Underwater magnetic anomaly signal noise reduction method based on improved empirical mode decomposition, computer equipment and storage medium

Abstract

The invention discloses an underwater magnetic anomaly signal noise reduction method based on improved empirical mode decomposition, computer equipment and a storage medium, and relates to the field of underwater magnetic anomaly signal noise reduction. Effective information extraction of underwater magnetic anomaly signals in the prior art is achieved, and the problems of end effect and mode aliasing are solved. According to the scheme, the method comprises the steps that a magnetic target model is established, an underwater small boat serves as a magnetic target, sea wave magnetic field noise and solar wind magnetic field noise are added into the magnetic target, and underwater magnetic anomaly simulation signals are obtained; using improved empirical mode decomposition to decompose underwater magnetic anomaly data, and using decomposed data to construct a high-dimensional matrix; and carrying out dimensionality reduction on the decomposed data by adopting a principal component analysis method, selecting out principal components of which the cumulative contribution rate is not less than 95% to form a new original signal, extracting magnetic anomaly signals covered under different noise amplitudes from the new original signal, and completing the noise reduction of the underwater magnetic anomaly signals. The invention relates to the field of improved empirical mode decomposition algorithm analogue simulation experiments.

Description

Technical Field

[0001] The invention relates to the field of underwater magnetic anomaly signal noise reduction. Background Art

[0002] Marine magnetic detection is one of the most effective means of searching for underwater magnetic bodies. Underwater magnetic anomaly signals are generated by superimposing the induced magnetic field generated by magnetic bodies on the marine magnetic background field, which causes significant distortion of the marine magnetic background field and changes the distribution of the geomagnetic field in the space around the location. Obtaining underwater magnetic anomaly signals through a magnetic detector is a good method for obtaining magnetic target information, and plays an important role in detecting ferromagnetic bodies and marine resources in the ocean. However, magnetic anomaly signals are subject to the influence of background noise and distance, which makes it difficult to directly distinguish the signals detected by the magnetic detector, and filtering and denoising are required to effectively extract detection information.

[0003] The underwater magnetic anomaly signal is a low-frequency weak signal, which is theoretically composed of the geomagnetic field and the magnetic anomaly magnetic field generated by the ferromagnetic target. However, in practical applications, magnetic anomaly signals are often accompanied by various high-frequency and low-frequency interference magnetic fields. Traditional filtering and denoising methods cannot effectively retain magnetic anomaly signals while eliminating or reducing noise. In recent years, the method of empirical mode decomposition has been used to filter magnetic anomaly signals, aiming to improve the adaptability of screening data, and has achieved good results. The patent with document number CN201811270149.6 discloses a geomagnetic filtering method based on multi-scale empirical mode decomposition. However, since empirical mode decomposition still has problems in endpoint effects and modal aliasing, many improved empirical mode decomposition methods have been derived. The patent with document number CN201610748459.9 discloses a geomagnetic measurement signal denoising method based on improved empirical mode decomposition. Although the above-mentioned documents and patents are effective methods for denoising magnetic anomaly signals, when processing signals with complex signal components, nonlinearity and high noise intensity, a small part of the noise will still be distributed in the main IMF. This is because the decomposed IMF components are not mutually orthogonal, so the endpoint effect and modal aliasing problems are not well solved. Summary of the invention

[0004] The purpose of the present invention is to solve the existing problem of effective information extraction of underwater magnetic anomaly signals and to solve the problems of endpoint effect and modal aliasing.

[0005] To achieve the above object, the present invention provides the following technical solution: a method for denoising underwater magnetic anomaly signals based on improved empirical mode decomposition, comprising the following steps: S1, establishing a magnetic target model, taking an underwater boat as a magnetic target r0(x,, y, z), adding sea wave magnetic field noise and solar wind magnetic field noise to the magnetic target r0(x,, y, z), and obtaining an underwater magnetic anomaly simulation signal;

[0006] S2, according to the underwater magnetic anomaly simulation signal obtained in S1, the underwater magnetic anomaly data is decomposed by using improved empirical mode decomposition, and a high-dimensional matrix is ​​constructed using the decomposed data;

[0007] S3. For the high-dimensional matrix constructed in S2, the principal component analysis method is used to reduce the dimension of the decomposed data, and the principal components with a cumulative contribution rate of not less than 95% are selected to form a new original signal. The magnetic anomaly signals masked by different noise amplitudes are extracted from the new original signal to complete the noise reduction of the underwater magnetic anomaly signal.

