Method for measuring and identifying morphologically dependent electromagnetic response electrical parameters based on linear background stripping

Abstract

This invention belongs to the field of electrical and magnetic variable measurement technology, specifically involving a method for measuring and identifying electromagnetic response electrical parameters based on linear background stripping and morphology-related parameters. The method involves a preset moving identification window width; within the selected window range, early and late multi-channel electrical responses of equal time periods are extracted from the first and last channels, respectively, and the summation and average are used to generate early and late characterizing multi-channel electrical responses; linear background removal is performed on the early and late characterizing multi-channel electrical responses to obtain early and late multi-channel abnormal morphology vectors; the correlation coefficient of the early multi-channel electrical responses with the late abnormal morphology is calculated to obtain the induced electromotive force (ICF) response identification factor at the center of the window; the window is moved point by point to obtain the distribution of ICF response identification factors corresponding to the overall data profile and normalized; the distribution and intensity characteristics of the ICF response in transient electromagnetic data are identified through its amplitude distribution, effectively improving the accuracy of ICF response identification in transient electromagnetic data.

Description

Technical Field

[0001] This invention belongs to the field of electrical and magnetic variable measurement technology, and particularly relates to a method for measuring and identifying electromagnetic response electrical parameters based on linear background stripping and morphology-related parameters. Background Technology

[0002] Transient electromagnetic detection (TEM) involves transmitting pulsed electromagnetic fields into the subsurface and observing and measuring the decay characteristics of the secondary eddy current signals induced in the subsurface medium over time, thereby obtaining information on the electrical distribution of the subsurface medium. This method is widely used in resource exploration, engineering geological surveys, and other fields. Its core is the high-precision measurement and analysis of electrical variables such as electromagnetic induced electromotive force, electrical response amplitude, and decay characteristics.

[0003] In actual measurements, the transient electromagnetic signals observed often contain the superposition of induced eddy current electrical response and induced polarization effect. The induced polarization effect causes the measured electrical response curve to exhibit typical characteristics such as abnormal early amplitude rise, rapid mid-to-late stage decay, and even negative values, directly affecting the accuracy of electrical parameter measurements and the reliability of underground structure inversion.

[0004] Currently, the identification and extraction of induced polarization responses in transient electromagnetic signals mainly rely on manual experience and lack automated, quantitative methods for signal measurement and analysis. Traditional identification methods only focus on late-stage negative value characteristics, lacking sufficient sensitivity for measuring weak induced polarization responses that do not exhibit negative values. They are also susceptible to noise interference and cannot achieve accurate identification based on the morphology, amplitude variation, and spatial correlation of the electrical response.

[0005] Meanwhile, existing measurement and analysis methods fail to effectively remove the trend-based electrical response caused by background strata, resulting in the masking of weak induced polarization anomalies and making it difficult to accurately extract electrical signal anomalies. Therefore, developing a method capable of removing linear background, extracting electrical response morphological features, and realizing automatic measurement and identification of induced polarization responses is of great significance for improving the accuracy of transient electromagnetic variable measurement and enhancing the reliability of anomaly identification. Summary of the Invention

