An air-ground electromagnetic exploration time-frequency data fusion inversion method

Abstract

This invention relates to the field of electromagnetic detection technology and provides a method for fusing and inverting time-frequency data from ground-to-air electromagnetic detection. The method includes: Step 1, time-domain ground-to-air electromagnetic detection data conversion: baseline correction, meanization, signal superposition and denoising, Fourier transform extraction of amplitude spectrum and calculation of frequency point sequence; Step 2, time-frequency detection data fusion: baseline correction, signal superposition and denoising, normalization, combination and sorting of frequency point sequence and spectrum data; Step 3, time-frequency fusion data inversion: setting an initial model of underground parameters, establishing an inversion objective function, obtaining the Jacobian matrix, using the least squares method to calculate the model parameter changes, repeating until all measurement points have been inverted to obtain the resistivity distribution. The ground-to-air electromagnetic detection time-frequency data fusion and inversion algorithm proposed in this invention can strengthen the correlation between shallow and deep information, reduce the ambiguity in the shallow-time-deep-frequency transition region, and improve the certainty of inversion results and the efficiency of data interpretation.

Description

Technical Field

[0001] This invention belongs to the field of electromagnetic detection technology, and in particular relates to a method for fusing and inverting time-frequency data from ground-to-air electromagnetic detection. Background Technology

[0002] Ground-to-air electromagnetic (GTE) is an artificial-source geophysical exploration technology that uses remote sensing to acquire information about the electrical structure of subsurface media. This technology establishes an active electromagnetic field by deploying high-power electrical excitation sources on the ground and uses a signal acquisition system mounted on a flight platform (such as a helicopter or drone) to detect the secondary field carrying electrical information about the subsurface structure. In the time-domain (TDV) GTE mode, a fixed-frequency bipolar pulse is used as the excitation pulse signal for the ground excitation source, and the induced electromotive force captured by the airborne signal acquisition system is a time-domain attenuated signal carrying information about shallow subsurface layers. In the frequency-domain (FV) GTE mode, a multi-frequency pseudo-random pulse is used as the excitation pulse for the ground excitation source, and the induced electromotive force captured by the airborne signal acquisition system is a frequency-domain attenuated signal carrying information about deep subsurface layers.

[0003] The time-domain signal inversion results show the resistivity variation with shallow subsurface depth, while the frequency-domain signal inversion results show the resistivity variation with deep subsurface depth. There is an unclear transition region between the two at intermediate depths. Currently, there are two main techniques for interpreting time-domain and frequency-domain geomagnetic sounding signals. One is fully separate data inversion, which involves repeatedly fitting forward models of geoelectric parameters to the time / frequency domain data, resulting in two sets of electrical parameters that respectively conform to the trends of the time / frequency domain data. This technique completely ignores the correlation between the two at different depths, and the independent data fitting results in multiple solutions, leading to errors and anomalies in the inversion results. The other is semi-separate data inversion, which simultaneously updates the geoelectric parameter models of the time / frequency domain data during each signal fitting process, ultimately obtaining one set of results that respectively conform to the trends of the time / frequency domain data. While this type of technique draws on the advantages of joint inversion, it requires applying weights to the two types of data during the signal fitting process. Shallow and deep information are weighted towards fitting the time and frequency domain signals, respectively, and the parameter weights in the depth transition region cannot be effectively determined. This makes the inversion results dependent on the selection of weight coefficients, thus increasing the uncertainty of the inversion results. Furthermore, using separate data inversion methods is extremely time-consuming for large-scale coal mine underground structure detection, resulting in low inversion efficiency. Therefore, a ground-air electromagnetic detection time-frequency data fusion inversion method is proposed. Summary of the Invention

[0004] The purpose of this invention is to provide a method for fusing and inverting time-frequency data from ground-to-air electromagnetic detection, aiming to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for fusing and inverting time-frequency data from ground-to-air electromagnetic detection includes the following steps:

[0007] Step 1: Time-domain ground-to-space electromagnetic detection data conversion;

[0008] Step 2: Fusion of time and frequency detection data;

[0009] Step 3: Time-frequency fusion data inversion;

[0010] Step 1 includes the following steps:

[0011] Step 11: Perform baseline correction on the time-domain ground-space electromagnetic detection dataset collected for each survey line;

[0012] Step 12: According to the signal acquisition sequence of the flight path, average the transmission and reception parameters of multiple consecutive measurements near N measurement points on a single measurement line to obtain the transmission and reception parameter sets of all individual measurement points from time-domain detection and save them independently.