[0008] Furthermore, a preferred embodiment is provided, in which in S1, when the distance between the magnetic detector and the magnetic target r0(x,, y, z) is not less than 2.5 times, the magnetic target r0(x,, y, z) is a point.

[0009] Further, a preferred embodiment is provided, in which the magnetic field at the magnetic target r0 (x, y, z) is:

[0010]

[0011] Among them, M (Mx, My, Mz) is the magnetic moment vector of the magnetic target. Since the magnetic permeability of seawater is approximately equal to the magnetic permeability of vacuum, that is, u0 takes the magnetic permeability of vacuum, r = |r| is the distance from the magnetic target to the magnetic detector, and r0 is the relative position vector between the magnetic target and the magnetic detector in the spatial rectangular coordinate system.

[0012] Further, a preferred embodiment is provided, in which the magnetic field generated by the magnetic target r0(x, y, z) in space is decomposed into three components, and the formula is expressed as follows:

[0013]

[0014] Among them B x , B y , B z is the component of the magnetic field generated by the magnetic target in space along the x, y, and z axes, I is the magnetic inclination, D is the magnetic declination, r x 、r y 、r z are the relative position components between the magnetic target and the magnetic detector along the x, y, and z directions respectively.

[0015] Further, a preferred embodiment is provided, wherein S2 is:

[0016] S2.1: Add Nm white noises of different amplitudes to the simulated underwater magnetic anomaly signal generated in S1 to generate Nm new signals {x p (t), p = 1, 2, ..., Nm};

[0017] S2.2 uses the empirical mode decomposition method to decompose the Nm signals with white noise added, where any signal is decomposed into i components. The underwater magnetic anomaly signal can be expressed as follows:

[0018]

[0019] Among them, c p,i (t) represents the original signal X after adding white noise to the pth p (t) The useful signal components obtained by decomposition are i in total, and their frequencies are arranged from high to low, r p (t) represents the useless components added each time decomposition; p,i (t) and r p (t) are integrated and averaged to obtain the final magnetic anomaly signal, and the decomposed result is as follows:

[0020]

[0021] Among them, s i (t) is c p,i (t) is the result of integrated averaging, k(t) is r p (t) The result after integration and averaging;

[0022] S2.3: The IMF component of the magnetic anomaly signal decomposed in S2.2 above is denoted as a m , thus forming a high-dimensional magnetic anomaly characteristic matrix A, the specific form is as follows:

[0023] A=[a1,a2,a3,K,a m ].

[0024] Further, a preferred implementation is provided, wherein S3 is:

[0025] S3.1: According to the high-dimensional matrix constructed in S2, let a' m =a m -E(a m ), where E(a m ) is the individual sample expectation, then the high-dimensional magnetic anomaly feature matrix after conversion is as follows:

[0026] A'=[a'1,a'2,a'3,K,a' m ]

[0027] S3.2: Construct the covariance matrix C of A' through the converted high-dimensional magnetic anomaly feature matrix. The specific expression is as follows:

[0028] C=A' T ×A' / (k-1)

[0029] Among them, A′ T is the transposed matrix of matrix A', k is the number of column vectors of each sample;

[0030] S3.3: Use singular value decomposition to find the eigenvalue λ and eigenvector U of the covariance matrix C. The formula is:

[0031] C=PΛP T

[0032] In the above formula: Λ is a diagonal matrix, and its element eigenvalues ​​λ are arranged in descending order. This is the eigenvalue matrix of the covariance matrix. Each column of P is the eigenvector U corresponding to the corresponding eigenvalue. This is the eigenvector matrix of the covariance matrix corresponding to the eigenvalue, P T is the rotation matrix of P; the sample data is rotated by the rotation matrix P T The linear transformation of is used to obtain the principal components arranged by variance and with decreasing energy.

