Real-time thunder and lightning positioning method based on FPGA signal acquisition and preprocessing

Abstract

The invention discloses a real-time thunder and lightning positioning method. A plurality of observation station GNSS (Global Navigation Satellite System) synchronously acquire thunder and lightning signals to obtain waveform data; performing FFT band-pass filtering and denoising processing on the lightning signal waveform data; a sliding window dynamic threshold is adopted to extract peak value characteristics, peak value characteristics (tp, Ap and Wp) of a waveform in different time windows are obtained, a local noise standard deviation is calculated, only an effective peak value is reserved, and real-time return of lightning data is realized while positioning precision is ensured. The calculation center reconstructs the waveform based on the peak value characteristics (tp, Ap, Wp), the reconstructed waveform more accurately reflects the characteristics of the lightning signal, and the time delay measurement error caused by signal distortion is reduced. And calculating the mutual correlation of the reconstructed lightning waveform data of every two observation stations, extracting the maximum correlation time delay, calculating a TDOA equation set by adopting a weighted nonlinear least square method according to the signal time delay information, and outputting a three-dimensional lightning positioning result. And evaluating the reliability of the positioning result through residual analysis and geometric accuracy factors.

Description

Technical Field

[0001] The present invention relates to the technical fields of meteorology, atmospheric electricity and space science, and particularly relates to a real-time lightning location method based on peak waveform reconstruction and cross-correlation. Background Art

[0002] Lightning is one of the important phenomena in atmospheric electromagnetic activities and is of great significance to the research in the fields of meteorology, atmospheric electricity and space science. Lightning location is of great scientific significance for detecting the charge structure, charge density distribution in thunderstorm clouds, and their evolution with the development of thunderstorm processes, and for revealing the physical mechanisms of lightning occurrence and development processes. There are many lightning location methods in the prior art. For example, the Time of Arrival (TOA) method and the Time Difference of Arrival (TDOA) method are currently commonly used lightning location technologies. Since the lightning sampling rate is relatively high, the lightning signal sampling rate needs to be ≥10 MS / s, and the amount of sampled data for each lightning occurrence is very large. Generally, the data volume of a lightning event reaches the GB level and cannot be transmitted back in real time through a conventional communication network. Generally, it relies on local data storage and then transmits back after processing, resulting in positioning delays and unable to meet the requirements of scenarios with second-level response. The existing real-time methods only extract the peak time and amplitude for matching, but the lightning waveform has multiple peaks superimposed, and the mis-matching rate is as high as 15% - 20%, and the peak matching efficiency is low. The signals received by the detection stations usually contain noise and other interferences. Low-frequency (VLF / LF) signals are vulnerable to power frequency noise and surface charge interference, resulting in unclear peak characteristics of the signals. Lightning signals will be affected by noise, terrain occlusion and electromagnetic interference during the propagation process, resulting in signal distortion. Therefore, it is very necessary to develop a real-time lightning location method. Summary of the Invention

[0003] The purpose of the present invention is to provide a real-time lightning location method, which uses FPGA signal acquisition and preprocessing, synchronously acquires lightning signals at multiple stations through GNSS and performs FFT band-pass filtering to remove noise, extracts peak characteristics using a sliding window dynamic threshold, calculates the local noise standard deviation, and only retains the effective peaks, greatly compressing the original signal data volume and realizing real-time transmission of lightning signals. Through waveform feature compression and reconstruction technology, the waveform is reconstructed based on peak parameters ( t p , A p , W p ). The waveform is fitted within the range of | t - t p |≤ W p and rectangular function waveform reconstruction is adopted. The cross-correlation time difference optimization algorithm is adopted, the cross-correlation function is calculated by FFT acceleration, and the time delay between stations is accurately calculated through cross-correlation analysis, and the time delay resolution is improved to 0.1 Δt(10 ns). Combine the weighted non - linear least - squares method to solve the TDOA equations, obtain the lightning location results, and use residual screening and GDOP evaluation to improve the location reliability.

