Lightning electromagnetic pulse cluster signal recognition and classification method based on symplectic geometry-mcODM

Abstract

The present invention discloses a lightning electromagnetic pulse cluster signal recognition and classification method based on symplectic geometry-mcODM, comprising the following steps: S1. Collection of lightning electromagnetic pulse cluster signals; S2. Projection of lightning electromagnetic signals in symplectic geometry space; S3. Classification of lightning electromagnetic pulse cluster signals; S4. Inputting lightning electromagnetic signals to be classified into symplectic geometry-mcODM, and classifying lightning using the model built in step S3. The present invention has new improvements and innovations in signal processing and classification methods, and can improve the classification accuracy and robustness of lightning electromagnetic pulse cluster signals.

Description

Technical Field

[0001] The present invention belongs to the technical field of lightning magnetic signal processing, and more specifically, relates to a lightning electromagnetic pulse cluster signal recognition and classification method based on symplectic geometry-mcODM. Background Art

[0002] Lightning is one of the major serious natural disasters, which can cause forest and oil depot fires, power supply and communication information system failures or damage, and pose a major threat to aerospace, mining and some important and sensitive high-tech equipment. After the 1980s, the hazards caused by lightning increased significantly, especially in areas closely related to high-tech, such as aerospace, national defense, communications, electricity, computers, and the electronics industry. The widespread use of large-scale and ultra-large-scale integrated circuits that are extremely sensitive to lightning electromagnetic interference has greatly increased the probability of being struck by lightning.

[0003] The study of lightning transmission mechanism can be applied to lightning protection technology. Electromagnetic pulse clusters are a common lightning discharge phenomenon in the process of lightning transmission, which characterizes different transmission states of lightning. Broadband interferometer three-dimensional observation of negative ground-to-ground lightning without pulse cluster events (Plateau Meteorology, 2013). Pulse clusters can be classified based on different waveform characteristics, such as regular pulse trains (RPT) and irregular pulse trains (CPT). In the complex transmission process of lightning, regular and irregular pulses are usually staggered to form new types of pulse clusters, such as the symmetrical pulse clusters mentioned in "Characteristics of Regular Pulse Bursts Generated From Lightning Discharges" (Frontiers, 2021), and Wang Pingting mentioned in her master's thesis "Analysis of Low-frequency Magnetic Field Signal Characteristics of Lightning Electromagnetic Pulse Clusters" (2019): (1) RPT ending with CPT, (2) RPT starting with CPT, (3) RPT with CPT in the middle, and (4) four mixed pulse clusters of RPT starting and ending with CPT. The above are all pulse clusters formed during the transmission of intra-cloud lightning. When ground lightning occurs, artificial lightning, upward leaders of high-rise buildings and towers can also produce similar pulse clusters "Characteristics of Electromagnetic Signals During the Initial Stage of Negative Rocket-Triggered Lightning" (Journal of Geophysical Research, 2018).

[0004] What lightning transmission characteristics do different types of lightning pulse clusters correspond to? How do these characteristics guide us to take active lightning protection measures?

[0005] From the current perspective, to solve the above problems, we must first extract different types of pulse cluster signals from the lightning electromagnetic model. However, this work is rather cumbersome if done manually, and the waveform characteristics may be interfered by environmental noise signals, making it difficult to identify.

[0006] Therefore, it is urgent to design a method for identifying and classifying electrical pulse cluster signals. Summary of the invention

[0007] The purpose of the present invention is to solve the problems existing in the prior art. Therefore, the present invention provides a method for identifying and classifying lightning electromagnetic pulse cluster signals based on symplectic geometry-mcODM, including obtaining lightning electromagnetic signals by setting up a high-sensitivity single-station low-frequency magnetic field measurement system composed of an outdoor antenna and an indoor acquisition system. The signal is projected into the symplectic geometry space, and the trajectory form and signal form of the signal are analyzed to enhance the characteristic expression ability of the lightning electromagnetic pulse cluster in the signal. Then, according to the boundary conditions of regular pulse clusters (RPT) and irregular pulse clusters (CPT), the model parameters that maximize the boundary between RPT and CPT are solved to enhance the classification ability of lightning electromagnetic pulse clusters. The lightning electromagnetic pulse cluster signal classification method based on symplectic geometry-mcODM provided by the present invention is suitable for single-base lightning observation stations in complex scenarios.

