the motor device causes strong mechanical vibration, the mechanical vibration is transmitted to the AUV shell through the fixed carrier, the surface vibration of the plate shell interacts with surrounding media of the turbulator to generate noise fluctuation, and finally, the observation turbulence signal is polluted. The adjustment of the motion states of different AUV platforms is accompanied by the change of the main frequency component of the vibration source, noise types are alternated and noise masking aggravates the complexity of noise sources, and the complex noise sources provide higher requirements for subsequent denoising work.
The cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition generally works as follows:
and obtaining the inherent noise vibration center frequency of the AUV according to the prior method, performing multi-scale fixed-frequency decomposition on the equivalent original turbulence shear signal and the vibration acceleration signal obtained by AUV observation, and decomposing the equivalent original turbulence shear signal and the vibration acceleration signal into M narrow-band modal components and 1 residual broadband component, wherein the M narrow-band modal components and the 1 residual broadband component correspond to the prior center frequency. And carrying out one-to-one cross spectrum denoising on the shearing modal component corresponding to the center frequency and the acceleration modal components (flow direction, direction spreading and normal direction) in the three orthogonal directions. Because the sum of the decomposed modal components is equal to the original signal, the turbulence signal is reconstructed by combining each eigenmode function and the residual mode, and then the purified turbulence signal after denoising can be obtained. The noise elimination in a single frequency band can finish the platform inherent noise elimination with high pertinence, and the noise elimination comprises the elimination of the motor strong noise and the motor weak noise, so that the interference of the platform vibration on the subsequent analysis of the turbulence signal can be effectively reduced.
The cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition is described in detail below. As shown in fig. 1, the cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition comprises the following steps:
and step 1, carrying out multi-scale fixed frequency decomposition on the measured original shearing signal and the original vibration acceleration signal according to the prior inherent noise frequency sequence to obtain a narrow-band shearing mode component to be denoised, a narrow-band vibration acceleration mode component, and a residual broadband shearing mode component and a residual broadband vibration acceleration mode component which do not need to be denoised.
According to the invention, turbulence information in the marine environment is acquired through the AUV carrying turbulence meter, so that an original vibration acceleration signal and an original shearing signal which are observed by the AUV instrument during the period of submergence, horizontal navigation and floating are obtained.
The vibration noise of the AUV apparatus refers to noise generated by the vibration of the AUV apparatus body or body parts, which affects the surrounding object medium. In general, vibration noise generated by excitation of different vibration sources has different corresponding vibration characteristics, and influences on observation results are different. The inherent vibration noise source of the AUV mainly comprises four types of propulsion motors, roll motors, pitching motors and buoyancy driving pumps. In the ocean observation process, the AUV finishes the running of the submerging, horizontal sailing and floating work by controlling the opening of different motors. When various motor devices are started to work, strong mechanical vibration can be generated, various mechanical vibration can be transmitted to a shell structure of the AUV through a fixed carrier, and shell surface vibration and a medium around a turbulent flow shearing probe arranged at the front end of the AUV instrument interact to generate fluctuation, so that a shearing signal acquired by the turbulent flow shearing probe is polluted.