[0004] The technical solution adopted by the present invention is: to provide a real - time lightning location method based on FPGA signal acquisition and pre - processing, including the following steps: S1. N stations acquire lightning signal waveform data; S2. Denoise the lightning signal waveform data; S3. Extract the peak characteristics of the waveform signal, obtain the positive and negative peak amplitudes, absolute time, and half - peak width information of the waveform within different time windows, and transmit them back to the calculation center; S4. The calculation center reconstructs the waveform according to the peak characteristic information; S5. Calculate the cross - correlation of the reconstructed lightning waveform data between every two stations and extract the maximum correlation time delay; S6. According to the signal time - delay information, use the weighted non - linear least - squares method to solve the TDOA equations and output the three - dimensional lightning location results; S7. Evaluate the reliability of the location results through residual analysis and geometric dilution of precision.

[0005] In the steps S1, S2, and S3, the FPGA logic gate circuit is used at the station to pre - process the collected lightning data, extract the peak information, and transmit it to the calculation center in real - time through the 5G network; in the steps S4, S5, S6, and S7, waveform reconstruction, cross - correlation calculation, lightning location, and result evaluation are carried out at the calculation center.

[0006] The specific method for the S1 to obtain multiple lightning signal waveform data is: Let the lightning signal waveform collected by the i - th station be s i ( t ), where t is time, i = 1, 2, …, N , N is the number of stations, the sampling time interval of each station is Δ t , the number of sampling points is M , and baseline correction eliminates the DC offset to make the baseline 0; the signal can be expressed in discrete form: , Time synchronization is achieved among stations using the GNSS timing module, and the synchronization error is less than 25 n s , and the sampling rate f s ≥ 10 MS / s.

[0007] The S2 denoising process uses the Fourier transform denoising method, which specifically includes: for the signal S i ( t ) perform a fast Fourier transform (FFT) to obtain the frequency-domain signal S i ( f ), identify the frequency range where the noise is located, and remove the high-frequency noise and low-frequency trend through a band-pass filter H BP ( f ). Convert the processed frequency-domain signal back to the time domain through an inverse Fourier transform (IFFT), and further perform baseline correction on the denoised signal to eliminate the DC offset, obtaining the denoised lightning signal. The Fourier transform denoising formula is as follows: , After removing the high-frequency noise, perform an inverse Fourier transform (IFFT) on the filtered frequency-domain signal to restore the time-domain signal, obtaining the denoised signal s i de ( t ): , where s i de ( t ) is the i -th denoised signal, IFFT is the inverse fast Fourier transform, which is used to convert the frequency-domain signal back to the time domain, S i ( f ) is the frequency domain of the i -th signal, H BP ( f ) is the frequency-domain response of the filter.

[0008] The S3 peak feature extraction method is as follows: in the signal s i de ( t ), peak feature extraction uses a sliding time window to detect positive and negative peaks, and set the amplitude threshold to 3 σ i ( σ i is the standard deviation). Peaks that satisfy ∣ A p ∣>3 σ i are valid peaks, and record the absolute time t p and amplitude A pand the full width at half maximum W p ; The set of positive and negative peaks in the detected waveform is: , wherein, t p + and t p − are the times of the positive and negative peaks respectively, A p + and A p − are the amplitudes of the positive and negative peaks respectively; Record its amplitude A p , absolute time t p and the full width at half maximum W p ; The full width at half maximum W p The calculation formula of is: , wherein, t left and t right are the time points at half of the peak amplitude respectively.

[0009] The S4 waveform reconstruction described above is based on the extracted peak feature points ( t p , A p , W p ), fitting a single peak with a rectangle, and the data outside the peak is zero. The rectangle function is: , Superimpose all positive and negative peaks to generate a reconstructed waveform: , where: s i re ( t ) is the i th reconstructed signal, ∑ p∈Peaki+ and ∑ p ∈Peaki− are the summations of positive peaks and negative peaks respectively, A p + and A p −are the amplitudes of the positive peak and the negative peak, respectively, t p + and t p − are the time positions of the positive peak and the negative peak, respectively, W p + and W p − are the widths of the positive peak and the negative peak, respectively.

[0010] The S5 cross-correlation mentioned above adopts linear cross-correlation with zero padding. For the reconstructed waveforms of station i and station j S i re ( t ) and S j re ( t ), before cross-correlation, it is necessary to align the waveform baselines and calculate the cross-correlation function: , R ij ( τ ) is the cross-correlation function of the lightning waveforms of the i -th station and the j -th station, which describes the similarity degree of the two waveforms at different time delays τ ; s i re ( t ) and s j re ( t ): They are the reconstructed signals of the i -th station and the j -th station, and the discrete form is: (fill with zeros when out of range), In the cross-correlation function R ij ( τ ), find the delay time that maximizes the function value and extract the maximum correlation time delay τ ij : , where, τ ij represents the i -th station and the jThe maximum correlation time delay of lightning waveforms between stations, extract the maximum correlation coefficient , and the corresponding time delay .