[0013] Step 13: Superimpose and denoise multiple consecutive sets of signals near N measurement points on a single measurement line, and save them independently as the time-domain attenuation data vector of a single measurement point;

[0014] Step 14: Extract the amplitude spectrum of the time-domain decay data vector in the frequency domain using Fourier transform;

[0015] Step 15: Calculate the frequency sequence of the corresponding spectrum using the sampling frequency of the time-domain data;

[0016] Step 16: Repeat steps 13 to 15 to perform time-frequency conversion on the data of all measuring points on the measuring line to obtain the corresponding spectrum dataset and frequency sequence;

[0017] Step 2 includes the following steps:

[0018] Step 21: Perform baseline correction on the frequency domain ground-to-air electromagnetic detection dataset collected for each survey line. According to the signal acquisition sequence of the flight route, average the transmission and reception parameters of multiple consecutive measurements near N measurement points on a single survey line to obtain the transmission and reception parameter sets of all individual measurement points from frequency domain detection and save them independently.

[0019] Step 22: Superimpose and denoise multiple consecutive sets of signals near N measurement points on a single measurement line to obtain spectrum data and corresponding frequency point sequences;

[0020] Step 23: Normalize the spectral data from the time domain and frequency domain detection modes respectively to obtain the normalized spectral data for each measurement point;

[0021] Step 24: For a single measurement point, combine the two sets of frequency point sequences from the time domain and frequency domain detection modes into a single frequency point sequence, and sort them from largest to smallest to obtain the time-frequency fusion frequency point sequence of the single measurement point;

[0022] Step 25: Combine and sort the normalized spectrum data according to the corresponding frequency point positions to obtain the time-frequency fusion spectrum data of a single measurement point;

[0023] Step 26: Repeat steps 22 to 25 to sort and combine the frequency sequence and spectrum data of all measurement points on the measurement line to obtain the corresponding comprehensive spectrum dataset and frequency sequence;

[0024] Step 3 includes the following steps:

[0025] Step 31: Set the initial model of the underground parameters for the i-th measuring point, including the resistivity and thickness of each underground layer, and take the logarithm of the model parameters;

[0026] Step 32: Establish the objective function for the inversion of fused data;

[0027] Step 33: Calculate the Jacobian matrix of each parameter using frequency domain ground-to-space electromagnetic forward modeling data;

[0028] Step 34: Use the least squares method to calculate the changes in model parameters until the current model parameters meet the fitting error of the fused data or reach the maximum number of iterations;

[0029] Step 35: Repeat steps 31 to 34 until all the fused data from N measurement points have been inverted, and obtain the time-frequency fused resistivity distribution of the entire measurement line.

[0030] Furthermore, the specific processes of steps 13 to 16 are as follows:

[0031] Step 13: Superimpose and denoise multiple consecutive sets of signals near N measurement points along a single measurement line to obtain the time-domain attenuation data vector for a single measurement point. And save it separately;

[0032] Step 14: Extract the time-domain decay data vector using Fourier transform. Amplitude spectrum in the frequency domain

[0033]

[0034] Where i is the measurement point number, j is the time-domain data number of the time-domain ground-to-air electromagnetic data of a single measurement point, and k is the frequency-domain data number of the time-domain ground-to-air electromagnetic data of a single measurement point.

[0035] Step 15: Calculate the frequency sequence f of the corresponding spectrum using the sampling frequency of the time-domain data. TEM,i ;

[0036]

[0037] Where Fl=[0,1,2,…,L], L is Data length, Fs TEM It is the sampling frequency of the time-domain data;

[0038] Step 16: Repeat steps 13-15 to process the data of all measuring points on the survey line. Perform time-frequency conversion to obtain the corresponding spectrum dataset. Sum of frequency points f TEM :

[0039]

[0040] f TEM ={f TEM,1 ,f TEM,2 ,f TEM,3 ,…,f TEM,N-1 ,f TEM,N}

[0041] Furthermore, in step 12, the transmit / receive parameter set includes the receive amplification factor mR from the time-domain probe. TEM,i Receiver point coordinates CoR TEM,i emission current IT TEM,i and the coordinates of the emission source CoT TEM,i .