[0008] The technical solution to achieve the purpose of the present invention is to provide a lightning electromagnetic pulse cluster signal recognition and classification method based on symplectic geometry-mcODM, comprising the following steps:

[0009] S1. Collection of lightning electromagnetic pulse cluster signals;

[0010] S2. Projection of lightning electromagnetic signals in symplectic geometric space;

[0011] S3. Classification of lightning electromagnetic pulse cluster signals;

[0012] S4. Input the lightning electromagnetic signal to be classified into the symplectic geometry-mcODM, and classify the lightning using the model built in step S3.

[0013] Furthermore, S1 specifically includes the following steps:

[0014] Weak lightning magnetic signals are obtained by setting up a high-sensitivity single-station low-frequency magnetic field measurement system consisting of an outdoor antenna and an indoor acquisition system.

[0015] The indoor acquisition system is provided with timing information by a high-precision GPS receiver, and the outdoor antenna is two horizontal orthogonal magnetic induction coils, which are connected to a chopper and amplification circuit to obtain an electromagnetic signal sample set with RPT and CPT.

[0016] Furthermore, S2 specifically includes the following steps: let the lightning electromagnetic pulse cluster signal x in the sample set = {x1, x2, …, x L}, where L is the number of sampling points of the signal; x is reconstructed through phase space to obtain:

[0017]

[0018] Where k is the embedding dimension of the phase space reconstruction of the lightning electromagnetic pulse cluster signal, τ is the time delay, and m = L-(k-1)τ; k is an important indicator that determines the quality of the phase space reconstruction of the lightning electromagnetic pulse cluster signal; the appropriate k value is determined by the power spectrum density estimation method:

[0019]

[0020] Among them, f max is the maximum peak frequency of the power spectral density estimate, f s is the sampling frequency, floor(·) is the rounding operation;

[0021] The Hamiltonian matrix W of the signal is constructed by X = [P T 0;0-P T ], where the matrix P is X T X; then the orthogonal symplectic matrix of the lightning signal can be expressed as:

[0022]

[0023] Among them, Q is an orthogonal symplectic matrix with symplectic matrix properties; since the manifold structure of the Hamiltonian matrix is ​​retained during the construction of the symplectic matrix, H is an upper triangular matrix, and its eigenvalue can be obtained by calculating the W matrix; from the Hamiltonian matrix properties, it can be known that:

[0024] λ(W)=λ(H)

[0025]

[0026] If the number of eigenvalues ​​is Then the P matrix eigenvalue σ i The relationship can be expressed by formula (5):

[0027] σ1>σ2>…>σ k (5)

[0028] make is i The eigenvector ofi 、M i They are Q i N i , then the reconstruction matrix is The reconstructed phase space matrix Z is equal to Use the diagonal mean transformation to transform the matrix Z into a matrix of length L Symplectic geometry components, The sum of the symplectic geometric components is Z;

[0029] make Where 1≤i≤m, at this time, if but otherwise The diagonal mean transfer matrix is ​​expressed as:

[0030]

[0031] Determine a set of Z i The corresponding one-dimensional array r1, r2, ..., r of length L L ; Then formula (6) can be rewritten as:

[0032]

[0033] Formula (7) is the first reconstructed SGCs component. The final SGCs are obtained by reconstructing the initial SGCs with similar frequencies. The final SGCs is expressed as:

[0034]

[0035] Update the lightning electromagnetic pulse cluster signal x in the sample set to R1.

[0036] Furthermore, S3 specifically includes the following steps: Let (x i ,y i ) is the sample in the sample set after gen is updated, where x i is the update signal of the i-th sample, y i is the label value of the RPT or CPT corresponding to the sample; then mcODM is:

[0037]

[0038] Among them, Ω(w) is the regularization term that penalizes the complexity of the model; w is the weight coefficient; η and λ are penalty functions; m is the total number of samples; is the marginal mean; i With ε i For Margin and the marginal mean; hence is the marginal difference; in order to ensure that scaling w does not affect the final classification result, let the marginal mean be 1, then the sample (x i ,y i ) is Its optimal marginal allocation machine can be re-planned as:

[0039]

[0040] Among them, μ∈(0,1] is a parameter that weighs the two deviations; θ∈(0,1] is a zero-loss segment parameter that controls the number of support vectors;

[0041] For multiple classifiers, let the regularization term be l is the scoring parameter, H is the assumed set of scoring functions; then the mathematical description of mcODM is:

[0042]

[0043] According to another aspect of the present invention, a computer-readable storage medium is provided, on which a computer program is stored. When the program is executed by a processor, the steps of the lightning electromagnetic pulse cluster signal recognition and classification method based on symplectic geometry-mcODM of the present invention are implemented.

[0044] According to another aspect of the present invention, there is provided a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method for identifying and classifying lightning electromagnetic pulse cluster signals based on symplectic geometry-mcODM of the present invention when executing the program.