And constructing a prior observation platform of the pool, and performing targeted judgment and analysis on the inherent vibration frequency characteristics brought by the single starting process of the propulsion motor, the roll motor, the pitch motor and the buoyancy driving pump. The AUV instrument is placed in a pool in a static mode, and 60 seconds of vibration acceleration time domain fluctuation data are obtained; independently starting a propulsion motor, and acquiring vibration acceleration time domain fluctuation data of 60 seconds under the influence of the propulsion motor; closing a propulsion motor, and independently starting a roll motor to acquire vibration acceleration time domain fluctuation data under the influence of the roll motor for 60 seconds; closing a roll motor, independently starting a pitching motor, and acquiring vibration acceleration time domain fluctuation data under the influence of the pitching motor for 60 seconds; and closing the pitching motor, and independently starting the buoyancy driving pump to acquire vibration acceleration time domain fluctuation data under the influence of the buoyancy driving pump for 60 seconds. And carrying out Fourier transform on the vibration acceleration data of the time domain fluctuation, obtaining corresponding frequency domain data, and judging and analyzing the inherent vibration frequency domain characteristics when different motors are started. And acquiring spectrum peaks brought by different motors during operation, wherein the spectrum peaks correspond to inherent vibration noise brought by different motors to AUV observation data, and counting inherent vibration noise frequencies under independent working states of different motors, so that the inherent vibration noise frequencies are used as priori center frequency information corresponding to modal components to be denoised in a subsequent multi-scale fixed frequency decomposition process.
In this embodiment, the multi-scale fixed-frequency decomposition mainly comprises two parts, namely known center frequency modal component decomposition and residual modal component acquisition of unknown center frequency. The known center frequency modal component decomposition includes a narrowband shear modal component su k (t) decomposing and narrow-band vibration acceleration modal component au k (t) decomposing the two parts. Residual modal component acquisition of center frequency is divided into residuesResidual broadband shear mode component su re Acquisition of (t) and residual wideband vibration acceleration modal component au re And (t) obtaining.
Because the two groups of signal sequences are based on the same center frequency omega in the multi-scale fixed-frequency decomposition process of the original shearing signal and the original vibration acceleration signal k ={ω 12 ,…,ω m The method comprises the steps of carrying out decomposition, so that the number of the decomposed shearing mode components is the same as that of vibration acceleration mode components, and one shearing mode component and one vibration acceleration mode component are in one-to-one correspondence at the same center frequency, so that the orderly carrying out of subsequent cross spectrum denoising is ensured. The step 1 specifically comprises the following steps:
step 1.1. The original shear signal s (t) consists of m narrow-band shear modal components su to be denoised k (t) and 1 residual broadband shear mode component su re (t) a composition.
Wherein the prior method can be used for obtaining the center frequency omega corresponding to the narrow-band shearing modal component k ={ω 12 ,…,ω m }。
The determined center frequency corresponds to the natural noise frequency during vibration of the AUV scope, i.e. is equal to the center frequency omega k The corresponding shear mode component su k (t) is one of the shearing data components that is primarily disturbed by noise generated by the operation of the AUV motor.
It is necessary to denoise it with an efficient algorithm to improve the signal-to-noise ratio of the shear observed signal.
Step 1.2. At m narrowband shear mode components su for a known corresponding center frequency k And (t) in the decomposition process, iteratively solving the shearing modal component and Lagrange operator by using an alternating direction multiplication operator based on each known center frequency.
At this time, the corresponding center frequency is ω km
The formula for iteratively solving the shearing modal component and Lagrange operator by using the alternating direction multiplication operator is as follows:
wherein,in order to shear the modal component,is Lagrange operator;for each modal scale parameter, which is used to control each modal bandwidth,to update coefficients iteratively, generallyRepresenting a unit pulse function.
And step 1.3, repeating the iteration formula in the step 1.2 until the iteration constraint condition is met.
If the iteration constraint condition is met, completing the iterative computation of the shearing mode component to obtain the required narrow-band shearing mode component su k (t); and if the following iteration constraint conditions are not met, performing n+1st iteration calculation by the n-th iteration.
The iteration constraint is as follows:
wherein,representing the n +1 th shear mode component,representing the nth shear mode component.
Step 1.4. From the original shear signal s (t) and the narrow-band shear modal component su that has been obtained k (t) obtaining a residual wideband shear modal component su re (t);su re (t)= s(t) -
Where m is the data of the acquired narrowband shear mode component.