[0033] Further, a preferred implementation is provided, and the calculation formula used to calculate the cumulative contribution rate of the characteristic value is as follows:

[0034]

[0035]

[0036] Where: j is the jth eigenvector, bj is the information contribution rate of the jth eigenvalue, α q is the cumulative contribution rate of the first q eigenvalues, λ l The eigenvalue λ is accumulated from L=1 to m or q.

[0037] Further, a preferred implementation is provided, wherein the formula for constructing the principal component composed of the eigenvectors is:

[0038] i new principal components y composed of eigenvectors i The principal component matrix Δ can be defined as:

[0039] Δ=P T A

[0040] The specific expansion form is:

[0041]

[0042] Among them, u 1j -u mj is the eigenvalue corresponding to each eigenvector in the m principal components.

[0043] Solution 3: A computer device includes a memory and a processor, wherein the memory stores a computer program. When the processor runs the computer program stored in the memory, the processor executes any one of the methods described above.

[0044] Solution 4: A computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of any of the methods described above are implemented.

[0045] The present invention is beneficial in that:

[0046] The present invention proposes a method for denoising underwater magnetic anomaly signals based on improved empirical mode decomposition. It is necessary to use an improved empirical mode decomposition method that can not only realize effective information extraction of underwater magnetic anomaly signals, but also solve the endpoint effect and modal aliasing problems to obtain underwater magnetic anomaly information.

[0047] The present invention provides a principal component analysis method. The emergence of the principal component analysis method provides a new idea for this demand. Since the modal aliasing phenomenon of empirical mode decomposition and its various improved methods is caused by the incomplete orthogonality between IMF components, and principal component analysis can convert an array with correlation into an orthogonal array, principal component analysis can be performed on the IMF components to make them completely orthogonal, thereby suppressing the modal aliasing phenomenon.

[0048] The underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition of the present invention has low complexity and does not introduce new noise, and has great advantages in processing signals with complex components, nonlinearity and large noise intensity. The use of principal component analysis combined with empirical mode decomposition to filter and denoise the simulated underwater magnetic anomaly signal plays an important role in detecting ferromagnetic objects and marine resources in the ocean.

[0049] The present invention establishes a magnetic dipole model to simulate magnetic anomalies, sets simulation parameters, and adds sea wave magnetic fields and solar wind magnetic fields as noise, thereby simulating underwater magnetic anomaly signals to a great extent.

[0050] The improved empirical mode decomposition method proposed in the present invention starts from the characteristics of the signal itself, completely orthogonally decomposes the signal, and the magnetic anomaly signals and noise with different characteristics are separated into different IMF components. By comparing with the classical EMD algorithm and the EEMD improved algorithm, it is found that the improved empirical mode decomposition algorithm proposed in the present invention is superior to EEMD and EMD in the removal of modal aliasing and denoising effects on the simulated experimental signal, and achieves the best effect.

[0051] The present invention is also applicable to the field of simulation experiments for improving empirical mode decomposition algorithms. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a flow chart of a method for denoising underwater magnetic anomaly signals based on improved empirical mode decomposition as described in implementation mode 1.

[0053] Figure 2 This is a magnetic anomaly simulation waveform diagram of an underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition described in implementation mode 1.

[0054] Figure 3 This is a simulated waveform diagram of a magnetic anomaly signal containing sea wave magnetic field and solar wind magnetic field noise according to an underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition described in implementation mode one.

[0055] Figure 4 Schematic diagram of IMF components obtained after collective mode decomposition in the present invention.

[0056] Figure 5 It is a schematic diagram of the cumulative contribution rate of the main components of the magnetic anomaly signal involved in the seventh embodiment of the present invention.

[0057] Figure 6 This is a comparison diagram of magnetic anomaly signal reconstruction involved in the present invention.

[0058] Figure 7 This is a comparison diagram of magnetic anomaly signal reconstruction using the common EEMD algorithm.

[0059] Figure 8 It is a schematic diagram of evaluating the denoising algorithm involved in the present invention.

[0060] In the figure, 1 is the correlation coefficient, 2 is the root mean square error, and 3 is the signal-to-noise ratio. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical solutions and advantages of the implementation methods of the present application clearer, the technical solutions in the implementation methods of the present application will be clearly and completely described below in conjunction with the drawings in the implementation methods of the present application. Obviously, the described implementation methods are only part of the implementation methods of the present application, not all of the implementation methods.

