Extremely low frequency ocean electromagnetic signal detection method based on dynamic Kalman filtering

Abstract

The present invention discloses a very low frequency ocean electromagnetic signal detection method based on dynamic Kalman filtering, including: obtaining a noisy very low frequency ocean electromagnetic signal time series; inputting the noisy very low frequency ocean electromagnetic signal time series into a dynamic Kalman filter to obtain a filtered electromagnetic signal time series; performing a short-time Fourier transform on the filtered electromagnetic signal time series to obtain a corresponding denoised time spectrum; extracting characteristic parameters of the very low frequency ocean electromagnetic signal in the denoised time spectrum. The present invention introduces historical state information into the filter, which can effectively suppress typical noise interference including wave induced noise and low-frequency random noise, greatly improves the overall signal-to-noise ratio of ocean electromagnetic data in the very low frequency band, and realizes enhanced detection of the target very low frequency signal baseband based on time-frequency domain feature correlation analysis, which has broad application prospects and value in the field of marine remote sensing information perception.

Description

Technical Field

[0001] The present invention relates to the technical field of signal detection, and in particular to an extremely low frequency ocean electromagnetic signal detection method based on dynamic Kalman filtering. Background Art

[0002] In order to resist the corrosion of seawater, ships are usually equipped with cathodic protection measures. The resulting protection current is modulated by the propeller to form a time-harmonic alternating current with the propeller's rotation frequency as the fundamental frequency, which then forms a target extremely low frequency electromagnetic field with significant harmonic characteristics. As an inherent physical property of a ship target, the extremely low frequency electromagnetic field has good penetration in seawater, slow propagation attenuation, and carries obvious target motion signal characteristics, which can serve as an important basis for long-distance detection and identification of marine targets.

[0003] Although the propagation attenuation of extremely low frequency electromagnetic signals is slow, the actual ocean environment contains various types of electromagnetic noise interference, which seriously reduces the signal-to-noise ratio of the measured data, bringing great difficulties to the long-distance detection of extremely low frequency signals. As the main noise sources that have a greater impact on the detection of extremely low frequency electromagnetic signals, wave induced noise and low-frequency random noise both have time-varying and random characteristics, and their generation mechanisms and signal domain characteristics are different.

[0004] In view of the impact of wave-induced noise and low-frequency random noise on the detection of extremely low frequency electromagnetic signals, it is necessary to suppress noise from the perspective of signal processing to improve the signal-to-noise ratio of the data and provide an effective method for long-distance detection of target extremely low frequency electromagnetic signals. Most of the existing marine electromagnetic noise suppression technologies are based on the characteristic differences between noise and signal in the frequency domain to achieve noise filtering, and have achieved relatively ideal signal detection effects in many problems. However, the premise of the existing methods is based on the separability of noise and signal characteristics in the frequency domain, which is difficult to be directly used for the detection of extremely low frequency signals. The spectrum of wave-induced noise is a broadband spectrum, similar to the spectrum of low-frequency random noise, both covering the extremely low frequency band where the target signal is located. It is difficult to effectively separate the target signal and noise through frequency domain filtering technology. Therefore, it is necessary to propose an effective denoising algorithm based on the characteristics of extremely low frequency electromagnetic signals and marine environmental noise to improve the detection capability of extremely low frequency electromagnetic signals under low signal-to-noise ratio conditions. Summary of the invention

[0005] The embodiment of the present invention provides an extremely low frequency ocean electromagnetic signal detection method based on dynamic Kalman filtering, so as to solve the problem that the existing denoising technology is difficult to be directly used for the detection of extremely low frequency signals due to the separability of noise and signal in the frequency domain.