[0006] To address the technical problems in existing technologies, such as the heavy reliance on human experience in identifying transient electromagnetic induced electrical responses, insufficient ability to identify weak induced electrical responses without negative values, and low accuracy due to environmental noise interference in single-point analysis, this invention aims to provide a method for measuring and identifying electromagnetic response electrical parameters based on linear background stripping and morphology-related parameters. By constructing multi-channel profile spatial windows, reconstructing early and late-stage electrical response features, fitting and stripping linear background fields, constraining morphological vector polarity, and measuring heterogeneous morphological correlation, this method achieves automatic, quantitative, and highly robust identification of the distribution and intensity of transient electromagnetic induced electrical responses, effectively improving the detection accuracy of polarization anomalies in complex geological environments.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A method for measuring and identifying electromagnetic response electrical parameters based on linear background stripping and morphology-related parameters includes the following steps: S1: Obtain the transient electromagnetic multi-channel data profile to be processed, the multi-channel data profile including N distributed along the survey line direction. s There are 1 measurement point and T time-channel electrical response values ​​corresponding to each measurement point; based on the overall length L of the multi-channel data profile, the width N of the motion recognition window for background correlation removal is preset. w The width N of the motion recognition window w The selection range is set to the total length L of the survey line or the total number of survey points N. s Within one-tenth to one-quarter of the range; S2: Within the currently selected motion recognition window, perform the extraction and reconstruction operations of the early and late characterization electrical responses: starting from the first pass, extract N... T Early-stage multi-channel electrical response measurements were performed, and the same number of channels N were truncated from the last channel forward. T The late-stage multi-channel electrical response, in which the number of time channels N is truncated. T The value is taken from 3 to 5; the early and late multi-channel electrical responses are summed and averaged separately to generate an early characterizing multi-channel electrical response E within the window. e And a late-stage characterization of the multi-channel electrical response E l ; S3: Early characterization of the multichannel electrical response E based on a first-order polynomial function e and late-stage characterization of multichannel electrical response E l Linear background fitting was performed separately to extract the trend-based electrical response background generated by regional stratigraphic structure; the original early characterization multi-channel electrical response E was then used. e Subtract the early multichannel linear background electrical response generated by the fitting. The late-stage characterization of the multi-channel electrical response E l Subtract the late-stage multichannel linear background electrical response generated by the fitting. In order to obtain early multi-channel anomaly morphology vectors and late-stage multi-channel anomaly morphology vector ; S4: For the acquired early multi-channel anomaly morphology vectors and late-stage multi-channel anomaly morphology vector Polarity constraint processing is performed; based on the physical mechanism that the induced polarization response exhibits abnormal rise in the early stage and abnormal sink in the late stage, logical operations are performed on the abnormal morphology vector to eliminate morphological interference caused by non-induced polarization factors and retain effective abnormal morphological information related to induced polarization physical characteristics. S5: Calculate the correlation coefficient α between the processed early multi-channel electrical response and the late-stage anomaly morphology based on the inner product algorithm. The correlation coefficient α reflects the degree of reverse coupling between early and late-stage anomalies in spatial morphological distribution within the window. Extract the negative part of the correlation coefficient α and perform absolute value processing, defining it as the induced electroelectric response identification factor α at the center measurement point of the current moving identification window. IP ; S6: Move the identification window point by point along the survey line direction, repeat steps S2 to S5 until the entire multi-track data profile is traversed to obtain the distribution sequence of induced electric response identification factors corresponding to the entire survey line; normalize the sequence, and by analyzing the amplitude distribution of the normalized induced electric response identification factors, identify the spatial distribution range and intensity characteristics of the induced electric response in the transient electromagnetic data.

[0008] Preferably, in step S1, the width N of the recognition window is moved. w The determination process also includes an adaptive adjustment mechanism: The multi-channel data profile is pre-scanned, and the average coherence length of the data in the spatial domain is analyzed. If the lateral heterogeneity of the geological structure in the detection area exceeds a preset threshold, then a window width N is selected. w The window width N is one-tenth of the survey line length; if the geological structure of the detection area is relatively flat and the spatial coherence length is greater than the preset length threshold, then the window width N is selected. w It is one-quarter of the length of the survey line; The adaptive adjustment mechanism ensures that the motion recognition window contains sufficient background information to support background field extraction, while also ensuring that the window can capture polarization anomalies with local characteristics.

[0009] Preferably, step S2 further includes, within a selected window range, extracting three to five early and late multi-channel electrical responses with the same number of channels at different times, summing and averaging the selected early and late multi-channel electrical responses respectively, to generate the early characterizing multi-channel electrical response E. e and late-stage characterization of multichannel electrical response E l : Early characterization of multichannel electrical response E e The calculation formula is as follows: ; Late characterization of multichannel electrical response E l The calculation formula is as follows: ; Where, N T The number of time channels to be captured is typically 3 to 5; T is the total number of time channels; E[i] represents the multi-channel response of the i-th time channel, and E[j] represents the multi-channel response of the j-th time channel; By summing and averaging the multichannel electrical responses, the interference of random noise on the single-channel electrical response is suppressed, thereby enhancing the signal-to-noise ratio of the characterization signal.

[0010] Preferably, in step S2, before performing the summation and averaging process on the extracted early and late traces, a process of normalizing and aligning each time trace is also included: For each time channel selected within the window, calculate the average amplitude of that time channel at all measurement points within the current window; Divide the original electrical response value of each measuring point within the time channel by the mean amplitude to eliminate the magnitude difference caused by electromagnetic attenuation between different time channels, and ensure that the spatial morphological features of different time scales have equal weight in the subsequent reconstruction process.

[0011] Preferably, in step S3, the early and late stages of the multi-channel electrical response characterization are fitted using a first-order polynomial function, and the fitting function expression is as follows: ; Where x is the lateral distance coordinate of the profile, E b To generate a fitted linear background electrical response value, a and b are the coefficients of a first-order polynomial, and their expressions are as follows: ; ; Where, N w x is the width of the selected time window. i E(i) represents the horizontal distance coordinate corresponding to the i-th position in the window; E(i) represents the response value corresponding to the i-th position in the current motion recognition window.