[0011] The TDOA equation set of the S6 positioning algorithm adopts the following formula: , where, ( x , y , z ) is the lightning location, ( x i , y i , z i ) and ( x j , y j , z j ) are the positions of station i and station j respectively, c is the speed of light, ε ij is the observation residual, i , j = 1, 2,..., N and i ≠ j , N ≥5; The TDOA algorithm is based on the least squares initial estimate to calculate the initial value ( x 0 ,y 0 ,z 0 ); construct the objective function and solve it iteratively: , where, , weight ; For the z ≥0 constraint, the projection method is used for processing, and the iteration terminates when the iteration step size ||Δx|| < 1m or the function change rate or the number of iterations ≥ 100.

[0012] The S7 reliability evaluation described above includes: if the root mean square (RMS) of the time delay residuals of multiple station pairs is greater than 0.1 μs , then recalculate after removing the station pairs with amplitude correlation lower than 0.8, τ ijme is the measured value, τ ij ca is the calculated value; Calculate the Geometric Dilution of Precision (GDOP) for error assessment: , where: J T is the observation matrix J transpose of, J T J is a m × m matrix, called the normal equation matrix, ( J T J ) −1 is the inverse matrix of the normal equation matrix, trace(( J T J ) −1 ) is the trace of the inverse matrix, that is, the sum of the diagonal elements of the inverse matrix; When GDOP > 5, the residual screening mechanism is automatically triggered, including: a) Residual analysis, calculate the time difference residual of each station pair , if greater than 0.2 μ s is marked as an outlier; b) Geometric Dilution of Precision (GDOP) error assessment, when GDOP > 3, automatically remove the 20% station pairs with the largest residuals and recalculate; c) Result output, generate the positioning result including three-dimensional coordinates, elevation, confidence level (95% elliptical area) and SNR-weighted.

[0013] The beneficial effects of the present invention are as follows: Through the preprocessing of lightning data at the measuring station, only the peak characteristics (time, amplitude, half-peak width) are transmitted, and the amount of single lightning data is reduced to the MB level, with a data compression rate > 90%. It supports real-time 5G backhaul and real-time lightning positioning, solving the problem that the prior art cannot meet the requirements of scenarios with second-level response. Key features are extracted from the original signal, and through peak waveform reconstruction, the reconstructed waveform can more accurately reflect the characteristics of the lightning signal, reducing the measurement error of time delay caused by signal distortion. There may be amplitude and phase differences in the lightning signals received by different stations. Waveform reconstruction can unify the signal characteristics, facilitating subsequent cross-correlation analysis. The cross-correlation function can effectively calculate the time delay between two signals. Even in the case of signal noise or distortion, the relative time difference of the signals can be determined by finding the maximum value of the cross-correlation function. Lightning signals are usually complex electromagnetic pulses. The cross-correlation method can process signals with different frequency components. The time delay calculated by the cross-correlation method, combined with the TDOA algorithm, can significantly improve the accuracy of lightning positioning. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 is a block diagram of the steps of the lightning positioning method of the present invention; Figure 2 is a principle block diagram of the positioning method of the present invention; Figure 3 is a comparison diagram of the original waveform and the reconstructed waveform of the lightning signal; Figure 4 is Figure 3 a partial enlarged view of; Figure 5 is a comparison diagram of the cross-correlation results of the lightning signal waveforms. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0015] For those skilled in the art to better understand the present invention, the present invention will be further described in conjunction with Figures 1 - 5 The content mentioned in the embodiments is not a limitation to the present invention.

[0016] The real-time lightning positioning method based on FPGA signal acquisition and preprocessing of the present invention includes the following steps: N measuring stations acquire lightning signal waveform data; perform denoising processing on the lightning signal waveform data; extract the peak characteristics of the waveform signal to obtain the positive and negative peak amplitudes, absolute time, and half-peak width information of the waveform within different time windows, and transmit them back to the calculation center; the calculation center performs waveform reconstruction according to the peak characteristic information; calculate the cross-correlation of the reconstructed lightning waveform data between every two measuring stations, and extract the maximum correlation time delay; according to the signal time delay information, use the weighted nonlinear least squares method to solve the TDOA equations, and output the three-dimensional lightning positioning result; evaluate the reliability of the positioning result through residual analysis and geometric dilution of precision. The block diagram of the steps of the lightning positioning method of the present invention is shown in Figure 1 .