[0006] On the one hand, an embodiment of the present invention provides an extremely low frequency ocean electromagnetic signal detection method based on dynamic Kalman filtering, comprising:

[0007] Obtain the time series of noisy extremely low frequency ocean electromagnetic signals;

[0008] Inputting the noisy extremely low frequency ocean electromagnetic signal time series into a dynamic Kalman filter to obtain a filtered electromagnetic signal time series;

[0009] Perform short-time Fourier transform on the filtered electromagnetic signal time series to obtain the corresponding denoised time-frequency spectrum;

[0010] Extract characteristic parameters of extremely low frequency ocean electromagnetic signals in the denoising spectrum;

[0011] Among them, the dynamic Kalman filter filtering processing of the noisy extremely low frequency ocean electromagnetic signal time series includes:

[0012] Dynamic Kalman filter is used to update the system state and covariance of the noisy extremely low frequency ocean electromagnetic signal time series;

[0013] Determine the optimal estimate of the current state based on the updated system state and covariance;

[0014] Update the covariance corresponding to the best estimate.

[0015] The extremely low frequency ocean electromagnetic signal detection method based on dynamic Kalman filtering in the present invention has the following advantages:

[0016] 1. The present invention combines the time domain and time-frequency domain analysis methods of the signal to study the characteristics of ocean electromagnetic noise, and explores the separable properties of ocean electromagnetic noise and extremely low frequency electromagnetic signals. Based on this, a dynamic filtering algorithm for electromagnetic noise based on the Kalman model is proposed, which provides a new idea for the problem of enhanced detection of weak signals in the case of time-frequency feature aliasing.

[0017] 2. The filtering algorithm proposed in the present invention has strong robustness and can dynamically adjust and optimize the filtering model parameters according to changes in the environment. Compared with traditional time-frequency domain filtering technology, it is less dependent on prior knowledge of signals and noise, and is suitable for long-distance target detection scenarios with time-varying and random marine electromagnetic interference.

[0018] 3. The present invention has been verified on measured marine electromagnetic data. The proposed method can successfully extract the extremely low frequency signal of the target of interest under negative signal-to-noise ratio conditions. At the same time, the designed iterative filtering algorithm has high computational efficiency and can realize real-time processing of marine electromagnetic data. It has certain application value and potential in actual detection problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 A flow chart of a very low frequency ocean electromagnetic signal detection method based on dynamic Kalman filtering provided by an embodiment of the present invention;

[0021] Figure 2 The measured ocean noisy electromagnetic signal time series provided by the embodiment of the present invention;

[0022] Figure 3 A spectrum diagram of a measured ocean noisy electromagnetic signal provided by an embodiment of the present invention;

[0023] Figure 4 A time-frequency diagram of a measured ocean noisy electromagnetic signal provided by an embodiment of the present invention;

[0024] Figure 5 A schematic diagram of a time domain impulse signal provided by an embodiment of the present invention;

[0025] Figure 6 An impact time-frequency diagram corresponding to a time-domain impact signal provided by an embodiment of the present invention;

[0026] Figure 7 The time series of the ocean noisy electromagnetic signal in the first experiment provided by the embodiment of the present invention;

[0027] Figure 8 A time series of electromagnetic signals after filtering in Experiment 1 provided by an embodiment of the present invention;

[0028] Fig. 9 The time-frequency diagram of the ocean noisy electromagnetic signal in Experiment 2 provided by the embodiment of the present invention;

[0029] Fig.10 This is a time-frequency diagram of the electromagnetic signal after filtering in Experiment 2 provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] Figure 1A flowchart of a very low frequency ocean electromagnetic signal detection method based on dynamic Kalman filtering provided by an embodiment of the present invention. An embodiment of the present invention provides a very low frequency ocean electromagnetic signal detection method based on dynamic Kalman filtering, the method comprising:

[0032] S100, obtaining a time series of noisy extremely low frequency ocean electromagnetic signals.

[0033] For example, the acquired noisy extremely low frequency ocean electromagnetic signal time series can be recorded as x(m), such as Figure 2 shown.

[0034] S120, inputting the noisy extremely low frequency ocean electromagnetic signal time series into a dynamic Kalman filter to obtain a filtered electromagnetic signal time series.

[0035] For example, for a time series x(m) containing weak extremely low frequency electromagnetic signals, it is input into the dynamic Kalman filter in sampling order, and the filtered electromagnetic signal time series is obtained through iterative calculation.

[0036] S140, performing short-time Fourier transform on the filtered electromagnetic signal time series to obtain a corresponding denoised time-frequency spectrum.