[0012] First-order polynomial fitting generates early multi-channel linear background electrical response. and late-stage multi-channel linear background electrical response The early and late-stage multichannel anomaly morphology vectors are obtained by subtracting the linear background electrical response from the characterizing multichannel electrical response. and The expression to be evaluated is: ; .

[0013] Preferably, in step S3, before performing the linear background fitting operation, the early characterization of the multichannel electrical response E is further performed. e and late-stage characterization of multichannel electrical response E l The steps for median filtering preprocessing are as follows:

[0014] A sliding median filter with a preset step size is used to smooth the electrical response vector to remove outliers caused by sensor jumps, local metal pipeline interference, or random strong noise. This prevents the outliers from pulling the trend of the linear background electrical response during the least squares fitting process, ensuring that the abnormal morphological vector after background removal can accurately reflect the morphological distortion of the underground polaribody.

[0015] Preferably, in step S4, the polarity constraint logic operation performed on the early and late multi-channel anomaly morphology vectors is specifically as follows: Determining early multi-channel anomaly morphology vectors If the value of each element in the array is negative, it is forced to be zero, and only the positive lifting component reflecting the induced charging effect is retained. Determining late-stage multi-channel anomaly morphology vectors If the value of an element is positive, it is forcibly set to zero, and only the negative sedimentation component reflecting the induced discharge effect is retained. Through the aforementioned polarity constraint, the induced polarization physical feature screening is enforced at the algorithm's underlying layer to exclude unidirectional electromagnetic induction morphological fluctuations caused by high surface conductivity interference or terrain undulations.

[0016] Preferably, step S5 further includes performing an inner product operation on the processed early and late multi-channel anomaly morphology vectors to calculate the correlation coefficient α between the early multi-channel electrical response and the late anomaly morphology: ; Here, dot represents the inner product operation, and the resulting abnormal shape correlation coefficient α is the correlation coefficient between the early multi-channel electrical response and the later abnormal morphology. The transient electromagnetic induced electrical response is characterized by an early rise and a late sink, therefore the early and late stages will show a negative correlation. Furthermore, the larger the α value, the greater the morphological difference, indicating a stronger induced electrostatic response. The absolute value of the obtained abnormal shape correlation coefficient is then taken as the induced electrostatic response identification factor α at the center measurement point of the window. IP .

[0017] Preferably, after obtaining the normalized induced polarization response identification factor distribution sequence in step S6, the method further includes an adaptive threshold determination criterion to output an induced polarization warning signal: Calculate the mean μ and standard deviation σ of the normalized identification factor sequence of the whole profile, and set the judgment threshold Th=μ+kσ according to the noise level of the detection environment, where the coefficient k ranges from 1.5 to 2.5; The normalized identification factor α norm The comparison is performed point by point with the judgment threshold Th. When α normWhen the value is >Th, it is determined that there is a significant induced electrical response in the area of ​​the measuring point. In combination with spatial consistency verification, isolated high-value points are eliminated, thereby realizing automatic and robust positioning of underground mineralization alteration zones or polarization anomalies.

[0018] The beneficial effects of this invention include: 1. Breaking away from the limitations of traditional induced polarization response identification that relies solely on single-point decay curves, this method introduces a moving identification window to achieve in-depth mining of spatial correlation information across multiple measurement points. Utilizing morphological features on the cross-section for identification effectively overcomes the susceptibility of single-point data to random noise and instrument drift, significantly improving the robustness and reliability of the identification results.

[0019] 2. A first-order polynomial fitting method is used to extract and remove the linear background field from the early and late characterization electrical responses. This technique effectively eliminates the trend-based electromagnetic response caused by the resistivity structure of the surrounding rock, allowing the weak induced polarization anomalies buried deep in the background field to be clearly revealed. Through background removal processing, this invention exhibits extremely high detection sensitivity for weak induced polarization responses that do not show negative values ​​and only exhibit changes in the attenuation slope.

[0020] 3. Fully utilizing the fundamental physical law of "early rise and late fall" in transient electromagnetic induced polarization (EMI), a scientific EMI response identification factor is constructed by applying polarity constraints to the abnormal morphology vector and combining it with an inner product correlation algorithm. This criterion has clear physical significance and can accurately distinguish between EMI response and simple low-resistance or high-resistance induced electrical response, greatly reducing the false alarm rate and missed alarm rate of EMI identification.