[0042] Furthermore, the specific processes of steps 22 to 26 are as follows:

[0043] Step 22: Superimpose and denoise multiple consecutive sets of signals near N measurement points along a single measurement line to obtain spectrum data. and the corresponding frequency sequence f FEM,i ;

[0044] Step 23: Normalize the spectral data from both the time and frequency domain detection modes to obtain the normalized spectral data for each measurement point. and

[0045]

[0046] Among them, mR TEM,i It is the receiving amplification factor from time-domain probe, IT TEM,i It is the transmit current from time-domain probe, mR FEM,i It is the receiving amplification factor from frequency domain detection, IT FEM,i It is the transmit current from frequency domain detection;

[0047] Step 24: For a single measurement point, combine the two sets of frequency sequence f from the time domain and frequency domain detection modes. TEM,i and f FEM,i The frequencies are combined into a set of frequency points and sorted from largest to smallest to obtain the time-frequency fusion frequency point sequence f for a single measurement point. EM,i :

[0048] f EM,i =sort[f TEM,i ,f FEM,i ];

[0049] Step 25: Normalize the spectrum data and By combining and sorting the corresponding frequency points, the time-frequency fusion spectrum data of a single measurement point can be obtained.

[0050] Step 26: Repeat steps 22-25 to sort and combine the frequency sequence and spectrum data of all measurement points on the measurement line to obtain the corresponding comprehensive spectrum dataset. Sum of frequency points f EM :

[0051]

[0052] f EM ={f EM,1 ,f EM,2 ,f EM,3 ,…,f EM,N-1 ,f EM,N}

[0053] Furthermore, in step 21, the transmit / receive parameter set includes the receive amplification factor mR from frequency domain detection. FEM,i Receiver point coordinates CoR FEM,i emission current IT FEM,i and the coordinates of the emission source CoT FEM,i .

[0054] Furthermore, the specific steps of step 3 are as follows:

[0055] Step 31: Set the initial model m of the underground parameters for the i-th measuring point. 0,i Including the resistivity ρ of each underground layer i and layer thickness h i And take the logarithm of the model parameters:

[0056]

[0057] Step 32: Establish the objective function Φ for the inversion of fused data. EM,i :

[0058]

[0059] in, It is normalized detection fusion data. It is based on the fusion frequency point f EM,i and the initial model m i The spectral data obtained from forward modeling, where λ is the regularization parameter and C is the smoothness matrix;

[0060] Step 33: Calculate the Jacobian matrix G of each parameter using frequency-domain ground-to-space electromagnetic forward modeling data. EM,i :

[0061]

[0062] Where, Δρ g It is a resistivity element, Δh g It is a layer thickness micro-element;

[0063] Step 34: Use the least squares method to calculate the change in model parameters Δm. i :

[0064]

[0065] Among them, e TEM,i and e FEM,i These are the normalized detection errors of electromagnetic detection data in the time domain and frequency domain, respectively.

[0066] The least squares method is repeated until the current model parameter m is reached. c,i Satisfy the fitting error of the fused data or reach the maximum number of iterations:

[0067] m c,i =m i +Δm i ;

[0068] Step 35: Repeat steps 31-34 until all the fused data from N measurement points have been inverted, obtaining the time-frequency fused resistivity distribution of the entire measurement line.

[0069] m c ={m c,1 ,m c,2 ,m c,3 ,…,m c,N-1 ,m c,N}

[0070] Compared with the prior art, the beneficial effects of the present invention are:

[0071] The ground-to-air electromagnetic detection time-frequency data fusion inversion algorithm proposed in this invention can enhance the correlation between shallow and deep information, reduce the ambiguity in the shallow time-deep frequency transition region, and improve the certainty of inversion results and the efficiency of data interpretation. Attached Figure Description

[0072] Figure 1 This is a flowchart of the present invention.

[0073] Figure 2 This is a flowchart of the time-domain ground-to-space electromagnetic detection data conversion process in this invention.

[0074] Figure 3 This is a flowchart of time-frequency detection data fusion in this invention.

[0075] Figure 4 This is a flowchart of time-frequency fusion data inversion in this invention. Detailed Implementation

[0076] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0077] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0078] like Figure 1 As shown, an embodiment of the present invention provides a method for fusing and inverting time-frequency data from ground-to-air electromagnetic detection, comprising the following steps:

[0079] Step 1: Time-domain ground-to-space electromagnetic detection data conversion;

[0080] Step 2: Fusion of time and frequency detection data;

[0081] Step 3: Time-frequency fusion data inversion.