[0037] For example, the filtered electromagnetic signal time series Perform a short-time Fourier transform (STFT) to obtain the corresponding denoised time-frequency spectrum

[0038]

[0039] Among them, w(m) is the window function, n is its sample offset, and M is the length of the noisy extremely low frequency ocean electromagnetic signal time series x(m). The denoised time-frequency diagram is shown in Figure 4 shown.

[0040] S160, extracts characteristic parameters of extremely low frequency ocean electromagnetic signals in the denoised spectrum.

[0041] For example, based on the denoised time spectrum Extracting the frequency of enhanced extremely low frequency electromagnetic signals using image visualization analysis methods and the starting time t of its appearance in the observed time series s and the deadline t e , as the detection result of the ocean's extremely low frequency electromagnetic signals, and then provide a basis for the accurate perception of target information.

[0042] Among them, S120 specifically includes:

[0043] S121, using a dynamic Kalman filter to update the system state and covariance of the noisy extremely low frequency ocean electromagnetic signal time series.

[0044] Exemplarily, the state equation x(k) and observation equation z(k) of the dynamic Kalman filter are first established:

[0045] x(k)=Ax(k-1)+Bu(k)+w(k)

[0046] z(k)=Hx(k)+v(k)

[0047] Wherein, x(k), u(k) and w(k) are the state, control quantity and process noise of the system at time k, z(k) and v(k) are the observation value and measurement noise of the system at time k, A and B are system parameters, and H is a parameter of the measurement system. In the embodiment of the present invention, w(k) and v(k) are assumed to be Gaussian white noise, and the corresponding covariances Q and R do not change with the change of the system state.

[0048] Before filtering, it is also necessary to set the values ​​of the noise covariance parameters Q and R in the observation and measurement process, and initialize the system state x(1|1) and the corresponding covariance P(1|1). In order to reduce the fluctuations in the initial process after filtering, the initial system state x(1|1) is selected as the mean of the electromagnetic signal time series. At the same time, in order to ensure the convergence of the iterative algorithm, the covariance P(1|1) of the initial state of the filter is not 0.

[0049] Next, iterate at time k. Let k = 2, and use the established dynamic Kalman filter to predict the current state x(k|k-1) and the corresponding covariance P(k|k-1) based on the state of the system at the previous moment, that is, x(1|1), to obtain the prediction result of the current state of the system:

[0050] x(k|k-1)=Ax(k-1|k-1)+Bu(k)

[0051] Among them, x(k|k-1) is the result of prediction using the state at the previous moment, and x(k-1|k-1) is the optimal estimate of the state at the previous moment.

[0052] Then update the value of the covariance P(k|k-1) of x(k|k-1):

[0053] P(k|k-1)=AP(k-1|k-1)A T +Q

[0054] Among them, P(k-1|k-1) is the covariance corresponding to x(k-1|k-1), A T Represents the transposed matrix of A.

[0055] S122, determining an optimal estimate of the current state according to the updated system state and covariance.

[0056] Exemplarily, the optimal estimated value x(k|k) at the current moment is calculated by combining the state prediction result at the current moment and the actual measurement result z(k) obtained by the electromagnetic sensor:

[0057] x(k|k)=x(k|k-1)+K g (k)[z(k)-Hx(k|k-1)]

[0058] Among them, the Kalman gain term K at the current moment g (k) Calculate as follows:

[0059] K g (k)=P(k|k-1)H T / [HP(k|k-1)H T +R]

[0060] S123, updating the covariance corresponding to the optimal estimated value.

[0061] Exemplarily, the covariance P(k|k) of the optimal estimate x(k|k) at the current moment is updated to prepare for the next round of iterative operations:

[0062] P(k|k)=[IK g (k)H]P(k|k-1)

[0063] Where I is the identity matrix.

[0064] After completing the above update process, let k=k+1, and repeat the above steps until k is greater than the length M of the noisy extremely low frequency ocean electromagnetic signal time series x(m), and the iterative process of the dynamic Kalman filter ends.

[0065] In a possible embodiment, after S160, the method further includes:

[0066] S180, performing fast Fourier transform and short-time Fourier transform on the noisy extremely low frequency ocean electromagnetic signal time series to obtain a corresponding noisy spectrum and a noisy time-frequency spectrum.