[0017] The specific method for obtaining N lightning signal waveform data is as follows: Let the lightning signal waveform collected by the i th measuring station be s i ( t ), where t is time, i = 1, 2, …, N , N is the number of measuring stations, the sampling time interval of each measuring station is Δ t , the number of sampling points is M , and the baseline correction eliminates the DC offset to make the baseline 0. The signal can be expressed in discrete form: , Time synchronization between each measuring station is achieved by using a GNSS timing module, and the synchronization error is less than 25 n s , and the sampling rate f s ≥ 10 MS / s.

[0018] TDOA lightning location requires a master station plus three slave stations to establish an equation set. To remove the ambiguous solution, at least one redundant slave station is required. Therefore, the number of measuring stations N should be ≥ 5. The TDOA location of the present invention uses the CHAN algorithm, which is a non-iterative algorithm for solving the location equation and is an approximate implementation method of maximum likelihood estimation in a line-of-sight environment. It has the characteristics of small computational workload and accurate results. Its disadvantage is that when the signal propagation path is non-line-of-sight, the performance of the CHAN algorithm deteriorates rapidly. Therefore, the setting of the measuring stations should avoid the occlusion of high-rise buildings and mountains.

[0019] The measuring station uses an FPGA lightning signal acquisition system, and the sampling frequency f s should be ≥ 10 MS / s, meeting the Nyquist sampling requirements for the high-frequency components of lightning pulses (the main frequency ≤ 5 MHz). The number of sampling points M = 10,000, covering the time window T w = 1 ms. The original signal is baseline-corrected to eliminate the DC offset. The high-speed A / D conversion module of the acquisition system converts the analog signal output by the lightning detection sensor into a digital signal. The GNSS timing unit outputs a high-precision clock frequency as the sampling reference clock of the high-speed A / D conversion module to ensure the accuracy of the signal conversion time. The GNSS timing unit uses a BDS / GPS receiver. The FPGA logic gate circuit realizes analog-to-digital conversion control, lightning detection data acquisition, timestamp acquisition, denoising processing, and peak feature extraction.

[0020] The purpose of denoising the lightning signal waveform data is to improve the quality and usability of the signal for more accurate subsequent analysis and processing. Lightning signals are often interfered by various noises during the acquisition and transmission processes. These noises may come from the natural environment, the device itself, or other external factors. Denoising processing can effectively reduce or eliminate these noise components and highlight the effective information of the signal. The lightning signal waveform data denoising processing of the present invention adopts the Fourier transform denoising method, which specifically includes: for the signal S i ( t ) perform a fast Fourier transform (FFT) to obtain the frequency-domain signal S i ( f ), identify the frequency range where the noise is located, and remove the high-frequency noise and low-frequency trend through a band-pass filter H BP ( f ). Convert the processed frequency-domain signal back to the time domain through an inverse Fourier transform (IFFT), further perform baseline correction on the denoised signal to eliminate the DC offset, and obtain the denoised lightning signal. The Fourier transform denoising formula is as follows: , After removing the high-frequency noise, perform an inverse Fourier transform (IFFT) on the filtered frequency-domain signal to restore the time-domain signal, and perform baseline correction using moving average filtering (window length 1 ms) to eliminate the offset and obtain the denoised signal s i de ( t ): , wherein, s i de ( t ) is the i th denoised signal, IFFT is the inverse fast Fourier transform, which is used to convert the frequency-domain signal back to the time domain, S i ( f ) is the frequency domain of the i th signal, H BP ( f ) is the frequency-domain response of the filter.