[0082] like Figure 2 As shown, in a preferred embodiment of the present invention, step 1 includes the following steps:

[0083] Step 11: Perform baseline correction on the time-domain ground-space electromagnetic detection dataset collected for each survey line;

[0084] Step 12: Following the signal acquisition sequence of the flight path, average the transmitted and received parameters from multiple consecutive measurements near N measurement points along a single measurement line to obtain the set of transmitted and received parameters from each measurement point in the time domain (including the received amplification factor mR from the time domain probe). TEM,i Receiver point coordinates CoR TEM,i emission current IT TEM,i and the coordinates of the emission source CoT TEM,i (and other information) and save them independently;

[0085] Step 13: Superimpose and denoise multiple consecutive sets of signals near N measurement points along a single measurement line to obtain the time-domain attenuation data vector for a single measurement point. And save it separately;

[0086] Step 14: Extract the time-domain decay data vector using Fourier transform. Amplitude spectrum in the frequency domain

[0087]

[0088] Where i is the measurement point number, j is the time-domain data number of the time-domain ground-to-air electromagnetic data of a single measurement point, and k is the frequency-domain data number of the time-domain ground-to-air electromagnetic data of a single measurement point.

[0089] Step 15: Calculate the frequency sequence f of the corresponding spectrum using the sampling frequency of the time-domain data. TEM,i ;

[0090]

[0091] Where Fl=[0,1,2,…,L], L is Data length, Fs TEM It is the sampling frequency of the time-domain data;

[0092] Step 16: Repeat steps 13-15 to process the data of all measuring points on the survey line. Perform time-frequency conversion to obtain the corresponding spectrum dataset. Sum of frequency points f TEM :

[0093]

[0094] f TEM ={f TEM,1 ,f TEM,2 ,f TEM,3 ,…,f TEM,N-1 ,f TEM,N}

[0095] like Figure 3 As shown, in a preferred embodiment of the present invention, step 2 includes the following steps:

[0096] Step 21: Perform baseline correction on the frequency domain air-to-ground electromagnetic detection dataset collected for each survey line. Following the signal acquisition sequence of the flight path, average the transmission and reception parameters of multiple consecutive measurements near N measurement points on a single survey line to obtain the transmission and reception parameter set for each measurement point from the frequency domain detection (including the receiving amplification factor mR from the frequency domain detection). FEM,i Receiver point coordinates CoR FEM,i emission current IT FEM,i and the coordinates of the emission source CoT FEM,i (and other information) and save them independently;

[0097] Step 22: Superimpose and denoise multiple consecutive sets of signals near N measurement points along a single measurement line to obtain spectrum data. and the corresponding frequency sequence f FEM,i ;

[0098] Step 23: Normalize the spectral data from both the time and frequency domain detection modes to obtain the normalized spectral data for each measurement point. and

[0099]

[0100] Among them, mR TEM,i It is the receiving amplification factor from time-domain probe, IT TEM,i It is the transmit current from time-domain probe, mR FEM,i It is the receiving amplification factor from frequency domain detection, IT FEM,i It is the transmit current from frequency domain detection;

[0101] Step 24: For a single measurement point, combine the two sets of frequency sequence f from the time domain and frequency domain detection modes. TEM,i and f FEM,i The frequencies are combined into a set of frequency points and sorted from largest to smallest to obtain the time-frequency fusion frequency point sequence f for a single measurement point. EM,i :

[0102] f EM,i =sort[f TEM,i ,f FEM,i ];

[0103] Step 25: Normalize the spectrum data and By combining and sorting the corresponding frequency points, the time-frequency fusion spectrum data of a single measurement point can be obtained.

[0104] Step 26: Repeat steps 22-25 to sort and combine the frequency sequence and spectrum data of all measurement points on the measurement line to obtain the corresponding comprehensive spectrum dataset. Sum of frequency points f EM :

[0105]

[0106] f EM ={f EM,1 ,f EM,2 ,f EM,3 ,…,f EM,N-1 ,f EM,N}

[0107] like Figure 3 As shown, in a preferred embodiment of the present invention, step 3 includes the following steps:

[0108] Step 31: Set the initial model m of the underground parameters for the i-th measuring point. 0,i Including the resistivity ρ of each underground layer i and layer thickness h i And take the logarithm of the model parameters:

[0109]

[0110] Step 32: Establish the objective function Φ for the inversion of fused data. EM,i :

[0111]

[0112] in, It is normalized detection fusion data. It is based on the fusion frequency point f EM,i and the initial model m i The spectral data obtained from forward modeling, where λ is the regularization parameter and C is the smoothness matrix;

[0113] Step 33: Calculate the Jacobian matrix G of each parameter using frequency-domain ground-to-space electromagnetic forward modeling data. EM,i :

[0114]

[0115] Where, Δρ g It is a resistivity element, Δh g It is a layer thickness micro-element;

[0116] Step 34: Use the least squares method to calculate the change in model parameters Δm. i :

[0117]

[0118] Among them, e TEM,i and e FEM,i These are the normalized detection errors of electromagnetic detection data in the time domain and frequency domain, respectively.