[0067] For example, for Figure 2 The electromagnetic signal time series x(m) shown in the figure is transformed by Fast Fourier Transform (FFT) to obtain its noisy spectrum X(k):

[0068]

[0069] The noisy spectrum diagram drawn based on the noisy spectrum calculation result X(k) is as follows Figure 3 shown.

[0070] S200, combining the noisy spectrum and the noisy time-frequency spectrum, determines the occurrence time of the wave-induced noise and the energy frequency band range of the low-frequency random noise.

[0071] For example, by combining the above noisy spectrum and noisy time-frequency diagram, the time-frequency correlation characteristics of the wave-induced noise in the observation data segment are analyzed, and the occurrence time t of each wave-induced noise interference is estimated. i ,i=1,2,…,N. At the same time, the time-frequency distribution characteristics of the ocean low-frequency random noise are analyzed by combining the noisy spectrum and the noisy time-frequency spectrum, and the main frequency band range of its energy is estimated [f min ,f max ].

[0072] S220, establishing a time domain impulse signal.

[0073] For example, the simulation is established as follows Figure 5 The time domain impulse signal h(t) is shown as 5δ(t-4)-5δ(t-6).

[0074] S240, performing short-time Fourier transform on the time-domain impulse signal to obtain a corresponding impulse time-frequency spectrum.

[0075] Exemplarily, a short-time Fourier transform is performed on the time-domain impulse signal to obtain the impulse time spectrum H(t,Ω):

[0076] H(t,Ω)=∫h(τ)g(τ-t)e -jΩτ dτ

[0077] Among them, g(τ) is the window function, τ is the integral variable, t and Ω are the shift variable and frequency variable respectively. The impact time-frequency diagram is drawn according to the impact time-frequency calculation result H(t,Ω) as shown in Figure 6 shown.

[0078] S260, determining the amplitude of the wave induced noise according to the frequency spectrum of the impact.

[0079] Exemplarily, according to the established time domain impulse signal and impulse time spectrum, the amplitude A of each wave induced noise appearing in the noisy extremely low frequency ocean electromagnetic signal time series x(m) is estimated. i ,i=1,2,…,N。

[0080] The embodiment of the present invention also provides an extremely low frequency ocean electromagnetic signal detection system, including:

[0081] A sequence acquisition module is used to obtain the time series of noisy extremely low frequency ocean electromagnetic signals;

[0082] A filtering module, used for inputting the noisy extremely low frequency ocean electromagnetic signal time series into a dynamic Kalman filter to obtain a filtered electromagnetic signal time series;

[0083] A short-time Fourier transform module is used to perform short-time Fourier transform on the filtered electromagnetic signal time series to obtain the corresponding denoised time spectrum;

[0084] Parameter extraction module, used to extract characteristic parameters of extremely low frequency ocean electromagnetic signals in the denoising spectrum;

[0085] Among them, the filtering module includes:

[0086] The system update submodule is used to update the system state and covariance of the noisy extremely low frequency ocean electromagnetic signal time series using a dynamic Kalman filter;

[0087] The optimal estimation submodule is used to determine the optimal estimation value of the current state according to the updated system state and covariance;

[0088] The covariance updating submodule is used to update the covariance corresponding to the optimal estimate.

[0089] In a possible embodiment, it also includes: a signal conversion module, which is used to perform fast Fourier transform and short-time Fourier transform on the time series of noisy extremely low frequency ocean electromagnetic signals to obtain corresponding noisy spectrum and noisy time-frequency spectrum; a noise determination module, which is used to determine the occurrence time of wave-induced noise and the energy frequency band range of low-frequency random noise by combining the noisy spectrum and the noisy time-frequency spectrum; a signal establishment module, which is used to establish a time-domain impact signal; an impact transformation module, which is used to perform short-time Fourier transform on the time-domain impact signal to obtain a corresponding impact time-frequency spectrum; an amplitude determination module, which is used to determine the amplitude of the wave-induced noise according to the impact time-frequency spectrum.