[0045] Compared with the prior art, the present invention has at least the following beneficial effects:

[0046] Compared with the traditional lightning electromagnetic pulse cluster signal classification method, the present invention can improve the lightning electromagnetic pulse cluster feature quality and the lightning electromagnetic pulse cluster classification accuracy by projecting the lightning electromagnetic pulse cluster signal into the symplectic geometric space;

[0047] Compared with other algorithms, symplectic geometry-mcODM pays more attention to the boundary samples between different categories and maximizes the sample boundary distance. Therefore, symplectic geometry-mcODM has better generalization performance, adaptability, robustness, speed and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings of the embodiments will be briefly introduced below. Obviously, the drawings in the following description only relate to some embodiments of the present invention, but are not intended to limit the present invention.

[0049] Figure 1 It is a flow chart of the present invention;

[0050] Figure 2 It is a schematic diagram of the present invention for identifying regular and irregular pulse clusters of lightning electromagnetic signals. DETAILED DESCRIPTION

[0051] To make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the embodiments of the present invention will be described clearly and completely in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them.

[0052] Unless otherwise defined, technical or scientific terms used herein shall have the same meaning as commonly understood by one of ordinary skill in the art to which the invention belongs.

[0053] The following is combined with Figure 1-2 The specific embodiments of the present invention are described in detail with examples.

[0054] Example 1: Figure 1-2 As shown, the present invention provides the following technical solution, a lightning electromagnetic pulse cluster signal recognition and classification method based on symplectic geometry-mcODM, including the following contents:

[0055] S1. Collection of lightning electromagnetic pulse cluster signals;

[0056] First, a high-sensitivity single-station low-frequency magnetic field measurement system consisting of an outdoor antenna and an indoor acquisition system is set up to obtain weak lightning magnetic signals. The indoor acquisition system is provided with timing information by a high-precision GPS receiver (time accuracy is better than 50 nanoseconds). The outdoor antenna is two horizontal orthogonal magnetic induction coils, followed by a chopper and amplifier circuit. The overall 3dB bandwidth of the magnetic sensor is 40KHz-500KHz. The antenna receives the signal and transmits it to the room via an 8m shielded coaxial cable. The high-speed acquisition card performs A / D conversion and acquisition, and the sampling rate is set to 5MHz. The measurement system uses a trigger method to collect data, with a record length of 0.1s and a pre-trigger position of 20%. Thus, a large number of electromagnetic signal sample sets with RPT and CPT are obtained.

[0057] S2. Projection of lightning electromagnetic signals in symplectic geometric space;

[0058] Let the lightning electromagnetic pulse cluster signal x in the sample set = {x1, x2, …, x L}, where L is the number of sampling points of the signal. Reconstruct x through phase space to obtain:

[0059]

[0060] Where k is the embedding dimension of the phase space reconstruction of the lightning electromagnetic pulse cluster signal, τ is the time delay, and m = L-(k-1)τ. k is an important indicator that determines the quality of the phase space reconstruction of the lightning electromagnetic pulse cluster signal. The appropriate k value is usually determined by the power spectrum density estimation method:

[0061]

[0062] Among them, f max is the maximum peak frequency of the power spectral density estimate, f s is the sampling frequency, and floor(·) is the rounding operation.

[0063] The Hamiltonian matrix W of the signal is constructed by X = [P T 0;0-P T ], where the matrix P is X T X. Then the orthogonal symplectic matrix of the lightning signal can be expressed as:

[0064]

[0065] Where Q is an orthogonal symplectic matrix with symplectic matrix properties. Since the manifold structure of the Hamiltonian matrix is ​​preserved during the construction of the symplectic matrix, Q T WQ still has the relevant properties of the Hamilton matrix. H is an upper triangular matrix, and its eigenvalues ​​can be obtained by calculating the W matrix. From the properties of the Hamilton matrix, we can know that:

[0066] λ(W)=λ(H)

[0067]

[0068] If the number of eigenvalues ​​is Then the P matrix eigenvalue σ i The relationship can be expressed by formula (5).

[0069] σ1>σ2>…>σ k (5)

[0070] make is i The eigenvector of i 、M i They are Q i N i , then the reconstruction matrix is The reconstructed phase space matrix Z is equal to The matrix Z is transformed into k symplectic geometric components of length L using the diagonal mean transformation. The sum of the symplectic geometric components is Z.