[0062] Implementation method 1, see Figures 1 to 5 This embodiment provides a method for reducing noise of underwater magnetic anomaly signals based on improved empirical mode decomposition, and the method comprises the following steps:

[0063] S1. Establish a magnetic target model, take the underwater boat as the magnetic target r0(x, y, z), add the wave magnetic field noise and the solar wind magnetic field noise to the magnetic target r0(x, y, z), and obtain the underwater magnetic anomaly simulation signal;

[0064] S2, according to the underwater magnetic anomaly simulation signal obtained in S1, the underwater magnetic anomaly data is decomposed by using improved empirical mode decomposition, and a high-dimensional matrix is ​​constructed using the decomposed data;

[0065] S3. For the high-dimensional matrix constructed in S2, the principal component analysis method is used to reduce the dimension of the decomposed data, and the principal components with a cumulative contribution rate of not less than 95% are selected to form a new original signal. The magnetic anomaly signals masked by different noise amplitudes are extracted from the new original signal to complete the noise reduction of the underwater magnetic anomaly signal.

[0066] See also Figures 1 to 5 This embodiment is described. The magnetic declination angle in Songhua River area of ​​Harbin City, Heilongjiang Province is 3°43', the magnetic inclination angle is 68°68', and the magnetic moment of the magnetic object is 400A·m. 2 , the CPA distance is 50m, where the CPA distance is the distance between the magnetic detector and the target position point greater than 2.5 times the length of the magnetic target. Assume that the target moving speed is 10m / s, the sampling rate is 10Hz, and the magnetic field at the distance from the magnetic target r0(0,,0,50).

[0067] In this embodiment, the magnetic sensor measures the total magnetic field, which is composed of the vector superposition of the magnetic anomaly magnetic field generated by the magnetic target and the geomagnetic background magnetic field. The amplitude of the geomagnetic background magnetic field is 40000-60000nT, and the amplitude of the magnetic anomaly magnetic field signal is 0.01-1000nT according to different application scenarios. The magnetic anomaly signal generated during simulation is 0.01-1000nT, and the wave magnetic field noise and solar wind magnetic field noise are added and submerged in the geomagnetic magnetic field background, thereby forming an underwater magnetic anomaly signal that is closer to reality.

[0068] Implementation method 2. This implementation method is a further limitation of the underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition provided in implementation method 1. In S1, when the distance between the magnetic detector and the magnetic target r0 (x, y, z) is not less than 2.5 times, the magnetic target r0 (x, y, z) is a point.

[0069] This embodiment further limits the selection conditions of the magnetic target r0(x,, y, z), and limits the selection of the magnetic target r0(x,, y, z) to a magnetic target r0(x,, y, z) when the distance between the magnetic detector and the position point of the underwater boat is not less than 2.5 times.

[0070] Embodiment 3: This embodiment further limits the underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition provided in Embodiment 2. The magnetic field at the magnetic target r0 (x, y, z) is:

[0071]

[0072] Among them, M (Mx, My, Mz) is the magnetic moment vector of the magnetic target. Since the magnetic permeability of seawater is similar to the magnetic permeability of vacuum, that is, u0 takes the magnetic permeability of vacuum, r = |r| is the distance from the magnetic target to the magnetic detector. r0 is the relative position vector between the magnetic target and the magnetic detector in the spatial rectangular coordinate system.

[0073] Embodiment 4: This embodiment further limits the underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition described in embodiment 1. The magnetic field generated by the magnetic target r0 (x, y, z) in space is decomposed into a three-component form, and the formula is expressed as follows:

[0074]

[0075] Among them B x , B y , B z is the component of the magnetic field generated by the magnetic target in space along the x, y, and z axes, I is the magnetic inclination, and D is the magnetic declination. x 、r y 、r z are the relative position components between the magnetic target and the magnetic detector along the x, y, and z directions respectively.