[0090] An embodiment of the present invention further provides a computer-readable storage medium, in which a plurality of computer instructions are stored. The plurality of computer instructions are used to enable a computer to execute the above method.

[0091] The effect of the present invention is further illustrated by the following simulation experiment:

[0092] (1) Experimental simulation conditions:

[0093] The data used in this experiment is a typical ocean electromagnetic signal time series. The data was collected by an inductive electromagnetic sensor at Fengjiahe Wharf in Qingdao City, Shandong Province in August 2018. The total data length is 1 hour and the sampling rate is 200Hz. It contains wave induced noise and low-frequency random noise, as well as the target extremely low frequency harmonic signal of interest. The fundamental frequency band where the extremely low frequency harmonic signal is located is severely interfered by low-frequency random noise. At the same time, the induced noise accompanying the waves will also submerge the extremely low frequency harmonic signal, resulting in a low overall signal-to-noise ratio of the electromagnetic observation data, which poses certain challenges to the detection of extremely low frequency signals. This experiment was conducted on a 3.20GHz main frequency. Core TM The experiment was carried out on a 64-bit Windows 10 operating system with an i7-8700 CPU and 16GB of RAM, based on MATLAB R2013b software.

[0094] (2) Noise suppression evaluation criteria:

[0095] (2a) Time domain evaluation criteria: peak-to-peak reduction ratio η

[0096] The peak-to-peak reduction ratio is defined as follows:

[0097]

[0098] in, is the peak-to-peak value of the noise in the time domain before filtering, is the peak-to-peak value of the noise in the time domain after filtering. The larger the value of η, the more the peak-to-peak value of the noise after filtering decreases compared to before filtering, indicating that the noise suppression effect in the time domain is better.

[0099] (2b) Frequency domain evaluation criteria: Noise band upper limit δ

[0100] The upper limit of the noise band is defined as follows:

[0101]

[0102] Where H(f) is the spectrum of noisy data, f up is the frequency upper limit variable. The above formula finds the minimum value of the frequency upper limit that satisfies the inequality in the brackets and defines it as the noise band upper limit δ. The larger the noise band upper limit δ is, the wider the spectrum energy distribution range of the low-frequency random ocean noise is, and the more serious the interference to the extremely low-frequency signal band is.

[0103] (3) Experimental content:

[0104] Experiment 1

[0105] When the system parameters A and B are 1 and 0, the measurement system parameter H is 1, and the covariances Q and R of the observation and measurement process noise are 1×10-6 and 4×10 -4 Under the conditions that the initial state x(1|1) of the filter is set to 0.0406nT and the covariance P(1|1) of the initial state is set to 1, the dynamic Kalman filter designed by the present invention is used to filter the time series of extremely low frequency ocean electromagnetic signals. The experimental results are as follows: Figure 7-8 As shown, Figure 7 is the noisy electromagnetic signal time series, Figure 8 is the time series of the filtered electromagnetic signal.

[0106] from Figure 7-8 It can be seen that the present invention can achieve better noise time domain energy suppression effect by dynamically correcting the measurement results in combination with the historical state.

[0107] The peak-to-peak value and peak-to-peak reduction ratio of the noisy electromagnetic data and the filtered electromagnetic data in the time domain are calculated respectively. The calculation results are shown in Table 1.

[0108] Table 1 Peak-to-peak value and peak-to-peak reduction ratio of electromagnetic data in the time domain after noise and filtering

[0109] index Peak-to-peak value of noisy data Peak-to-peak value of filtered data Peak-to-peak drop Peak-to-peak reduction ratio Numeric 53.0022nT 38.0096nT 14.9926nT 0.2829

[0110] As can be seen from Table 1, the peak-to-peak value of the filtered electromagnetic data is reduced by 14.9926nT compared with the noisy electromagnetic data, and the peak-to-peak value reduction ratio is 0.2829. The present invention can effectively suppress the energy of ocean noise in the time domain through a dynamic noise filtering algorithm, and provide support for the detection of extremely low frequency signals under the interference of wave induced noise.