[0071] make Where 1≤i≤m, at this time, if but otherwise Therefore, the diagonal mean transfer matrix can be expressed as:

[0072]

[0073] Through formula (6), we can determine a set of i The corresponding one-dimensional array r1, r2, ..., r of length L L Then formula (6) can be rewritten as:

[0074]

[0075] Formula (7) is the first reconstructed SGCs component. By reconstructing the initial SGCs with similar frequencies, the final SGCs can be obtained. The final SGCs can be expressed as:

[0076]

[0077] Update the lightning electromagnetic pulse cluster signal x in the sample set to R1.

[0078] S3. Classification of lightning electromagnetic pulse cluster signals;

[0079] In order to ensure the marginal mean value when classifying lightning electromagnetic pulse cluster signals, let (x i ,y i ) is the sample in the sample set after gen is updated, where x i is the update signal of the i-th sample, y i is the label value of the RPT or CPT corresponding to the sample. Then mcODM is:

[0080]

[0081] Among them, Ω(w) is the regularization term that penalizes the complexity of the model; w is the weight coefficient; η and λ are penalty functions; m is the total number of samples; is the marginal mean; i With ε i For Margin and the marginal mean; hence is the marginal difference. In order to ensure that scaling w does not affect the final classification result, let the marginal mean be 1, then the sample (x i ,y i ) is Its optimal marginal allocation machine can be re-planned as:

[0082]

[0083] Among them, μ∈(0,1] is a parameter that weighs the two deviations (greater than or less than the marginal mean); θ∈(0,1] is a zero-loss segment parameter that controls the number of support vectors.

[0084] For multiple classifiers, let the regularization term be l is the scoring parameter, H is the assumed set of scoring functions; then the mathematical description of mcODM is:

[0085]

[0086] S4. Input the lightning electromagnetic signal to be classified into the symplectic geometry-mcODM, and classify the lightning using the model built in step S3.

[0087] Compared with the traditional lightning electromagnetic pulse cluster signal classification method, the present invention can improve the lightning electromagnetic pulse cluster feature quality and the lightning electromagnetic pulse cluster classification accuracy by projecting the lightning electromagnetic pulse cluster signal into the symplectic geometric space;

[0088] Compared with other algorithms, symplectic geometry-mcODM pays more attention to the boundary samples between different categories and maximizes the sample boundary distance. Therefore, symplectic geometry-mcODM has better generalization performance, adaptability, robustness, speed and stability.

[0089] In summary, compared with traditional methods, this invention has new improvements and innovations in signal processing and classification methods, and can improve the classification accuracy and robustness of lightning electromagnetic pulse cluster signals.

[0090] Embodiment 2:

[0091] The computer-readable storage medium of this embodiment stores a computer program thereon, which, when executed by a processor, implements the steps in the method for identifying and classifying lightning electromagnetic pulse cluster signals based on symplectic geometry-mcODM of Example 1.

[0092] The computer-readable storage medium of this embodiment may be an internal storage unit of the terminal, such as a hard disk or memory of the terminal; the computer-readable storage medium of this embodiment may also be an external storage device of the terminal, such as a plug-in hard disk, a smart memory card, a secure digital card, a flash memory card, etc. equipped on the terminal; further, the computer-readable storage medium may also include both an internal storage unit of the terminal and an external storage device.

[0093] The computer-readable storage medium of this embodiment is used to store computer programs and other programs and data required by the terminal. The computer-readable storage medium can also be used to temporarily store data that has been output or is to be output.

[0094] Embodiment 3:

[0095] The computer device of this embodiment includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps of the lightning electromagnetic pulse cluster signal recognition and classification method based on symplectic geometry-mcODM of Example 1 are implemented.

[0096] In this embodiment, the processor may be a central processing unit, or other general-purpose processors, digital signal processors, application-specific integrated circuits, readily available programmable gate arrays or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The memory may include a read-only memory and a random access memory, and provide instructions and data to the processor. A portion of the memory may also include a non-volatile random access memory. For example, the memory may also store information about the device type.

[0097] Those skilled in the art will appreciate that the disclosed content of the embodiments may be provided as methods, systems, or computer program products. Therefore, the present solution may adopt the form of hardware embodiments, software embodiments, or embodiments combining software and hardware. Moreover, the present solution may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program codes.

[0098] The present solution is described with reference to the method according to the embodiment of the present solution and the flowchart and / or block diagram of the computer program product. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions; these computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 one or more processes and / or methods Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0099] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 one or more processes and / or methods Figure 1 A function specified in one or more boxes.

[0100] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 one or more processes and / or methods Figure 1 A step that specifies a function in one or more boxes.

[0101] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0102] The examples described in the present invention are merely descriptions of the preferred implementation modes of the present invention, and are not intended to limit the concept and scope of the present invention. Without departing from the design concept of the present invention, various modifications and improvements made to the technical solutions of the present invention by engineers and technicians in this field should all fall within the protection scope of the present invention.