[0021] 4. The algorithm involved does not rely on complex nonlinear inversion or iterative optimization processes, resulting in extremely low computational overhead. The generated induced polarization response identification factor can intuitively and quantitatively reflect the distribution and intensity of underground polaritons, providing an accurate basis for subsequent data interpretation, parameter extraction, and borehole layout.

[0022] In summary, this invention solves the technical pain point of difficult and easily misjudged identification of induced electrical response in transient electromagnetic exploration through an innovative architecture that eliminates background correlation. It has important practical application value in improving the accuracy of electromagnetic data interpretation and the effectiveness of geological prospecting. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating the method for measuring and identifying electromagnetic response electrical parameters based on linear background stripping and morphology-related parameters according to the present invention.

[0024] Figure 2 This is a schematic diagram of the multi-channel data profile of the transient electromagnetic induced electro-electric response in a three-dimensional simulation.

[0025] Figure 3To identify the induced electroelectric response of measured transient electromagnetic data. Detailed Implementation

[0026] This invention provides a method for measuring and identifying electromagnetic response electrical parameters based on linear background stripping and morphology-related parameters. To make the technical solution and implementation effects of this invention clearer, the following is in conjunction with the appendix. Figures 1-3 The present invention will be further described in detail below: Example 1

[0027] In engineering practice of transient electromagnetic detection, the observed secondary field induced electromotive force signal is usually a superposition of pure induced eddy current response and excited polarization response. In conventional conductive media, the induced response monotonically decays over time and its polarity remains constant; however, when a polarizing body (metal sulfide, graphite, etc.) is present underground, an accumulation and release process of polarization charge occurs under electromagnetic field excitation. This process manifests as a delayed reverse current in the time domain, which is reflected in the observation curve as an abnormal rise in the late signal, a change in slope, and even a reversal of polarity. The electromagnetic response electrical parameter measurement and identification method based on linear background stripping and morphology-related parameters described in this invention utilizes the localized characteristics of the excited electromotive force response in the spatial profile and its polarity evolution characteristics in the time domain. By stripping the background in the spatial dimension and measuring the correlation in the time dimension, polarization information can be extracted.

[0028] A method for measuring and identifying electromagnetic response electrical parameters based on linear background stripping and morphology-related parameters is provided in the appendix. Figure 1 As shown, it includes the following steps: S1: Obtain the transient electromagnetic multi-channel data profile to be processed, the multi-channel data profile including N distributed along the survey line direction. s There are 1 measurement point and T time-channel electrical response values ​​corresponding to each measurement point; based on the overall length L of the multi-channel data profile, the width N of the motion recognition window for background correlation removal is preset. w The width N of the motion recognition window w The selection range is set to the total length L of the survey line or the total number of survey points N. s Within one-tenth to one-quarter of the range.

[0029] S2: Within the currently selected motion recognition window, perform the extraction and reconstruction operations of the early and late characterization electrical responses: starting from the first pass, extract N... T Early-stage multi-channel electrical response measurements were performed, and the same number of channels N were truncated from the last channel forward. T The late-stage multi-channel electrical response, in which the number of time channels N is truncated. T The value is taken from 3 to 5; the early and late multi-channel electrical responses are summed and averaged separately to generate an early characterizing multi-channel electrical response E within the window.e And a late-stage characterization of the multi-channel electrical response E l .

[0030] S3: Early characterization of the multichannel electrical response E based on a first-order polynomial function e and late-stage characterization of multichannel electrical response E l Linear background fitting was performed separately to extract the trend-based electrical response background generated by regional stratigraphic structure; the original early characterization multi-channel electrical response E was then used. e Subtract the early multichannel linear background electrical response generated by the fitting. The late-stage characterization of the multi-channel electrical response E l Subtract the late-stage multichannel linear background electrical response generated by the fitting. In order to obtain early multi-channel anomaly morphology vectors and late-stage multi-channel anomaly morphology vector .

[0031] S4: For the acquired early multi-channel anomaly morphology vectors and late-stage multi-channel anomaly morphology vector Polarity constraint processing is performed; based on the physical mechanism that the induced polarization response exhibits abnormal rise in the early stage and abnormal sink in the late stage, logical operations are performed on the abnormal morphology vector to eliminate morphological interference caused by non-induced polarization factors and retain effective abnormal morphological information related to induced polarization physical characteristics.

[0032] S5: Calculate the correlation coefficient α between the processed early multi-channel electrical response and the late-stage anomaly morphology based on the inner product algorithm. The correlation coefficient α reflects the degree of reverse coupling between early and late-stage anomalies in spatial morphological distribution within the window. Extract the negative part of the correlation coefficient α and perform absolute value processing, defining it as the induced electroelectric response identification factor α at the center measurement point of the current moving identification window. IP .