Step 1.5. The original vibration acceleration signal a (t) is composed of m narrow-band vibration acceleration modal components au to be denoised k (t) and 1 residual wideband vibration acceleration modal component au re (t) composition.
Wherein the center frequency omega corresponding to the narrowband shear component can be obtained by using an a priori algorithm k ={ω 12 ,…,ω m }。
The determined center frequency corresponds to the natural noise frequency during vibration of the AUV scope, i.e. is equal to the center frequency omega k The corresponding vibration acceleration modal component au k And (t) is one of the main sources of vibration noise generated when the AUV motor operates.
Step 1.6. At m narrow-band vibration acceleration modal components au for a known center frequency k And (t) in the decomposition process, iteratively solving vibration acceleration modal components and Lagrange operators by using an alternating direction multiplication operator based on each known center frequency.
At this time, the corresponding center frequency is ω km
The equation for iteratively solving the vibration acceleration modal component and Lagrange operator by using the alternating direction multiplication operator is as follows:
wherein,representing the modal component of the vibration acceleration,representing Lagrange operators.
And step 1.7, repeating the iteration formula in the step 1.6 until the iteration constraint condition is met.
If the iteration constraint condition is met, completing the vibration acceleration modal component iterative computation to obtain a narrow-band vibration acceleration modal component au k (t); if the iteration constraint condition is not satisfied, carrying out n+1st iteration calculation by the nth iteration.
The iteration constraint is as follows:
wherein,represents the n+1th vibration acceleration modal component,representing the nth vibration acceleration modal component.
Step 1.8. From the raw vibration acceleration signal a (t) and the narrow-band vibration acceleration modal component au that has been obtained k (t) acquiring a residual wideband vibration acceleration modal component au re (t);au re (t)= a(t) -
Wherein m is the data of the acquired narrow-band vibration acceleration modal component.
And 2, carrying out cross spectrum denoising on vibration acceleration modal components and shearing modal components which are in one-to-one correspondence with the inherent vibration noise characteristic frequency in a single frequency band to obtain shearing modal components after noise rejection.
Step 2.1. The shearing mode component su obtained by decomposition in step 1 k (t) is derived from the true shear signal componentAnd the modal component au of the acceleration under vibration k The noise component contribution of (t) is given by:
wherein su k (t) represents the prior center frequency omega obtained after multi-scale fixed frequency decomposition k The corresponding real shearing modal component;representing and prior center frequency omega k The corresponding real shearing modal component; au k (t) represents the prior center frequency omega obtained after multi-scale fixed frequency decomposition k The corresponding vibration acceleration modal component; b (B) k Representing the influence degree of vibration acceleration on the shear signal for a multi-source weighting function; * Representing a convolution calculation.
Step 2.2. Constructing a vibration acceleration modal component au corresponding to the prior center frequency by coherence of the shear signal and the vibration acceleration signal k (t) and a shearing mode component su k (t) transfer function between and calculating an influence factorIs a multi-source weighting function of (a):
wherein,and (3) withRespectively the shearing mode components su k (t) and vibration acceleration modality au k (t) a corresponding self-power spectrum;representing the shear mode component su k (t) and a vibration acceleration modal component au k A cross-power spectrum of (t),representing the shear mode component su k (t) and a vibration acceleration modal component au k And (t) cross-power spectrum conjugation.
Self-power spectrumAnd (3) withThe calculation formula of (2) is as follows:
definition of the definitionIs the spectrum of the signal.
Wherein,for the signal amplitude value,for the phase of the signal,is the angular frequency.
And (3) withRespectively represent the shearing mode components su k (t) and a vibration acceleration modal component au k The frequency spectrum of (t),and (3) withRespectively the shearing mode components su k (t) and a vibration acceleration modal component au k And (t) spectrum conjugation.
Cross power spectrumThe calculation formula of (2) is as follows:
step 2.3. Substituting the multi-source weighting function obtained in step 2.2 into the formula in step 2.1 to obtain the prior center frequency omega k Corresponding real shearing mode component
And 3, reconstructing the denoised shearing mode component and residual broadband shearing mode component data.