[0021] The purpose of extracting the peak characteristics of the lightning signal waveform is to identify and extract representative and key information from the complex lightning signal for subsequent signal analysis, processing, and application. For the extraction of the peak characteristics of the waveform signal, obtain the positive and negative peak amplitudes, absolute time, and half-peak width information of the waveform within different time windows and transmit them back to the calculation center. The peak characteristic extraction method is: for the signals i de ( t ) Among them, peak feature extraction uses a sliding time window to detect positive and negative peaks. The amplitude threshold is set to 3 σ i ( σ i is the standard deviation), satisfying ∣ A p ∣ > 3 σ i of the peak is a valid peak, and record the absolute time of the positive and negative peaks t p , amplitude A p and half-peak width W p . The set of positive and negative peaks detected in the waveform is: , where, t p + and t p − are the times of the positive and negative peaks respectively, A p + and A p − are the amplitudes of the positive and negative peaks respectively; record their amplitudes A p , absolute time t p and half-peak width W p . When the amplitude of the positive peak A p + is greater than 3 σ i , it is determined as a valid positive peak. When the absolute value of the amplitude of the negative peak ∣ A p − ∣ is greater than 3 σ i , it is determined as a valid negative peak. σ i is the standard deviation of the signal, which is used to measure the degree of signal fluctuation. Using 3 σi as the threshold is because in the normal distribution, about 99.7% of the data will fall within ±3 σ of the mean. Therefore, peaks with amplitudes exceeding 3 σi are likely to be outliers or noise. For each peak point determined to be valid tp , its full width at half maximum (FWHM) needs to be calculated W p . The specific steps are as follows: Determine the peak center and find the central time point of the peak t p , that is, the time position where the peak appears. Halve the amplitude and calculate half of the peak amplitude, i.e., A p / 2, where A p is the amplitude of the peak. Search for time points, starting from t p , search left and right respectively to find the time points when the signal amplitude first drops to A p / 2 t left and t right . The full width at half maximum W p is the difference between these two time points. The formula for calculating the full width at half maximum W p is: , where, t left and t right are the time points at half of the peak amplitude respectively. Adopt the sliding window dynamic threshold algorithm to extract only the effective peak parameters that satisfy | A p |>3σ. The amount of single-peak data is greatly compressed, meeting the real-time transmission requirements of the 5G network.

[0022] The above steps are completed on the measuring station using the FPGA logic gate circuit embedded operating system. After the lightning signals collected by the measuring station are preprocessed, the 16-bit data output from the FPGA has a speed of 80 MHZ and a data volume of 160 MB / s. However, due to the network speed limitation when transmitting back through the 5G network, the lightning signal acquisition speed is much greater than the data transmission speed. Therefore, using DDR as a cache can effectively solve this problem and make up for the speed difference between lightning signal acquisition and data transmission. The data is temporarily stored in the DDR data storage unit using DMA, and the ARM processing unit embedded operating system extracts the peak features ([[]] t p , A p , W p ) from the cache of the DDR data storage unit after receiving the notification of the collected data, and transmits them back to the computing center through the data transmission unit. The principle block diagram of the positioning method of the present invention is shown in Figure 2 .

[0023] The computing center receives lightning data information and performs waveform reconstruction based on the peak feature information. The reconstruction of the lightning signal wave peak refers to restoring the peak characteristics of the signal through the processing and analysis of the lightning signal, so as to more accurately identify and analyze the signal. The waveform reconstruction is based on the extracted peak feature points ( t p , A p , W p ). The rising edge of the lightning pulse approximates the characteristics of a quadratic function. The present invention uses a rectangular function for reconstruction, which can more strictly constrain the steep front edge of the lightning pulse and avoid the smoothing effect of the parabola model on the rising edge. The waveform reconstruction is based on the extracted peak feature points ( t p , A p , W p ). A single peak is fitted with a rectangle, and the data outside the peak is zero. The rectangular function is: , The rectangular function appears as a square pulse with a constant amplitude in the time domain. The horizontal axis (time axis) defines the time window of the pulse, and the center point is t p (positive peak) or t p − (negative peak). The vertical axis (amplitude axis) defines the amplitude intensity of the pulse A p + (positive peak) or A p − (negative peak). The pulse width is controlled by W p + (positive peak) or W p − (negative peak) to control the symmetric extension range of the window. At t = t p , the function reaches the maximum value A p , and the function value is zero at other times.