[0076] See also Figure 3 To illustrate this implementation mode, the simulation parameters of this implementation mode are as follows: the number of noise source points is 500, the number of noise superpositions is 30, the sea water depth is 10m, the magnetic inclination is 68°68', w is a random angular frequency roughly in the range of 0.5 to 2.5Hz, and the magnetic declination is 3°43'. According to the Weaver wave model, low-frequency wave magnetic field noise with a noise range of 0.01-1nT is generated; the solar wind noise is a low-frequency non-Gaussian random noise, the magnetic field change period is 0.2 to 1000s, and the amplitude ranges from 0.1nT to tens of nT. The solar wind noise is simulated by random sequences and mathematical functions to generate solar wind noise with a frequency of 0.001-5Hz.

[0077] Specifically, the above-mentioned sea wave magnetic field and solar wind magnetic field noise are added to the magnetic anomaly signal to generate an underwater magnetic anomaly signal that is closer to reality.

[0078] Embodiment 5: This embodiment further limits the underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition provided in Embodiment 1, wherein S2 is:

[0079] S2.1: Add Nm white noises of different amplitudes to the simulated underwater magnetic anomaly signal generated in S1 to generate Nm new signals {x p (t), p = 1, 2, ..., Nm};

[0080] S2.2 uses the empirical mode decomposition method to decompose the Nm signals with white noise added, where any signal is decomposed into i components. The underwater magnetic anomaly signal can be expressed as follows:

[0081]

[0082] Among them, c p,i (t) represents the original signal X after adding white noise to the pth p (t) The useful signal components obtained by decomposition are i in total, and their frequencies are arranged from high to low, r p (t) represents the useless components added each time decomposition; p,i (t) and r p (t) are integrated and averaged to obtain the final magnetic anomaly signal, and the decomposed result is as follows:

[0083]

[0084] Among them, s i (t) is c p,i (t) is the result of integrated averaging, k(t) is r p (t) The result after integration and averaging;

[0085] S2.3: The IMF component of the magnetic anomaly signal decomposed in S2.2 above is denoted as a m , thus forming a high-dimensional magnetic anomaly characteristic matrix A, the specific form is as follows:

[0086] A=[a1,a2,a3,K,a m ].

[0087] Embodiment 6: This embodiment further limits the underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition provided in embodiment 1, and S3 is:

[0088] S3.1: According to the high-dimensional matrix constructed in S2, let a' m =a m -E(a m ), where E(a m ) is the individual sample expectation, then the high-dimensional magnetic anomaly feature matrix after conversion is as follows:

[0089] A'=[a'1,a'2,a'3,K,a' m ]

[0090] S3.2: Construct the covariance matrix C of A' through the converted high-dimensional magnetic anomaly feature matrix. The specific expression is as follows:

[0091] C=A' T×A' / (k-1)

[0092] Among them, A′ T is the transposed matrix of matrix A', k is the number of column vectors of each sample;

[0093] S3.3: Use singular value decomposition to find the eigenvalue λ and eigenvector U of the covariance matrix C. The formula is:

[0094] C=PΛP T

[0095] In the above formula: Λ is a diagonal matrix, and its element eigenvalues ​​λ are arranged in descending order. This is the eigenvalue matrix of the covariance matrix. Each column of P is the eigenvector U corresponding to the corresponding eigenvalue. This is the eigenvector matrix of the covariance matrix corresponding to the eigenvalue, P T is the rotation matrix of P; the sample data is rotated by the rotation matrix P T The linear transformation of is used to obtain the principal components arranged by variance and with decreasing energy.

[0096] Embodiment 7: This embodiment further limits the underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition provided in embodiment 1. The calculation formula used to calculate the cumulative contribution rate of the eigenvalue is as follows:

[0097]

[0098]

[0099] Where: j is the jth eigenvector, bj is the information contribution rate of the jth eigenvalue, α q is the cumulative contribution rate of the first q eigenvalues, λ l The eigenvalue λ is accumulated from L=1 to m or q.

[0100] See also Figure 5 This embodiment is described. In this embodiment, j is the information contribution rate of the jth eigenvalue, α p is the cumulative contribution rate of the first q eigenvalues. The cumulative contribution rate of the first 8 eigenvalues ​​reaches 97.37%, so q=8.