[0033] S6: Move the identification window point by point along the survey line direction, repeat steps S2 to S5 until the entire multi-track data profile is traversed to obtain the distribution sequence of induced electric response identification factors corresponding to the entire survey line; normalize the sequence, and by analyzing the amplitude distribution of the normalized induced electric response identification factors, identify the spatial distribution range and intensity characteristics of the induced electric response in the transient electromagnetic data.

[0034] In this embodiment, the recognition window width N is moved in step S1. w The determination process also includes an adaptive adjustment mechanism: The multi-channel data profile is pre-scanned, and the average coherence length of the data in the spatial domain is analyzed. If the lateral heterogeneity of the geological structure in the detection area exceeds a preset threshold, then a window width N is selected.w The window width N is one-tenth of the survey line length; if the geological structure of the detection area is relatively flat and the spatial coherence length is greater than the preset length threshold, then the window width N is selected. w It is one-quarter of the length of the survey line.

[0035] The adaptive adjustment mechanism ensures that the motion recognition window contains sufficient background information to support background field extraction, while also ensuring that the window can capture polarization anomalies with local characteristics.

[0036] Step S2 further includes extracting three to five early and late multi-channel electrical responses with the same number of channels within a selected window, summing and averaging the selected early and late multi-channel electrical responses to generate the early characterizing multi-channel electrical response E. e and late-stage characterization of multichannel electrical response E l : Early characterization of multichannel electrical response E e The calculation formula is as follows: ; Late characterization of multichannel electrical response E l The calculation formula is as follows: ; Where, N T The number of time channels to be captured is typically 3 to 5; T is the total number of time channels; E[i] represents the multi-channel response of the i-th time channel, and E[j] represents the multi-channel response of the j-th time channel.

[0037] By summing and averaging the multichannel electrical responses, the interference of random noise on the single-channel electrical response is suppressed, thereby enhancing the signal-to-noise ratio of the characterization signal.

[0038] Regarding the selection of time channels in step S2, an optimization strategy based on sampling rate and signal-to-noise ratio can also be provided. In different transient electromagnetic detection missions, the distribution of time channels often follows logarithmic or linear equal intervals. This invention selects the early channel N... T When selecting channels, priority is given to the first three stable windows after the primary field shutdown to avoid interference from the residual current of the instrument shutdown. However, when selecting late channels, the background noise level (standard deviation) of each channel is estimated in real time, and the last N channels with a signal-to-noise ratio greater than 3 are dynamically selected. T One channel. If the late signal has completely sunk below the noise floor, this method will automatically reduce N. T The value of can be selected, or higher-order low-pass filtering preprocessing can be performed on the last few available channels to ensure that the abnormal shape vectors participating in the inner product operation are not dominated by pure random noise.

[0039] Example 2

[0040] Based on Example 1, in step S2, before performing the summation and averaging process on the extracted early and late traces, a normalization and alignment process for each time trace is also included: For each time channel selected within the window, calculate the average amplitude of that time channel at all measurement points within the current window; Divide the original electrical response value of each measuring point within the time channel by the mean amplitude to eliminate the magnitude difference caused by electromagnetic attenuation between different time channels, and ensure that the spatial morphological features of different time scales have equal weight in the subsequent reconstruction process.

[0041] In step S3, the early and late stages of the multi-channel electrical response characterization are fitted using a first-order polynomial function. The fitting function expression is as follows: ; Where x is the lateral distance coordinate of the profile, E b To generate a fitted linear background electrical response value, a and b are the coefficients of a first-order polynomial, and their expressions are as follows: ; ; Where, N w x is the width of the selected time window. i E(i) represents the horizontal distance coordinate corresponding to the i-th position in the window; E(i) represents the response value corresponding to the i-th position in the current motion recognition window.

[0042] First-order polynomial fitting generates early multi-channel linear background electrical response. and late-stage multi-channel linear background electrical response The early and late-stage multichannel anomaly morphology vectors are obtained by subtracting the linear background electrical response from the characterizing multichannel electrical response. and The expression to be evaluated is: ; .

[0043] In step S3, before performing the linear background fitting operation, the early characterization of the multichannel electrical response E is further performed. e and late-stage characterization of multichannel electrical response E l The steps for median filtering preprocessing are as follows: A sliding median filter with a preset step size is used to smooth the electrical response vector to remove outliers caused by sensor jumps, local metal pipeline interference, or random strong noise. This prevents the outliers from pulling the trend of the linear background electrical response during the least squares fitting process, ensuring that the abnormal morphological vector after background removal can accurately reflect the morphological distortion of the underground polaribody.