[0024] All positive and negative peaks are superimposed to generate a reconstructed waveform. The positive and negative peaks are superimposed independently to avoid polarity confusion: , Where: s i re ( t ) is the i -th reconstructed signal, ∑ p∈Peaki+ and ∑ p ∈Peaki−They are the summations of the positive peaks and negative peaks respectively, A p + and A p − They are the amplitudes of the positive peaks and negative peaks respectively, t p + and t p − They are the time positions of the positive peaks and negative peaks respectively, W p + and W p − They are the widths of the positive peaks and negative peaks respectively. When processing the signal, in the region where ∣ t - t p ∣> W p the signal is set to zero, aiming to reduce the sidelobe interference in the frequency domain. The sidelobe interference is caused by the sudden change of the time-domain signal and appears as other frequency components except the main lobe in the frequency domain. By setting the signal to zero in the region where ∣ t - t p ∣> W p the time range of the signal can be effectively restricted and the generation of sidelobes can be reduced. Through rectangular function reconstruction, on the premise of sacrificing the frequency-domain smoothness, the ability to retain the key transient features of lightning signals is enhanced, which is especially suitable for lightning location. Calculate the correlation coefficient ρ ≥0.8, otherwise re-extract the peaks.

[0025] Calculate the cross-correlation of the reconstructed lightning waveform data between every two stations and extract the maximum correlation time delay. The cross-correlation adopts linear cross-correlation with zero padding. For the reconstructed waveforms i of station j and station S i re ( t ) and S j re ( t ), the waveform baselines need to be aligned before cross-correlation. For s i re ( t ) and S j re ( t ), zero padding is performed to a length of 2M, and the cross-correlation function is calculated using the following formula: , R ij ( τ ) is the cross - correlation function of the lightning waveforms of the i th and j th stations, which describes the similarity degree of the two waveforms at different time delays τ ; s i re ( t ) and s j re ( t ) are respectively the reconstructed signals of the i th and j th stations, and the discrete form is: , (padding with zeros when out of range); In the cross - correlation function R ij ( τ ), find the delay time that maximizes the function value and extract the maximum correlation time delay τ ij : , τ ij represents the maximum correlation time delay of the lightning waveforms between the i th and j th stations, , extract the maximum correlation coefficient and the corresponding time delay .

[0026] According to the signal time - delay information, use the weighted non - linear least - squares method to solve the TDOA equations and output the three - dimensional lightning location result. The TDOA equations of the location algorithm are as follows: , where, ( x , y , z ) is the lightning location, ( x i , y i , z i ) and ( x j , y j , z j ) are respectively the i th station and thej The position of c is the speed of light, ε ij is the observation residual, i , j = 1, 2, ..., N and i ≠ j , N ≥ 5; The TDOA algorithm is based on the least - squares initial estimate to calculate the initial value ( x 0 ,y 0 ,z 0 ); construct the objective function and solve it iteratively: , where, , the weight ; For the z ≥ 0 constraint, the projection method is used. When the iteration step size ||Δx|| < 1m or the function change rate or the number of iterations ≥ 100, it terminates. After the optimization algorithm converges, the three - dimensional position of the signal source ( x , y, z ) is obtained.

[0027] The reliability of the positioning result is evaluated through residual analysis and the geometric dilution of precision. The residual is the difference between the observed value and the model predicted value. In TDOA positioning, the residual represents the difference between the measured time difference and the time difference calculated based on the solved position. If the root - mean - square (RMS) of the time - delay residuals of multiple station pairs is greater than 0.1 μs , then the station pairs with amplitude correlation lower than 0.8 are removed and recalculated. τ ij me is the measured value, τ ij ca is the calculated value. The geometric dilution of precision (GDOP) is an index to evaluate the influence of the geometric layout of stations on the positioning accuracy. The smaller the GDOP value, the more ideal the geometric layout of the receiving points and the higher the positioning accuracy. Calculate the geometric dilution of precision (GDOP) for error evaluation: , where: J T is the transpose of the observation matrix J , J T J is am × m matrix, called the normal equation matrix, ( J T J ) −1 is the inverse matrix of the normal equation matrix, trace(( J T J ) −1 ) is the trace of the inverse matrix, that is, the sum of the diagonal elements of the inverse matrix. When GDOP > 5, the residual screening mechanism is automatically triggered, including: a) Residual analysis, calculating the time delay residuals of each pair of stations , if it is greater than 0.2 μ s, it is marked as an outlier; b) Geometric dilution of precision (GDOP) error assessment. When GDOP > 3, the 20% of the station pairs with the largest residuals are automatically removed and recalculated; c) Result output, generating the positioning results including three-dimensional coordinates, elevation, confidence level (95% elliptical area), and SNR-weighted.