[0101] Embodiment 8: This embodiment further limits the underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition provided in embodiment 1. The formula for constructing the principal component composed of the eigenvectors is:

[0102] i new principal components y composed of eigenvectors i The principal component matrix Δ can be defined as:

[0103] Δ=PT A

[0104] The specific expansion form is:

[0105]

[0106] Among them, u 1j -u mj is the eigenvalue corresponding to each eigenvector in the m principal components.

[0107] See also Figure 4 To illustrate the present embodiment, the present embodiment describes selecting a combination of principal components with a cumulative contribution rate of more than 95%, performing EMD decomposition on the composed new original signal, obtaining completely orthogonal IMF components, eliminating the modal aliasing effect between the components, and further separating the magnetic anomaly signals and noise with different characteristics into different IMF components.

[0108] Where i = m. Select the first 8 principal components for signal reconstruction to generate a new orthogonal original signal, where Figure 6 It is a comparison chart of the new orthogonal original signal and the original signal. Figure 7 This is a comparison chart of the new orthogonal original signal and the original signal after EEMD decomposition, which is reconstructed based on the first 8 IMFs selected based on experience. Figure 6 and Figure 7 We can clearly see that the new original signal constructed by the improved empirical mode decomposition has good feature retention. Except for the obvious noise that has not been filtered out at t=40s, that is, near the time point when the magnetic anomaly occurs, there is a significant denoising effect in other areas of the time axis.

[0109] The signal effect after EMD reconstruction is evaluated in terms of accuracy, precision, and noise control, such as Figure 8 The accuracy can be shown by the correlation coefficient, the precision can be measured by the root mean square error, and the noise control can be measured by the signal-to-noise ratio.

[0110] Depend on Figure 8It can be seen that the method proposed in the present invention can obtain a larger signal-to-noise ratio and a larger correlation coefficient with the original analog signal than the two preprocessing methods of EMD and EEMD, which reflects that the proposed method has better noise control and accuracy than the other two methods, and also has the smallest root mean square error with the original analog signal, which reflects that the accuracy of the signal reconstructed by the proposed method is greatly improved compared with the other two methods. Therefore, from the perspectives of accuracy, precision, and noise control of the reconstructed signal, the underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition proposed in the present invention has the best denoising effect and enhances the ability to acquire underwater magnetic anomaly signals. The present invention verifies the application value of the improved empirical mode decomposition method in underwater magnetic anomaly signal filtering, and the established model is also of reference significance for other types of simulation analysis and has universal applicability.

[0111] Embodiment 9: This embodiment proposes a computer device including a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes any one of the methods described above.

[0112] Embodiment 10: This embodiment proposes a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the method described in any one of embodiments 1 to 8 are implemented.

[0113] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of the systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram may represent a module, a program segment, or a portion of a code, and the above-mentioned module, program segment, or a portion of a code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the boxes may also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be used to perform the specified functions.

[0114] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0115] Although preferred embodiments of the present invention have been described, additional changes and modifications may be made to these embodiments by those skilled in the art once the basic inventive concepts are known. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention. Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for denoising underwater magnetic anomaly signals based on improved empirical mode decomposition, characterized in that: The following steps are included: S1. Establish a magnetic target model, take the underwater boat as the magnetic target r0(x, y, z), add the wave magnetic field noise and the solar wind magnetic field noise to the magnetic target r0(x, y, z), and obtain the underwater magnetic anomaly simulation signal; S2, according to the underwater magnetic anomaly simulation signal obtained in S1, the underwater magnetic anomaly data is decomposed by using improved empirical mode decomposition, and a high-dimensional matrix is ​​constructed using the decomposed data; S3. For the high-dimensional matrix constructed in S2, the principal component analysis method is used to reduce the dimension of the decomposed data, and the principal components with a cumulative contribution rate of not less than 95% are selected to form a new original signal. The magnetic anomaly signals masked by different noise amplitudes are extracted from the new original signal to complete the noise reduction of the underwater magnetic anomaly signal.

2. The underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition according to claim 1 is characterized in that: In S1, when the distance between the magnetic detector and the magnetic target r0(x, y, z) is not less than 2.5 times, the magnetic target r0(x, y, z) is a point.