Claims

1. A lightning electromagnetic pulse cluster signal recognition and classification method based on symplectic geometry-mcODM, characterized in that: The following steps are involved: S1. Collection of lightning electromagnetic pulse cluster signals; S2. Obtain the projection of lightning electromagnetic signal in symplectic geometric space by constructing signal phase space and symplectic matrix; S3. Classification of lightning electromagnetic pulse cluster signals; S4. Input the lightning electromagnetic signal to be classified into the symplectic geometry-mcODM, and classify the lightning using the model built in step S3; Wherein, step S3 specifically includes the following steps: let (x i ,y i ) is the sample in the sample set after gen is updated, where x i is the update signal of the i-th sample, y i is the label value of the RPT or CPT corresponding to the sample; then mcODM is: Among them, Ω(w) is the regularization term that penalizes the complexity of the model; w is the weight coefficient; η and λ are penalty functions; m is the total number of samples; is the marginal mean; i With ε i Marginal h (x i ,y i ) and the deviation from the marginal mean; hence is the marginal difference; in order to ensure that scaling w does not affect the final classification result, let the marginal mean be 1, then the sample (x i ,y i ) is |Υ h (x i ,y i )-1|, its optimal marginal allocation machine can be re-planned as: Among them, μ∈(0,1] is a parameter that weighs the two deviations; θ∈(0,1] is a zero-loss segment parameter that controls the number of support vectors; For multiple classifiers, let the regularization term be l is the scoring parameter, H is the assumed set of scoring functions; then the mathematical description of mcODM is:

2. The method according to claim 1, characterized in that S1 specifically includes the following steps: Weak lightning magnetic signals are obtained by setting up a high-sensitivity single-station low-frequency magnetic field measurement system consisting of an outdoor antenna and an indoor acquisition system. The indoor acquisition system is provided with timing information by a high-precision GPS receiver, and the outdoor antenna is two horizontal orthogonal magnetic induction coils, which are connected to a chopper and amplification circuit to obtain an electromagnetic signal sample set with RPT and CPT.

3. The method according to claim 1, characterized in that S2 specifically includes the following steps: let the lightning electromagnetic pulse cluster signal x in the sample set = {x1, x2, …, x L }, where L is the number of sampling points of the signal; x is reconstructed through phase space to obtain: Where k is the embedding dimension of the phase space reconstruction of the lightning electromagnetic pulse cluster signal, τ is the time delay, and m = L-(k-1)τ; k is an important indicator that determines the quality of the phase space reconstruction of the lightning electromagnetic pulse cluster signal; the appropriate k value is determined by the power spectrum density estimation method: Among them, f max is the maximum peak frequency of the power spectral density estimate, f s is the sampling frequency, floor(·) is the rounding operation; The Hamiltonian matrix W of the signal is constructed by X = [P T 0;0-P T ], where the matrix P is X T X; then the orthogonal symplectic matrix of the lightning signal is expressed as: Among them, Q is an orthogonal symplectic matrix with symplectic matrix properties; since the manifold structure of the Hamiltonian matrix is ​​retained during the construction of the symplectic matrix, H is an upper triangular matrix, and its eigenvalues ​​are obtained by calculating the W matrix; from the properties of the Hamiltonian matrix, it can be seen that: λ(W)=λ(H) If the number of eigenvalues ​​is Then the P matrix eigenvalue σ i The relationship is expressed by formula (5): σ1>σ2>…>σ k (5) make is i The eigenvector of i 、M i They are Q i N i , then the reconstruction matrix is The reconstructed phase space matrix Z is equal to Use the diagonal mean transformation to transform the matrix Z into a matrix of length L Symplectic geometry components, The sum of the symplectic geometric components is Z; make Where 1≤i≤m, at this time, if but otherwise The diagonal mean transfer matrix is ​​expressed as: Determine a set of Z i The corresponding one-dimensional array r1, r2, ..., r of length L L ; Then formula (6) can be rewritten as: Formula (7) is the first reconstructed SGCs component. The final SGCs are obtained by reconstructing the initial SGCs with similar frequencies. The final SGCs is expressed as: Update the lightning electromagnetic pulse cluster signal x in the sample set to R1.

4. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in the lightning electromagnetic pulse cluster signal recognition and classification method based on symplectic geometry-mcODM as described in any one of claims 1 to 3 are implemented.

5. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps in the lightning electromagnetic pulse cluster signal recognition and classification method based on symplectic geometry-mcODM as described in any one of claims 1 to 3 are implemented.

Images