[0044] Example 3

[0045] Based on Example 1 or Example 2, the polarity constraint logic operation performed on the early and late multi-channel anomaly morphology vectors in step S4 is specifically as follows: Determining early multi-channel anomaly morphology vectors If the value of each element in the array is negative, it is forced to be zero, and only the positive lifting component reflecting the induced charging effect is retained. Determining late-stage multi-channel anomaly morphology vectors If the value of an element is positive, it is forcibly set to zero, and only the negative sedimentation component reflecting the induced discharge effect is retained. Through the aforementioned polarity constraint, the induced polarization physical feature screening is enforced at the algorithm's underlying layer to exclude unidirectional electromagnetic induction morphological fluctuations caused by high surface conductivity interference or terrain undulations.

[0046] When applying polarity constraints, considering the possibility of extremely slight drift in the measured data, an absolute zero threshold is not used; instead, a small noise tolerance threshold δ is set. Specifically, the early vectors are allowed to satisfy... The component with less than δ is zero, so that the late vector satisfies The component of >−δ is zero. The value of δ is usually set to 0.5 times the standard deviation of the electrical response amplitude within the window. This improvement enables the algorithm to exhibit extremely strong numerical stability when processing massive amounts of data in a high-noise background, avoiding correlation coefficient fluctuations caused by minute numerical variations.

[0047] Step S5 further includes performing an inner product operation on the processed early and late multi-channel anomaly morphology vectors to calculate the correlation coefficient α between the early multi-channel electrical response and the late anomaly morphology: ; Here, dot represents the inner product operation, and the resulting abnormal shape correlation coefficient α is the correlation coefficient between the early multi-channel electrical response and the later abnormal morphology. The transient electromagnetic induced electrical response is characterized by an early rise and a late sink, therefore the early and late stages will show a negative correlation. Furthermore, the larger the α value, the greater the morphological difference, indicating a stronger induced electrostatic response. The absolute value of the obtained abnormal shape correlation coefficient is then taken as the induced electrostatic response identification factor α at the center measurement point of the window. IP .

[0048] After obtaining the normalized induced polarization response identification factor distribution sequence in step S6, the method further includes an adaptive threshold determination criterion to output an induced polarization warning signal. Calculate the mean μ and standard deviation σ of the normalized identification factor sequence of the whole profile, and set the judgment threshold Th=μ+kσ according to the noise level of the detection environment, where the coefficient k ranges from 1.5 to 2.5; The normalized identification factor α norm The comparison is performed point by point with the judgment threshold Th. When α norm When the value is >Th, it is determined that there is a significant induced electrical response in the area of ​​the measuring point. In combination with spatial consistency verification, isolated high-value points are eliminated, thereby realizing automatic and robust positioning of underground mineralization alteration zones or polarization anomalies.

[0049] Example 4

[0050] Based on Example 1, Example 2, or Example 3, a method for measuring and identifying electromagnetic response electrical parameters based on linear background stripping and morphology-related parameters is provided. Figure 2 Image (a) shows a blocky polarimetric geological body model with dimensions of 140m × 140m × 80m and a top burial depth of 40m. The polarimetric body is described using the Cole-Cole model parameters, specifically: zero-frequency conductivity of 0.01 S / m, charge rate of 0.5, time constant of 0.2 s, and frequency correlation coefficient of 0.5. The background is a high-resistivity half-space with a conductivity of 3 × 10⁻⁶. -4 S / m. A three-dimensional forward modeling simulation was performed on this model. Figure 2 (b) shows the transient electromagnetic multichannel electrical response profile of the central section of the model, where the dashed lines represent negative electrical responses. It can be seen that above the corresponding polarization anomaly, the multichannel electrical response exhibits a typical characteristic of an early bulge, a late depression, and ultimately, a negative value. Figure 2 In the middle (c), the single-point decay curve is located at the center of the profile. Overall, it shows an early large value, followed by a rapid decay to a negative value in the middle and late stages.

[0051] like Figure 3 As shown, this case study provides a method for measuring and identifying electromagnetic response electrical parameters based on linear background stripping and morphology-related parameters. Figure 3 (a) shows the measured transient electromagnetic multi-channel electrical response profile of a metal sulfide ore area. It can be seen that, influenced by the induced polarization effect of the metal sulfide ore body, the transient electromagnetic data exhibits multiple induced polarization response characteristics, characterized by a pronounced bulge in the early stage and a concave shape in the later stage, with some areas showing continuous negative electrical responses. The distribution of the induced polarization response identification factor was obtained using the technique of this invention on this multi-channel data, and the results are as follows: Figure 3 As shown in (b), the distribution of its amplitude can effectively reflect the distribution of the induced polarization response, enabling rapid identification of the induced polarization response. At the same time, the magnitude of the amplitude also reflects the intensity of the induced polarization response at that location, further characterizing the size and polarization strength of the underground polaribody.