3. The underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition according to claim 2 is characterized in that: The magnetic field at the magnetic target r0 (x, y, z) is: Wherein, M (Mx, My, Mz) is the magnetic moment vector of the magnetic target, u0 is the vacuum magnetic permeability, r = |r| is the distance from the magnetic target to the magnetic detector, and r0 is the relative position vector between the magnetic target and the magnetic detector in the spatial rectangular coordinate system.

4. The underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition according to claim 1 is characterized in that: The magnetic field B is decomposed into three components: Among them B x , B y , B z is the component of the magnetic field generated by the magnetic target in space along the x, y, and z axes, I is the magnetic inclination, D is the magnetic declination, r x 、r y 、r z are the relative position components between the magnetic target and the magnetic detector along the x, y, and z directions respectively.

5. The underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition according to claim 1 is characterized in that: The S2 is: S2.1: Add Nm white noises of different amplitudes to the simulated underwater magnetic anomaly signal generated in S1 to generate Nm new signals {x p (t), p = 1, 2, ..., Nm}; S2.2: Use the empirical mode decomposition method to decompose the Nm new signals respectively, where any signal is decomposed into i components. The underwater magnetic anomaly signal can be expressed as follows: Among them, c p,i (t) represents the original signal X after adding white noise to the pth p (t) The useful signal components obtained by decomposition are i in total, and their frequencies are arranged from high to low, r p (t) represents the useless components added each time decomposition, and c p,i (t) and r p (t) are integrated and averaged to obtain the final magnetic anomaly signal, and the decomposed result is as follows: Among them, s i (t) is c p,i (t) is the result of integrated averaging, k(t) is r p (t) The result after integration and averaging; S2.3: The IMF component of the magnetic anomaly signal decomposed in S2.2 above is denoted as a m , forming a high-dimensional magnetic anomaly feature matrix A: <h2 style=";text-align:left;direction:ltr">A = [a1,a2,a3,K,a<h2 style=";text-align:left;direction:ltr"> m <h2 style=";text-align:left;direction:ltr"> ]。 6. The underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition according to claim 1 is characterized in that: The S3 is: S3.1: According to the high-dimensional matrix constructed in S2, let a' m =a m -E(a m ), where E(a m ) is the individual sample expectation, then the high-dimensional magnetic anomaly feature matrix after conversion is as follows: A'=[a'1,a'2,a'3,K,a' m ] S3.2: Construct the covariance matrix C of A' through the converted high-dimensional magnetic anomaly feature matrix. The specific expression is as follows: C=A' T ×A' / (k-1) Among them, A′ T is the transposed matrix of matrix A', k is the number of column vectors of each sample; S3.3: Use singular value decomposition to find the eigenvalue λ and eigenvector U of the covariance matrix C. The formula is: C=PΛP T In the above formula: Λ is a diagonal matrix, and its element eigenvalues ​​λ are arranged in descending order. This is the eigenvalue matrix of the covariance matrix. Each column of P is the eigenvector U corresponding to the corresponding eigenvalue. This is the eigenvector matrix of the covariance matrix corresponding to the eigenvalue, P T is the rotation matrix of P; the sample data is rotated by the rotation matrix P T The linear transformation of is used to obtain the principal components arranged by variance and with decreasing energy.

7. The underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition according to claim 1 is characterized in that: The calculation formula used to calculate the cumulative contribution rate of the eigenvalue is as follows: Where: j is the jth eigenvector, b j is the information contribution rate of the jth eigenvalue, α q is the cumulative contribution rate of the first q eigenvalues, λ l The eigenvalue λ is accumulated from L=1 to m or q.

8. The underwater magnetic anomaly signal denoising method based on improved empirical mode decomposition according to claim 1 is characterized in that: The formula for constructing the principal component composed of eigenvectors is: i new principal components y composed of eigenvectors i The principal component matrix Δ can be defined as: Δ=P T A The specific expansion form is: Among them, u 1j -u mj is the eigenvalue corresponding to each eigenvector in the m principal components.

9. A computer device, comprising a memory and a processor, characterized in that: A computer program is stored in the memory. When the processor runs the computer program stored in the memory, the processor executes the method according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images