Based on denoised shearing modal componentsAnd residual broadband shear mode component su re (t) performing data reconstruction to obtain a clean shear signalThe reconstruction formula is as follows:
wherein,representing the clean shear signal after reconstruction,represents m denoised clean shear mode components, su re And (t) is a residual broadband shear mode component.
According to the cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition, which is disclosed by the invention, the influence of AUV multi-source vibration noise is well eliminated based on a cross spectrum turbulence denoising mode of multi-scale fixed frequency decomposition, and accurate and effective observation of ocean turbulence data is realized, so that the scientific requirement on efficient ocean information observation under the background of a complex ocean environment is met.
Example 2
This embodiment 2 describes a cross spectrum turbulence denoising system based on AUV multi-scale fixed frequency decomposition, which is based on the same inventive concept as the cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition described in the above embodiment 1.
Specifically, cross spectrum turbulence denoising system based on AUV multi-scale fixed frequency decomposition includes:
the multi-scale fixed frequency decomposition module carries out multi-scale fixed frequency decomposition on the measured original shearing signal and the original vibration acceleration signal according to the prior inherent noise frequency sequence to obtain a narrow-band shearing mode component and a narrow-band vibration acceleration mode component to be denoised, and a residual broadband shearing mode component and a residual broadband vibration acceleration mode component which do not need denoising;
the cross spectrum denoising module is used for denoising the cross spectrum in a single frequency band of the vibration acceleration modal component and the shearing modal component which are in one-to-one correspondence with the characteristic frequency of the inherent vibration noise to obtain a shearing modal component after noise rejection;
and the reconstruction module is used for reconstructing the denoised shearing mode component and the residual broadband shearing mode component data.
It should be noted that, in the cross spectrum turbulence denoising system based on AUV multi-scale fixed frequency decomposition, the implementation process of the functions and actions of each functional module is specifically shown in the implementation process of the corresponding steps in the method in the above embodiment 1, and will not be described herein.
Example 3
Embodiment 3 describes a computer apparatus for implementing the cross-spectrum turbulence denoising method based on AUV multi-scale fixed-frequency decomposition described in embodiment 1 above.
In particular, the computer device includes a memory and one or more processors.
Executable codes are stored in the memory, and when the processor executes the executable codes, the executable codes are used for realizing the cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition.
In this embodiment, the computer device is any device or apparatus having data processing capability, which is not described herein.
Example 4
Embodiment 4 describes a computer readable storage medium for implementing the cross-spectral turbulence denoising method based on AUV multi-scale fixed frequency decomposition described in embodiment 1 above.
Specifically, the computer readable storage medium in this embodiment 4 has a program stored thereon, which when executed by a processor, is configured to implement the steps of the cross-spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition described above.
The computer readable storage medium may be an internal storage unit of any device or apparatus having data processing capability, such as a hard disk or a memory, or may be an external storage device of any device having data processing capability, such as a plug-in hard disk, a Smart Media Card (SMC), an SD Card, a Flash memory Card (Flash Card), or the like, which are provided on the device.
The method overcomes the limitations of insufficient mobility and dimension in the prior art, can track and excavate the space-time distribution characteristics of the meter-scale coherent structure in the actual marine environment, provides a method support for tracking the main energetic vortex space-time evolution mechanism, and can promote further insight on the turbulence circulation evolution mechanism.
The foregoing description is, of course, merely illustrative of preferred embodiments of the present invention, and it should be understood that the present invention is not limited to the above-described embodiments, but is intended to cover all modifications, equivalents and alternatives falling within the spirit and scope of the present invention as defined by the appended claims.