[0052] In summary, the method of this invention can effectively identify a weak polarization anomaly in the mining area where the induced electromotive force (EMF) sign has not changed due to the deep ore body. On the original induced EMF curve, this anomaly only shows a slight slowdown in the late-stage decay rate, which traditional methods easily classify as induced fluctuations in high-resistivity surrounding rock. However, after step S3 of this invention strips away the linear downward trend of the surrounding rock, the anomaly morphology vector... A clear negative pulse is displayed at the corresponding location, and this pulse corresponds to the early anomalous morphological vector. The positive pulses overlap highly in space, and after the inner product operation in step S5, a significant recognition factor peak is generated.

[0053] For large-scale engineering applications, after the normalization process in step S6, this invention can introduce an adaptive threshold judgment based on probability distribution. Assuming the full-profile identification factor sequence follows a generalized extreme value distribution, a discrimination threshold T = μ + 2.5σ is set by calculating the mean μ and standard deviation σ of the sequence. Only when the identification factor exceeds this threshold is it marked as an induced polarization anomaly area on the interpretation map. This successfully eliminates geometric induction distortion caused by terrain at the end of the survey line, further reducing the false alarm rate. In actual hardware implementation or software integration, the algorithm described in this invention has extremely high computational efficiency because it does not involve complex matrix inversion or nonlinear iteration, enabling second-level automatic full-profile identification on conventional portable processing terminals. For high-dynamic, large-data-volume detection scenarios such as airborne transient electromagnetic fields, this method can be integrated into an airborne real-time processing system to delineate the polarimetric distribution range in real time during flight, providing decision support for intensified field detection.

[0054] The above embodiments are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any changes made based on the design principles of the present invention, or any non-creative modifications made thereon, shall fall within the scope of protection of the present invention.

Claims

1. A method of measuring and identifying morphologically dependent electromagnetic response electrical parameters based on linear background stripping, characterized in that, The method comprises the following steps: S1: acquire a transient electromagnetic multi-channel data profile to be processed, the multi-channel data profile containing N s measurement points distributed along the direction of the measurement line and T time channel electric response values corresponding to each measurement point; according to the overall length L of the multi-channel data profile, preset a moving identification window width N w for performing background correlation operation; S2: In the range of N w , the extraction and reconstruction operation of early and late period representation electric response is performed: N T early multi-channel electric responses are intercepted from the first trace, and N T late multi-channel electric responses are intercepted from the last trace; the intercepted early and late multi-channel electric responses are respectively processed by summation and averaging, so as to generate one early period representation multi-channel electric response E e and one late period representation multi-channel electric response E l in the window. S3: Linear background fitting of the generated early multi-channel electric response E e and late multi-channel electric response E l respectively; subtract the generated early multi-channel linear background electric response e , subtract the generated late multi-channel linear background electric response l , obtain early multi-channel anomaly shape vector and late multi-channel anomaly shape vector ;​​ S4: Perform polarity constraint processing on the acquired early multi-channel anomaly shape vectors and late multi-channel anomaly shape vectors based on the physical mechanism of the induced polarization response that appears as an anomaly uplift in the early stage and an anomaly subsidence in the late stage, perform logical operation on the anomaly shape vectors; S5: Based on the inner product algorithm, the abnormal morphological correlation coefficient a of the processed early multi-channel electric response compared with the late one is calculated; the negative value part of the correlation coefficient a is extracted and absolute value processing is carried out, which is defined as the induced polarization electric response identification factor a at the center measuring point position of the current moving identification window IP ; S6: moving the identification window point by point along the measuring line direction, repeatedly performing steps S2 to S5 until the entire multi-channel data profile is traversed, obtaining a distribution sequence of the IP response identification factor corresponding to the entire measuring line; and performing normalization processing on the sequence, identifying the spatial distribution range and intensity characteristics of the IP response in the transient electromagnetic data by analyzing the amplitude distribution of the normalized IP response identification factor.

2. The method of claim 1, wherein, Moving the recognition window width N in step S1 w The determination process also includes an adaptive adjustment mechanism: The multi-channel data profile is pre-scanned to analyze the average coherence length of data in the spatial domain; if the lateral heterogeneity of the geological structure of the detection area exceeds a preset threshold, a window width N w is selected, which is one tenth of the length of the survey line; if the geological structure of the detection area is relatively flat and the spatial coherence length is greater than a preset length threshold, a window width N w is selected, which is one fourth of the length of the survey line.