[0119] The least squares method is repeated until the current model parameter m is reached. c,i Satisfy the fitting error of the fused data or reach the maximum number of iterations:

[0120] m c,i =m i +Δm i ;

[0121] Step 35: Repeat steps 31-34 until all the fused data from N measurement points have been inverted, obtaining the time-frequency fused resistivity distribution of the entire measurement line.

[0122] m c ={m c,1 ,m c,2 ,m c,3 ,…,m c,N-1 ,m c,N}

[0123] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present invention, and these should also be considered within the scope of protection of the present invention. These modifications and improvements will not affect the effectiveness of the implementation of the present invention or the practicality of the patent.

Claims

1. A method for fusing and inverting time-frequency data from ground-to-air electromagnetic detection, characterized in that, Includes the following steps: Step 1: Time-domain ground-to-space electromagnetic detection data conversion; Step 2: Fusion of time and frequency detection data; Step 3: Time-frequency fusion data inversion; Step 1 includes the following steps: Step 11: Perform baseline correction on the time-domain ground-space electromagnetic detection dataset collected for each survey line; Step 12: According to the signal acquisition sequence of the flight path, average the transmission and reception parameters of multiple consecutive measurements near N measurement points on a single measurement line to obtain the transmission and reception parameter sets of all individual measurement points from time-domain detection and save them independently. Step 13: Superimpose and denoise multiple consecutive sets of signals near N measurement points on a single measurement line, and save them independently as the time-domain attenuation data vector of a single measurement point; Step 14: Extract the amplitude spectrum of the time-domain decay data vector in the frequency domain using Fourier transform; Step 15: Calculate the frequency sequence of the corresponding spectrum using the sampling frequency of the time-domain data; Step 16: Repeat steps 13 to 15 to perform time-frequency conversion on the data of all measuring points on the measuring line to obtain the corresponding spectrum dataset and frequency sequence; Step 2 includes the following steps: Step 21: Perform baseline correction on the frequency domain ground-to-air electromagnetic detection dataset collected for each survey line. According to the signal acquisition sequence of the flight route, average the transmission and reception parameters of multiple consecutive measurements near N measurement points on a single survey line to obtain the transmission and reception parameter sets of all individual measurement points from frequency domain detection and save them independently. Step 22: Superimpose and denoise multiple consecutive sets of signals near N measurement points on a single measurement line to obtain spectrum data and corresponding frequency point sequences; Step 23: Normalize the spectral data from the time domain and frequency domain detection modes respectively to obtain the normalized spectral data for each measurement point; Step 24: For a single measurement point, combine the two sets of frequency point sequences from the time domain and frequency domain detection modes into a single frequency point sequence, and sort them from largest to smallest to obtain the time-frequency fusion frequency point sequence of the single measurement point; Step 25: Combine and sort the normalized spectrum data according to the corresponding frequency point positions to obtain the time-frequency fusion spectrum data of a single measurement point; Step 26: Repeat steps 22 to 25 to sort and combine the frequency sequence and spectrum data of all measurement points on the measurement line to obtain the corresponding comprehensive spectrum dataset and frequency sequence; Step 3 includes the following steps: Step 31: Set the initial model of the underground parameters for the i-th measuring point, including the resistivity and thickness of each underground layer, and take the logarithm of the model parameters; Step 32: Establish the objective function for the inversion of fused data; Step 33: Calculate the Jacobian matrix of each parameter using frequency domain ground-to-space electromagnetic forward modeling data; Step 34: Use the least squares method to calculate the changes in model parameters until the current model parameters meet the fitting error of the fused data or reach the maximum number of iterations; Step 35: Repeat steps 31 to 34 until all the fused data from N measurement points have been inverted, and obtain the time-frequency fused resistivity distribution of the entire measurement line.