Claims (7)

1. The cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition is characterized by comprising the following steps:
step 1, carrying out multi-scale fixed frequency decomposition on an original shearing signal and an original vibration acceleration signal which are obtained through measurement according to an priori inherent noise frequency sequence to obtain a narrow-band shearing mode component to be denoised, a narrow-band vibration acceleration mode component, and a residual broadband shearing mode component and a residual broadband vibration acceleration mode component which do not need denoising;
the step 1 specifically comprises the following steps:
step 1.1. The original shear signal s (t) consists of m narrow-band shear modal components su to be denoised k (t) and 1 residual broadband shear mode component su re (t) obtaining the center frequency omega corresponding to the narrow-band shearing modal component k ={ω 12 ,…,ω m };
Step 1.2. At m narrowband shearing mode components su to be denoised for a known corresponding center frequency k In the decomposition process, based on each known center frequency, iteratively solving a shearing modal component and Lagrange operators by using an alternating direction multiplication operator;
step 1.3, repeating the iteration formula in the step 1.2 until the iteration constraint condition is met; the iteration constraint is as follows:
wherein τ is an iterative update coefficient;
represents the n+1th shear mode component, +.>Representing an nth shear mode component;
if the iteration constraint condition is met, completing the iterative computation of the shearing mode component to obtain the required narrow-band shearing mode component su k (t); if the following iteration constraint conditions are not met, carrying out n+1st iteration calculation by the nth iteration;
step 1.4. From the original shear signal s (t) and the narrow-band shear modal component su that has been obtained k (t) obtaining a residual wideband shear modal component su re (t);
Wherein m is the data of the acquired narrow-band shearing modal components, namely m narrow-band shearing modal components su to be denoised k (t);
Step 1.5. The original vibration acceleration signal a (t) is composed of m narrow-band vibration acceleration modal components au to be denoised k (t) and 1 residual wideband vibration acceleration modal component au re (t) composition;
obtaining the center frequency omega corresponding to the narrow-band shearing component k ={ω 12 ,…,ω m };
Step 1.6. At a known center frequency m narrow-band vibration acceleration modal components au to be denoised k In the decomposition process, based on each known center frequency, utilizing an alternating direction multiplication operator to iteratively solve vibration acceleration modal components and Lagrange operators;
step 1.7, repeating the iteration formula in the step 1.6 until the iteration constraint condition is met; the iteration constraint is as follows:
wherein,represents the n+1th vibration acceleration modal component, +.>Representing an nth vibration acceleration modal component;
if the iteration constraint condition is met, completing the vibration acceleration modal component iterative computation to obtain a narrow-band vibration acceleration modal component au k (t); if the iteration constraint condition is not satisfied, carrying out n+1st iteration calculation by the nth iteration;
step 1.8. From the raw vibration acceleration signal a (t) and the narrow-band vibration acceleration modal component au that has been obtained k (t) acquiring a residual wideband vibration acceleration modal component au re (t);
Wherein m is the data of the acquired narrow-band vibration acceleration modal components, namely m narrow-band vibration acceleration modal components au to be denoised k (t);
Step 2, cross spectrum denoising is carried out on vibration acceleration modal components and shearing modal components which are in one-to-one correspondence with the inherent vibration noise characteristic frequency in a single frequency band, and the shearing modal components after noise rejection are obtained;
the step 2 specifically comprises the following steps:
step 2.1. The shearing mode component su to be denoised obtained by decomposition in step 1 k (t) is defined by the a priori center frequency omega k Corresponding real shearing mode componentWith au k The noise component contribution of (t) is given by:
wherein B is k Characterizing the influence degree of vibration acceleration on a shear signal as a multi-source weighting function, wherein the influence degree is represented by convolution calculation;