3. The method of claim 2, wherein, In step S2, early and late multi-channel electric responses of three to five channels are respectively intercepted in the selected window range, and the selected early and late multi-channel electric responses are respectively processed by summation and averaging to generate early multi-channel electric response E e and late multi-channel electric response E l : Early characterization of multi-channel electroresponse E e The formula for calculation is as follows: ； Late time multi-channel electrical response E l The formula for calculating the late time multi-channel electrical response E is as follows: ； where N T is the number of time traces taken; T is the total number of time traces; E[i] represents the multi-trace response of the ith time trace, and E[j] represents the multi-trace response of the jth time trace. By summing and averaging the multi-channel electric response, the interference of random noise on the single-channel electric response is suppressed, and the signal-to-noise ratio of the characteristic signal is enhanced.

4. The method of claim 2, wherein, In step S2, before the summing and averaging processing of the intercepted early and late channels, a normalization alignment process of each time channel is further included: For each time channel selected in the window, the amplitude mean value of the time channel at all measuring points in the current window is calculated; and the original electric response value of each measuring point in the time channel is divided by the amplitude mean value.

5. The method of claim 1, wherein, In step S3, the early and late characteristic multi-channel electric responses are respectively fitted by a first-order polynomial function, and the fitting function expression is as follows: ； where x is the cross-sectional lateral distance coordinate, E b To generate the fitted linear background electrical response values, a and b are the first-order polynomial coefficients, and the expressions are solved for: ； ； wherein N w is a selected time window width; x i denotes a lateral distance coordinate corresponding to the i-th position in the window; E(i) denotes a response value corresponding to the i-th position in the current moving recognition window; First order polynomial fitting generates early multi-channel linear background electroresponse and late multi-channel linear background electroresponse Subtracting the linear background electroresponse from the multi-channel electroresponse yields early and late multi-channel anomaly shape vectors and The expression is: ； 。 6. The method of claim 5, wherein, In step S3, before performing the linear background fitting operation, further includes a step of performing median filtering preprocessing on the early characterization multi-channel electric response E e and the late characterization multi-channel electric response E l The step of performing median filtering preprocessing A sliding median filter with a preset step is used to smooth the electric response vector, and outliers caused by sensor jump, local metal pipeline interference or random strong noise are removed, so that the outliers do not drag and deviate the trend of the linear background electric response in the least square fitting process.

7. The method of claim 1, wherein, In step S4, the polarity constraint logic operation performed on the early and late multi-channel abnormal morphology vectors is specifically: Determining early multi-channel anomaly shape vectors If the element value is negative, it is forced to be zero, only keeping the positive uplift component reflecting the IP charge effect. Determining each element value in a late multi-channel anomaly shape vector If the element value is positive, it is forced to be zero, only keeping the negative subsidence component reflecting the induced polarization discharge effect. Through the polarity constraint, the IP physical feature screening is forced to be performed at the bottom of the algorithm, and the same direction electromagnetic induction fluctuation caused by the surface high conductivity interference or the terrain fluctuation is excluded.

8. The method of claim 1, wherein, In step S5, the inner product operation is further performed on the processed early and late multi-channel abnormal morphology vectors, and the abnormal morphology correlation coefficient α of the early multi-channel electric response compared with the late one is calculated: ； Wherein, dot represents the inner product operation, and α is the abnormal morphology correlation coefficient of the early multi-channel electric response compared with the late one; The transient electromagnetic induced electric response is characterized by early uplift and late subsidence, so that the early and late stages show a negative correlation feature, that is , and the greater the value of a, the greater the morphological difference, indicating that the induced electric response is stronger. The absolute value of the obtained abnormal shape correlation coefficient is taken as the induced polarization response identification factor α at the center of the window IP .

9. The method of claim 8, wherein, After obtaining the normalized IP electric response identification factor distribution sequence in step S6, a step of introducing an adaptive threshold judgment criterion to output an IP warning signal is further included: The mean μ and the standard deviation σ of the full profile normalized identification factor sequence are calculated, and the judgment threshold Th=μ+kσ is set according to the noise level of the detection environment, wherein the value range of the coefficient k is 1.5 to 2.5; The normalized recognition factor α norm Point-by-point comparison with the determination threshold Th, when α norm > Th, it is determined that there is a significant induced polarization response in the measurement point area, and isolated recognition high value points are removed in combination with spatial consistency check, realizing automatic and stable positioning of underground mineralization alteration zone or polarization anomaly body.

Images