[0028] Figure 3 is the comparison chart of the original waveform and the reconstructed waveform. The upper figure is the waveform of i station, and the lower figure is the waveform of j station. Figure 3 is the superposition chart of the original waveform and the reconstructed waveform within 1ms. The original waveform shows a curved undulation, and the reconstructed waveform shows a rectangular shape going straight up and down. Figure 4 is Figure 3 the partial enlarged view intercepted from 0 to 0.1ms. It can be seen from Figure 3 , Figure 4 that the original waveform has a large amount of data. Through reconstruction with a rectangular function, only the peak features are extracted. On the premise of sacrificing the frequency domain smoothness, the ability to retain the key transient features of lightning signals is enhanced, and the amount of data transmitted back is greatly reduced. Figure 5 is i the comparison chart of the cross-correlation results of lightning signal waveforms between j station and Figure 5 station. The upper figure is the cross-correlation result of the original waveform, and the lower figure is the cross-correlation result of the reconstructed waveform. It can be seen from Figure 5 that the original waveform directly reflects the instantaneous change of lightning current, including high-frequency details and steep rising edges. By cross-correlating the reconstructed waveforms and comparing the phase similarity of the original waveforms, the instantaneous matching degree of lightning signals can be accurately captured. The present invention is used for lightning location and does not require cross-correlation of the original waveforms,

[0029] The description and drawings of this application are only a specific embodiment and not restrictive. Those skilled in the art, under the inspiration of this application and without departing from the purpose of this application and the scope protected by the claims, can also make many forms, all of which are within the protection scope of this application.

Claims

1. A real-time lightning location method based on FPGA signal acquisition preprocessing, characterized in that: The following steps are involved: S1, N Each measuring station obtains lightning signal waveform data; S2, performing denoising processing on lightning signal waveform data; S3, extract the peak features of the waveform signal, obtain the positive and negative peak amplitudes, absolute time, and half-peak width information of the waveform in different time windows, and transmit it back to the computing center; S4, the computing center reconstructs the waveform according to the peak characteristic information; S5, calculating the cross-correlation of the lightning waveform data reconstructed between every two measuring stations, and extracting the maximum correlation time delay; S6. According to the signal time delay information, a weighted nonlinear least square method is used to solve the TDOA equations and output a three-dimensional lightning location result; S7. Evaluate the reliability of positioning results through residual analysis and geometric precision factor.

2. The real-time lightning location method based on FPGA signal acquisition preprocessing according to claim 1 is characterized in that: In steps S1, S2, and S3, FPGA logic gate circuits are used at the measuring station to pre-process the collected lightning data, extract peak information, and transmit it back to the computing center in real time through the 5G network; in steps S4, S5, S6, and S7, waveform reconstruction, cross-correlation calculation, lightning location, and result evaluation are performed at the computing center.

3. The real-time lightning location method based on FPGA signal acquisition preprocessing according to claim 1 is characterized in that: The specific method of S1 obtaining multiple lightning signal waveform data is as follows: i The lightning signal waveform collected by each measuring station is s i ( t ),in t For time, i = 1, 2, …, N , N is the number of stations, and the sampling time interval of each station is Δ t , the number of sampling points is M , baseline correction eliminates DC offset to a baseline of 0; the signal can be expressed in discrete form: ， GNSS timing modules are used between each measuring station to achieve time synchronization, with a synchronization error of less than 25n s , sampling rate f s ≥10MS / s.

4. The real-time lightning location method based on FPGA signal acquisition preprocessing according to claim 1 is characterized in that: The S2 denoising process adopts the Fourier transform denoising method, which specifically includes: S i ( t ) Perform fast Fourier transform (FFT) to obtain the frequency domain signal S i ( f ), identify the frequency range of the noise, and pass it through a bandpass filter H BP ( f ) removes high-frequency noise and low-frequency trends, converts the processed frequency domain signal back to the time domain through inverse Fourier transform (IFFT), further performs baseline correction on the denoised signal, eliminates DC offset, and obtains the denoised lightning signal. The Fourier transform denoising formula is as follows: ， After removing the high-frequency noise, the filtered frequency domain signal is subjected to an inverse Fourier transform (IFFT) to restore the time domain signal and obtain the denoised signal. s i de ( t ): ， in, s i de ( t ) is the i denoised signal, IFFT is the inverse fast Fourier transform, which is used to convert the frequency domain signal back to the time domain. S i ( f ) is the i The frequency domain of a signal, H BP ( f ) is the frequency domain response of the filter.