step 2.2. Constructing au by the coherence of the shear signal and the vibration acceleration signal k (t) and su k A transfer function between (t) to obtain a multi-source weighting function of the influence factor;
wherein,and->Su respectively k (t) and au k (t) a corresponding self-power spectrum; />Represents su k (t) and au k Cross-power spectrum of (t), +.>Represents su k (t) and au k Conjugation of cross-power spectra of (t);
step 2.3. Substituting the multi-source weighting function obtained in step 2.2 into the formula in step 2.1 to obtain the prior center frequency omega k Corresponding real shearing mode component
And 3, reconstructing the denoised shearing mode component and residual broadband shearing mode component data.
2. The cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition according to claim 1, wherein,
in the step 1, turbulence information in the marine environment is acquired through an AUV carrying turbulence meter, so that an original shearing signal and an original vibration acceleration signal which are observed by the AUV instrument during the period of submergence, horizontal navigation and floating are obtained.
3. The cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition according to claim 1, wherein,
in the step 1.2, the formulas for iteratively solving the shearing modal component and the Lagrange operator are as follows:
wherein,for the n+1th shear mode component, +.>k (t) is Lagrange operator; alpha k For each modal scale parameter, it is used for controlling each modal bandwidth, τ is the iteration update coefficient, δ (t) represents the unit pulse function;
in the step 1.6, the formulas for iteratively solving the vibration acceleration modal component and the Lagrange operator are as follows:
wherein,represents the n+1th vibration acceleration modal component, +.>k (t) represents Lagrange operator.
4. The cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition according to claim 1, wherein,
in the step 2.2, the self-power spectrumAnd->The calculation formula of (2) is as follows:
definition F (ω) =a (ω) e -jθω Is the spectrum of the signal;
wherein a (ω) is a signal amplitude, θ is a signal phase, ω=2pi f is an angular frequency;
F s (omega) and F a (ω) represents su respectively k (t) and au k The frequency spectrum of (t),and->Su respectively k (t) and au k And (t) spectrum conjugation.
5. The cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition according to claim 1, wherein,
in the step 2.2, the cross power spectrumThe calculation formula of (2) is as follows: />Wherein F is s (ω) represents su k Frequency spectrum of (t), I>For au k And (t) spectrum conjugation.
6. The cross spectrum turbulence denoising method based on AUV multi-scale fixed frequency decomposition according to claim 1, wherein,
the step 3 specifically comprises the following steps:
based on the shearing modal component after denoising, namely the shearing modal component is equal to the prior center frequency omega k Corresponding real shearing mode componentAnd residual broadband shear mode component su re (t) performing data reconstruction, and obtaining a reconstructed pure shear signal +.>The reconstruction formula is as follows:
7. an AUV multi-scale fixed frequency decomposition-based cross-spectrum turbulence denoising system for implementing the AUV multi-scale fixed frequency decomposition-based cross-spectrum turbulence denoising method as claimed in claim 1, comprising:
the multi-scale fixed frequency decomposition module carries out multi-scale fixed frequency decomposition on the measured original shearing signal and the original vibration acceleration signal according to the prior inherent noise frequency sequence to obtain a narrow-band shearing mode component and a narrow-band vibration acceleration mode component to be denoised, and a residual broadband shearing mode component and a residual broadband vibration acceleration mode component which do not need denoising;
the cross spectrum denoising module is used for denoising the cross spectrum in a single frequency band of the vibration acceleration modal component and the shearing modal component which are in one-to-one correspondence with the characteristic frequency of the inherent vibration noise to obtain a shearing modal component after noise rejection;
and the reconstruction module is used for reconstructing the denoised shearing mode component and the residual broadband shearing mode component data.
CN202311411491.4A 2023-10-30 2023-10-30 Cross spectrum turbulence denoising method and system based on AUV multi-scale fixed frequency decomposition Active CN117150223B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311411491.4A CN117150223B (en) 2023-10-30 2023-10-30 Cross spectrum turbulence denoising method and system based on AUV multi-scale fixed frequency decomposition