2. The method for fusing and inverting time-frequency data from ground-to-air electromagnetic detection according to claim 1, characterized in that, The specific processes of steps 13 to 16 are as follows: Step 13: Superimpose and denoise multiple consecutive sets of signals near N measurement points along a single measurement line to obtain the time-domain attenuation data vector for a single measurement point. And save it separately; Step 14: Extract the time-domain decay data vector using Fourier transform. Amplitude spectrum in the frequency domain Where i is the measurement point number, j is the time-domain data number of the time-domain ground-to-air electromagnetic data of a single measurement point, and k is the frequency-domain data number of the time-domain ground-to-air electromagnetic data of a single measurement point. Step 15: Calculate the frequency sequence f of the corresponding spectrum using the sampling frequency of the time-domain data. TEM,i ; Where Fl=[0,1,2,…,L], L is Data length, Fs TEM It is the sampling frequency of the time-domain data; Step 16: Repeat steps 13-15 to process the data of all measuring points on the survey line. Perform time-frequency conversion to obtain the corresponding spectrum dataset. Sum of frequency points f TEM :

3. The method for fusing and inverting time-frequency data from ground-to-air electromagnetic detection according to claim 1, characterized in that, In step 12, the transmit / receive parameter set includes the receive amplification factor mR from the time-domain probe. TEM,i Receiver point coordinates CoR TEM,i emission current IT TEM,i and the coordinates of the emission source CoT TEM,i .

4. The method for fusing and inverting time-frequency data from ground-to-air electromagnetic detection according to claim 1, characterized in that, The specific processes of steps 22 to 26 are as follows: Step 22: Superimpose and denoise multiple consecutive sets of signals near N measurement points along a single measurement line to obtain spectrum data. and the corresponding frequency sequence f FEM,i ; Step 23: Normalize the spectral data from both the time and frequency domain detection modes to obtain the normalized spectral data for each measurement point. and Among them, mR TEM,i It is the receiving amplification factor from time-domain probe, IT TEM,i It is the transmit current from time-domain probe, mR FEM,i It is the receiving amplification factor from frequency domain detection, IT FEM,i It is the transmit current from frequency domain detection; Step 24: For a single measurement point, combine the two sets of frequency sequence f from the time domain and frequency domain detection modes. TEM,i and f FEM,i The frequencies are combined into a set of frequency points and sorted from largest to smallest to obtain the time-frequency fusion frequency point sequence f for a single measurement point. EM,i : f EM,i =sort[f TEM,i ,f FEM,i ]; Step 25: Normalize the spectrum data and By combining and sorting the corresponding frequency points, the time-frequency fusion spectrum data of a single measurement point can be obtained. Step 26: Repeat steps 22-25 to sort and combine the frequency sequence and spectrum data of all measurement points on the measurement line to obtain the corresponding comprehensive spectrum dataset. Sum of frequency points f EM : f EM ={f EM,1 ,f EM,2 ,f EM,3 ,…,f EM,N-1 ,f EM,N }。 5. The method for fusing and inverting time-frequency data from ground-to-air electromagnetic detection according to claim 1, characterized in that, In step 21, the transmit / receive parameter set includes the receive amplification factor mR from frequency domain detection. FEM,i Receiver point coordinates CoR FEM,i emission current IT FEM,i and the coordinates of the emission source CoT FEM,i .

6. The method for fusing and inverting time-frequency data from ground-to-air electromagnetic detection according to claim 1, characterized in that, The specific steps of step 3 are as follows: Step 31: Set the initial model m of the underground parameters for the i-th measuring point. 0,i Including the resistivity ρ of each underground layer i and layer thickness h i And take the logarithm of the model parameters: Step 32: Establish the objective function Φ for the inversion of fused data. EM,i : in, It is normalized detection fusion data. It is based on the fusion frequency point f EM,i and the initial model m i The spectral data obtained from forward modeling, where λ is the regularization parameter and C is the smoothness matrix; Step 33: Calculate the Jacobian matrix G of each parameter using frequency-domain ground-to-space electromagnetic forward modeling data. EM,i : Where, Δρ g It is a resistivity element, Δh g It is a layer thickness micro-element; Step 34: Use the least squares method to calculate the change in model parameters Δm. i : Among them, e TEM,i and e FEM,i These are the normalized detection errors of electromagnetic detection data in the time domain and frequency domain, respectively. The least squares method is repeated until the current model parameter m is reached. c,i Satisfy the fitting error of the fused data or reach the maximum number of iterations: m c,i =m i +Δm i ; Step 35: Repeat steps 31-34 until all the fused data from N measurement points have been inverted, obtaining the time-frequency fused resistivity distribution of the entire measurement line. m c ={m c,1 ,m c,2 ,m c,3 ,…,m c,N-1 ,m c,N }。

Images