[0111] Experiment 2

[0112] The noisy electromagnetic data is subjected to short-time Fourier transform to obtain its noisy time-frequency diagram, and compared with the denoised time-frequency diagram of the electromagnetic data filtered by the method of the present invention. The experimental results are as follows: Figure 9-10 As shown, Fig. 9 is the noisy time-frequency plot of the noisy electromagnetic data, Fig.10 It is a denoised time-frequency diagram of electromagnetic data after filtering by the method of the present invention.

[0113] from Figure 9-10 It can be seen that the present invention can effectively suppress low-frequency random noise and successfully detect the 3 Hz weak extremely low-frequency fundamental signal submerged by the noise, thereby realizing accurate perception of the target signal under low signal-to-noise ratio conditions.

[0114] The noise frequency band upper limits of the filtered data and noisy electromagnetic data are compared, and the numerical index results are shown in Table 2.

[0115] Table 2 Comparison of noise band upper limit index between filtered data and noisy electromagnetic data by the method of the present invention

[0116] Data to be evaluated The data after filtering by the method of the present invention Noisy EM Data Noise band upper limit δ 2Hz 5Hz

[0117] As can be seen from Table 2, the upper limit of the noise frequency band of the filtered data of the method of the present invention is 2Hz, which is 3Hz lower than the original noisy electromagnetic data. The present invention can effectively suppress the energy of ocean noise in the frequency domain through a dynamic noise filtering method, and provide support for the detection of extremely low frequency signals under the interference of low frequency random noise.

[0118] In summary, the present invention can effectively suppress typical noise interference including wave induced noise and low-frequency random noise by introducing historical state information into the filter, greatly improving the overall signal-to-noise ratio of ocean electromagnetic data in the extremely low frequency band, and realizing enhanced detection of the target extremely low frequency signal fundamental frequency based on time-frequency domain feature correlation analysis. It has broad application prospects and value in the field of marine remote sensing information perception.

[0119] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. 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.

[0120] Obviously, those skilled in the art can 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 very low frequency ocean electromagnetic signal detection method based on dynamic Kalman filtering, characterized in that: include: Obtain the time series of noisy extremely low frequency ocean electromagnetic signals; Inputting the noisy extremely low frequency ocean electromagnetic signal time series into a dynamic Kalman filter to obtain a filtered electromagnetic signal time series; Performing short-time Fourier transform on the filtered electromagnetic signal time series to obtain a corresponding denoised time-frequency spectrum; Extracting characteristic parameters of extremely low frequency ocean electromagnetic signals in the denoised spectrum; The filtering process of the dynamic Kalman filter on the noisy extremely low frequency ocean electromagnetic signal time series includes: Using a dynamic Kalman filter to update the system state and covariance of the noisy extremely low frequency ocean electromagnetic signal time series; Determine the optimal estimate of the current state based on the updated system state and covariance; Updating the covariance corresponding to the optimal estimate; After extracting the characteristic parameters of the extremely low frequency ocean electromagnetic signal in the denoising spectrum, the method further includes: Performing fast Fourier transform and short-time Fourier transform on the noisy extremely low frequency ocean electromagnetic signal time series to obtain a corresponding noisy spectrum and a noisy time-frequency spectrum; Combining the noisy spectrum and the noisy time-frequency spectrum, determining the occurrence time of the wave-induced noise and the energy frequency band range of the low-frequency random noise; Establish time domain impact signal; Performing short-time Fourier transform on the time-domain impulse signal to obtain a corresponding impulse time-frequency spectrum; The amplitude of the wave induced noise is determined according to the impact time spectrum.

2. The method for detecting extremely low frequency ocean electromagnetic signals based on dynamic Kalman filtering according to claim 1 is characterized in that: When the dynamic Kalman filter is used to filter the noisy extremely low frequency ocean electromagnetic signal time series, the moment is iterated, and the optimal estimate and the corresponding covariance are updated cyclically until the moment after the iteration is greater than the length of the noisy extremely low frequency ocean electromagnetic signal time series.

3. The method for detecting extremely low frequency ocean electromagnetic signals based on dynamic Kalman filtering according to claim 1 is characterized in that: The characteristic parameters include: the frequency and the start and end time of the extremely low frequency ocean electromagnetic signal.

Images