5. The real-time lightning location method based on FPGA signal acquisition preprocessing according to claim 1 is characterized in that: The S3 peak feature extraction method is as follows: s i de ( t ), the peak feature extraction uses a sliding time window to detect positive and negative peaks, and the amplitude threshold is set to 3 σ i ( σ i is the standard deviation), satisfying ∣ A p ∣>3 σ i The peak value is the effective peak value, and the absolute time of the positive and negative peak values ​​is recorded. t p , Amplitude A p and half peak width W p ; The positive and negative peak sets in the detection waveform are: ， in, t p + and t p − are the times of the positive and negative peaks, respectively. A p + and A p − are the amplitudes of the positive and negative peaks, respectively; Record its amplitude A p , absolute time t p and half-peak width W p ; Half peak width W p The calculation formula is: ， in, t left and t right are the time points at half the peak amplitude.

6. The real-time lightning location method based on FPGA signal acquisition preprocessing according to claim 1 is characterized in that: The S4 waveform reconstruction is based on the extracted peak feature points ( t p , A p , W p ), a single peak is fitted with a rectangle, and the data outside the peak is zero. The rectangular function is: ， Add all positive and negative peaks to generate the reconstructed waveform: ， in: s i re ( t ) is the i reconstructed signal, ∑ p∈Peaki+ and∑ p ∈Peaki− are the sum of the positive peak and the negative peak respectively, A p + and A p − are the amplitudes of the positive and negative peaks, respectively, t p + and t p − are the time positions of the positive and negative peaks, respectively. W p + and W p − are the widths of the positive and negative peaks, respectively.

7. The real-time lightning location method based on FPGA signal acquisition preprocessing according to claim 1 is characterized in that: The S5 cross-correlation adopts linear cross-correlation with zero filling. i and station j The reconstructed waveform S i re ( t ) and S j re ( t ), the waveform baseline needs to be aligned before cross-correlation, and the cross-correlation function is calculated: , R ij ( τ ) is the i Stations and j The cross-correlation function of the lightning waveforms at the observation stations describes the relationship between the two waveforms at different time delays. τ The degree of similarity under s i re ( t ) and s j re ( t ): respectively i Stations and j The signal reconstructed by each station is in discrete form: (fill with zero when out of range), In the cross-correlation function R ij ( τ ) to find the delay time that maximizes the function value and extract the maximum relevant time delay τ ij : , τ ij Indicates i Stations and j The maximum correlation time delay of lightning waveforms between the measuring stations is , extract the maximum correlation coefficient And the corresponding time delay 8. The real-time lightning location method based on FPGA signal acquisition preprocessing according to claim 1 is characterized in that: The S6 positioning algorithm TDOA equation group adopts the following formula: , in, ( x , y , z ) is the lightning position, ( x i , y i , z i ) and( x j , y j , z j ) are the measuring stations i and station j location, c is the speed of light, ε ij is the observed residual, i , j =1,2,..., N and i ≠ j , N ≥5; The TDOA algorithm is based on the initial estimate of least squares and calculates the initial value ( x 0 ,y 0 ,z 0); construct the objective function and solve it iteratively: , in, , weight ; right z ≥0 constraint is processed by projection method. When the iteration step ||Δx||<1m or the function change rate Or terminate when the number of iterations ≥ 100.

9. The real-time lightning location method based on FPGA signal acquisition preprocessing according to claim 1, characterized in that: The S7 reliability evaluation includes: if the time delay residuals of multiple station pairs The root mean square (RMS) is greater than 0.1 μs , then remove the station pairs with amplitude correlation lower than 0.8 and recalculate. τ ij me is the measured value, τ ij ca is the calculated value; Calculate the geometric dilution of precision (GDOP) for error assessment: ， in: J T is the observation matrix J The transpose of J T J is a m × m The matrix of is called the normal equation matrix, ( J T J ) −1 is the inverse matrix of the normal equation matrix, trace(( J T J ) −1 ) is the trace of the inverse matrix, i.e., the sum of the diagonal elements of the inverse matrix; When GDOP>5, the residual screening mechanism is automatically triggered, including: a) Residual analysis, calculate the time delay residual of each station pair , if greater than 0.2 μ s is marked as an outlier; b) Geometric Dilution of Precision (GDOP) error assessment: when GDOP>3, the 20% station pairs with the largest residual errors are automatically eliminated and recalculated; c) Result output: generate positioning results including three-dimensional coordinates, elevation, confidence level (95% ellipse area) and signal-to-noise ratio weight.

Images