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311411491.4A CN117150223B (en) 2023-10-30 2023-10-30 Cross spectrum turbulence denoising method and system based on AUV multi-scale fixed frequency decomposition

Publications (2)

Publication Number Publication Date
CN117150223A CN117150223A (en) 2023-12-01
CN117150223B true CN117150223B (en) 2024-02-02

Family

ID=88884678

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311411491.4A Active CN117150223B (en) 2023-10-30 2023-10-30 Cross spectrum turbulence denoising method and system based on AUV multi-scale fixed frequency decomposition

Country Status (1)

Country Link
CN (1) CN117150223B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117421561B (en) * 2023-12-18 2024-03-12 中国海洋大学 Turbulence denoising method and system based on parameter optimization VMD (virtual machine direction detector) combined wavelet

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115688637A (en) * 2023-01-03 2023-02-03 中国海洋大学 Turbulent mixing intensity calculation method, system, computer device and storage medium
CN116380462A (en) * 2023-03-20 2023-07-04 南京航空航天大学 Bearing vibration signal noise reduction method for multi-scale mean value variation modal decomposition

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114004263B (en) * 2021-12-29 2022-05-03 四川大学 Large-scale equipment working condition diagnosis and prediction method based on feature fusion conversion

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115688637A (en) * 2023-01-03 2023-02-03 中国海洋大学 Turbulent mixing intensity calculation method, system, computer device and storage medium
CN116380462A (en) * 2023-03-20 2023-07-04 南京航空航天大学 Bearing vibration signal noise reduction method for multi-scale mean value variation modal decomposition

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Deep Denoising Method for Side Scan Sonar Images without High-quality Reference Data;Xiaoteng Zhou等;《 IEEE Xplore》;全文 *
GPS与加速度计融合桥梁变形信息提取模型研究;韩厚增;王坚;孟晓林;;中国矿业大学学报(03);全文 *

Also Published As

Publication number Publication date
CN117150223A (en) 2023-12-01

Similar Documents

Publication Publication Date Title
CN117150223B (en) Cross spectrum turbulence denoising method and system based on AUV multi-scale fixed frequency decomposition
CN108919240B (en) Underwater acoustic target radiation noise modulation spectrum reconstruction method based on group sparse structure
EP2297557B1 (en) Reconstructing an acoustic field
Saijyou et al. Data extrapolation method for boundary element method-based near-field acoustical holography
CN103529441A (en) Method and system for detecting and distinguishing passive synthetic aperture target signal
CN105866740A (en) Underwater sound matched field localization method based on compressed sensing
Shah et al. A high-resolution continuous-scan acoustic measurement method for turbofan engine applications
CN112577592A (en) Finite space plane near-field acoustic holography measuring method based on space Fourier transform
CN116068493A (en) Passive sound source positioning method for deep sea large-depth vertical distributed hydrophone
Sun et al. Self-noise spectrum analysis and joint noise filtering for the sea-wing underwater glider based on experimental data
Nevarez et al. Cnn sensor analytics with hybrid-float6 quantization on low-power embedded FPGAs
CN116992263A (en) Method for extracting, identifying and separating self-noise characteristics of underwater vehicle based on multitask learning
CN115062547A (en) Wind-driven sea surface and ship wake fluid micro-element orbit speed calculation method and system
Shang et al. Combined noise reduction and DOA estimation algorithm for MEMS vector hydrophone based on variational mode decomposition
Chu et al. MTSA-Net: a multiscale time self-attention network for ship radiated self-noise reduction
CN114252148A (en) Sound field reconstruction method based on prolate ellipsoid wave superposition
Qi et al. Holographic reconstruction research on the radiated acoustic field of the underwater vehicle
Tyshchenko et al. Modern Methods of Sound Propagation Modeling Based on the Expansion of Acoustic Fields over Normal Modes
Hu et al. Underwater moving noise source identification based on hybrid near-field acoustical holography
Wang et al. Pulsar signal denoising method based on empirical mode decomposition and independent component analysis
Qian et al. A semi-analytical method for the 3d elastic structural-acoustic radiation in shallow water
Hu et al. Underwater hybrid near-field acoustical holography based on the measurement of vector hydrophone array
Hu et al. Underwater patch near-field acoustical holography based on particle velocity and vector hydrophone array
CN117169862B (en) Deep sea broadband signal waveform rapid simulation method and system based on ray acoustics
CN111368418B (en) Multi-scale simulation analysis method for flow-induced noise

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant