Submarine high-conductivity target transient electromagnetic feature extraction and differential background suppression decision method

Abstract

The application discloses a seabed high-conductive target transient electromagnetic feature extraction and differential background suppression decision method, belongs to the field of seabed high-conductive target detection, and comprises the following steps: S1, setting parameters; S2, acquiring transient electromagnetic total field response data of each measuring point in a measuring line or a measuring area; S3, obtaining secondary field response data of the seabed high-conductive target; S4, forming a target attitude response feature set; S5, determining a buried depth attenuation coefficient; S6, obtaining differential processing response data; S7, determining a differential maximum amplitude; S8, forming a differential applicability evaluation index set; S9, constructing an intelligent analysis Agent for seabed transient electromagnetic detection, and generating a unified fusion feature vector; and S10, outputting a result. The seabed high-conductive target transient electromagnetic feature extraction and differential background suppression decision method realizes the whole-process quantification and integration of target feature extraction, background suppression evaluation and intelligent auxiliary decision.

Description

Technical Field

[0001] This invention relates to the field of seabed high conductivity target detection technology, and in particular to a method for extracting transient electromagnetic features of seabed high conductivity targets and a differential background suppression decision method. Background Technology

[0002] With the continuous growth in demand for marine resource development and utilization, subsea engineering construction, subsea pipeline operation and maintenance, and underwater security, the detection and identification of underwater unexploded ordnance, sunken metallic objects, and other highly conductive targets on the seabed has become increasingly important. Among them, underwater unexploded ordnance is characterized by high conductivity and small size. Its burial state and spatial attitude are uncertain, and it is often located in a conductive background formed by the coupling of seawater and sediment layers, which brings significant technical challenges to the interpretation of target electromagnetic response and accurate identification of anomalies.

[0003] Acoustic detection and passive magnetic detection are conventional techniques for underwater target detection, but their applications have significant limitations: acoustic detection is susceptible to sound scattering characteristics, multipath propagation effects, and environmental noise interference, resulting in insufficient detection stability; passive magnetic detection is only sensitive to ferromagnetic targets, and its target identification capability decreases significantly in complex magnetic backgrounds or weak magnetic target scenarios. Transient electromagnetic methods, by exciting conductive targets with artificial field sources to generate induced eddy currents, can induce a clear secondary field response in highly conductive metallic targets, possessing the technical advantage of being suitable for detecting underwater unexploded ordnance and various highly conductive targets.

[0004] Current research on transient electromagnetic detection of unexploded ordnance underwater mainly focuses on target identification, parameter inversion, or response characteristic modeling under ideal conditions. It has not yet systematically established the correspondence between target attitude, burial depth, and secondary field spatial response images in underwater environments. In the center loop point-by-point scanning detection mode, changes in target tilt angle, rotation angle, and burial depth couple and affect the peak response distribution, principal axis direction, and amplitude attenuation law. Relying solely on response images or curves for qualitative judgment cannot provide a stable and reliable basis for target attitude identification and burial depth estimation.

[0005] The transient electromagnetic field signal observed underwater in practice is composed of the target's secondary field response and the background responses of seawater and sediment layers. Symmetrical measurement point differential processing can reduce the smoothing background interference caused by near-common modes in space, but its effectiveness is highly dependent on the lateral uniformity of the background field. When there are significant undulations in the seabed topography, changes in the electrical properties of the sediment layer, or significant lateral differences in water depth, background interference is easily retained after differential processing, and even false anomalies may occur, misleading the target identification results. Meanwhile, traditional data processing workflows rely excessively on human experience, independently interpreting indicators such as response images, attenuation amplitude, and differential residues, making it difficult to form a unified and stable processing scheme under multi-parameter and multi-background conditions. Therefore, there is an urgent need to construct an integrated processing method that can simultaneously achieve target response feature extraction, depth-based attenuation relationship characterization, and differential background suppression applicability evaluation, and rely on intelligent analysis agents to complete multi-feature fusion reasoning and auxiliary decision-making.

[0006] Artificial intelligence technology has been widely applied in fields such as target recognition, signal processing, and intelligent decision-making. However, for the transient electromagnetic detection scenario of underwater unexploded ordnance, a single response amplitude, a single set of differential results, or a single type of image feature cannot fully characterize the target state and the degree of background interference. In the context of complex seawater-sediment layers, the coupled effects of factors such as target attitude, burial depth, background lateral non-uniformity, and observation noise further increase the difficulty of response interpretation. Summary of the Invention

[0007] The purpose of this invention is to provide a method for extracting transient electromagnetic features of highly conductive targets on the seabed and for differential background suppression decision-making, thereby solving the aforementioned technical problems.

[0008] To achieve the above objectives, this invention provides a method for extracting transient electromagnetic features of highly conductive seabed targets and for differential background suppression decision-making, comprising the following steps: S1. Set the seabed exploration environment parameters, high-conductivity target parameters, and transient electromagnetic observation parameters of the center loop, and determine the exploration background consisting of the seawater layer and sediment layer; S2. Based on the transient electromagnetic observation parameters of the center loop determined in S1, the transmitting coil and receiving coil deployed at the same point are moved synchronously using the point-by-point scanning method of the center loop to obtain the transient electromagnetic total field response data of each measuring point in the measuring line or measuring area. S3. Based on the transient electromagnetic observation parameters of the center loop determined in S1, perform the primary field calculation of the circular loop. Combined with the orthogonal magnetic dipole equivalent model, decompose the high-conductivity target on the seabed into three mutually orthogonal magnetic moment components, and calculate the secondary field response data of the high-conductivity target on the seabed. S4. Based on the secondary field response data of the high-conductivity target on the seabed obtained in S3, extract the peak distribution pattern, principal axis direction and component amplitude characteristics of the response of the high-conductivity target on the seabed under different tilt and rotation angles to form a target attitude response feature set. S5. Based on the secondary field response data of the high-conductivity target on the seabed obtained in S3, extract the maximum amplitude of the target response under different burial depth conditions and complete the normalization process, establish the empirical exponential decay relationship between the peak value of the target response and the burial depth, and determine the burial depth attenuation coefficient. S6. Based on the transient electromagnetic total field response data obtained in S2, select measurement points that are symmetrical about the center of the measurement line or the target projection position, perform differential post-processing of symmetrical measurement points, and obtain differential processed response data. S7. Based on the seabed exploration environment parameters determined in S1, a simplified background field model is constructed, and the background lateral non-uniform intensity is quantified. Combined with the differential processing response data obtained in S6, the maximum amplitude of the background before differential processing, the maximum amplitude of the residual background after differential processing, and the maximum amplitude of the target anomaly after differential processing are determined. S8. Based on the various amplitude parameters obtained in S7, calculate the residual ratio, background suppression rate and target anomaly contrast, quantitatively evaluate the applicability of differential post-processing for symmetrical measurement points, and form a set of differential applicability evaluation indicators. S9. Construct an intelligent analysis agent for seabed transient electromagnetic detection. The target attitude response feature set obtained in S4, the burial depth attenuation coefficient obtained in S5, the differential applicability evaluation index set obtained in S8, the center loop transient electromagnetic observation parameters determined in S1, and the detection environment noise level are fused from multiple sources to generate a unified fused feature vector. S10, based on the fused feature vector obtained from S9, uses an intelligent analysis agent to perform rule-based reasoning, machine learning discrimination, and credibility assessment, outputting the attitude discrimination result of the high-conductivity target on the seabed, burial depth level, target anomaly visibility level, differential processing credibility, background suppression strategy, and observation parameter adjustment suggestions.

[0009] Therefore, the present invention employs the above-mentioned transient electromagnetic feature extraction and differential background suppression decision-making method for high-conductivity targets on the seabed, which has the following beneficial effects: 1. Establish a quantitative correspondence between the attitude of underwater high-conductivity targets and the secondary field response images, providing a stable and reliable basis for target attitude recognition.

[0010] 2. Construct an empirical index decay model for burial depth to achieve a quantitative characterization of the response amplitude as a function of burial depth, providing a quantitative reference for estimating the target burial depth level; 3. Using symmetrical measurement point difference as a post-processing method for scanning data can effectively reduce the spatial common mode smoothing background and highlight local target anomalies; 4. By using three indicators—residual ratio, background suppression rate, and target anomaly contrast—the applicable boundaries of differential background suppression are clarified to avoid false anomalies and misjudgments in strong non-uniform backgrounds. 5. The technical process is clear and standardized, and can be adapted to the detection, interpretation and background suppression of various small-scale high-conductivity targets such as underwater unexploded ordnance, seabed pipelines, and sunken metal objects; 6. Relying on intelligent analysis agents to complete multi-source feature fusion reasoning, replacing independent judgment based on human experience, and forming a unified and stable intelligent processing decision; 7. The intelligent agent can output differential processing credibility, background suppression strategy and observation parameter adjustment suggestions, forming an iteratively optimized auxiliary decision-making system, which greatly improves the detection adaptability to complex seabed environments.

[0011] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0012] Figure 1This is a flowchart of the transient electromagnetic feature extraction and differential background suppression decision method for high-conductivity targets on the seabed according to the present invention.

[0013] Figure 2 This is a schematic diagram of the underwater unexploded ordnance detection scenario along the center loop as described in the embodiment. Figure 3 This is a schematic diagram of the equivalent model of the orthogonal magnetic dipole described in the embodiment; Figure 4 This is a schematic diagram illustrating the burial depth attenuation relationship described in the embodiment. Figure 5 This is a schematic diagram illustrating the principle of symmetrical measurement point differential post-processing as described in the embodiment. Figure 6 This is a schematic diagram illustrating the variation of differential residual and background suppression rate with lateral non-uniform intensity as described in the embodiment. Figure 7 This is a schematic diagram of the intelligent analysis agent fusion decision-making process described in the embodiment. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are merely illustrative of the embodiments of the present invention and are not intended to limit the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of this application. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout.

[0015] It should be noted that the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, such as a process, method, system, product, or server that includes a series of steps or units, not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such process, method, product, or device.

[0016] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0017] like Figure 1 As shown, the transient electromagnetic feature extraction and differential background suppression decision method for high-conductivity targets on the seabed includes the following steps: S1. Set the seabed exploration environment parameters, high-conductivity target parameters, and center loop transient electromagnetic observation parameters, and determine the exploration background consisting of the seawater layer and sediment layer.

[0018] S2. Based on the transient electromagnetic observation parameters of the center loop determined in S1, the transmitting coil and receiving coil deployed at the same point are moved synchronously using the point-by-point scanning method of the center loop to obtain the transient electromagnetic total field response data of each measuring point in the measuring line or measuring area.

[0019] S3. Based on the transient electromagnetic observation parameters of the center loop determined in S1, perform the primary field calculation of the circular loop. Combined with the orthogonal magnetic dipole equivalent model, decompose the high-conductivity target on the seabed into three mutually orthogonal magnetic moment components, and calculate the secondary field response data of the high-conductivity target on the seabed.

[0020] S4. Based on the secondary field response data of the high-conductivity target on the seabed obtained in S3, extract the peak distribution pattern, principal axis direction and component amplitude characteristics of the response of the high-conductivity target on the seabed under different tilt and rotation angles to form a target attitude response feature set.

[0021] S5. Based on the secondary field response data of the high-conductivity target on the seabed obtained in S3, extract the maximum amplitude of the target response under different burial depth conditions and complete the normalization process, establish the empirical exponential decay relationship between the peak value of the target response and the burial depth, and determine the burial depth attenuation coefficient.

[0022] S6. Based on the transient electromagnetic total field response data obtained in S2, select measurement points that are symmetrical about the center of the survey line or the target projection position, perform differential post-processing of symmetrical measurement points, and obtain differential processed response data.

[0023] S7. Based on the seabed exploration environment parameters determined in S1, a simplified background field model is constructed, and the background lateral non-uniform intensity is quantified. Combined with the differential processing response data obtained in S6, the maximum amplitude of the background before differential processing, the maximum amplitude of the residual background after differential processing, and the maximum amplitude of the target anomaly after differential processing are determined.

[0024] S8. Based on the various amplitude parameters obtained in S7, calculate the residual ratio, background suppression rate and target anomaly contrast, quantitatively evaluate the applicability of differential post-processing for symmetrical measurement points, and form a set of differential applicability evaluation indicators.

[0025] S9. Construct an intelligent analysis agent for seabed transient electromagnetic detection. The agent integrates the target attitude response feature set obtained in S4, the burial depth attenuation coefficient obtained in S5, the differential applicability evaluation index set obtained in S8, the transient electromagnetic observation parameters of the center loop determined in S1, and the noise level of the detection environment to generate a unified fused feature vector.

[0026] S10, based on the fused feature vector obtained from S9, uses an intelligent analysis agent to perform rule-based reasoning, machine learning discrimination, and credibility assessment, outputting the attitude discrimination result of the high-conductivity target on the seabed, burial depth level, target anomaly visibility level, differential processing credibility, background suppression strategy, and observation parameter adjustment suggestions.

[0027] In this embodiment, after step S10, step S11 is also included: updating and iterating the intelligent analysis agent: for newly added simulated samples, experimental samples or field observation samples, the corresponding response features, burial depth attenuation features and differential evaluation indicators are extracted, and the interpretation results after manual verification or field verification are used as sample labels to supplement the sample library; the knowledge rule threshold is updated or the machine learning discrimination model is retrained according to the sample library, thereby improving the adaptability of the intelligent analysis agent under different seawater depths, sediment layer electrical properties and background non-uniformity conditions.

[0028] In step S1, the electrical and geometric parameters of the high-conductivity seabed target to be detected are defined, and the operating parameters and deployment rules of the center loop observation system are clarified.

[0029] In step S2, transient electromagnetic response values ​​are acquired at each measuring point using a uniform time channel. The measured transient electromagnetic total field response is decomposed into two parts: background response and target anomaly response. The expression for the transient electromagnetic total field response is as follows: ; In the formula, Horizontal coordinate place, time The transient electromagnetic total field response; This represents the background response generated jointly by the seawater layer and the sedimentary layer. This is an abnormal response generated by sensing high-conductivity targets on the seabed.

[0030] Step S3 specifically includes the following steps: S31. Based on the center loop observation parameters determined in step S1, and according to the principle of transient electromagnetic excitation of the center loop, calculate the primary magnetic induction intensity generated by the circular transmitting coil at the target position, and obtain the primary field distribution at the target position: ; In the formula, The magnetic induction intensity at the target location; Permeability of free space; This refers to the number of turns of the transmitting coil; For transmitting current; The radius of the transmitting coil; The horizontal radial distance of the target position relative to the center of the transmitting coil; The vertical burial depth of the target location relative to the plane of the transmitting coil, and , The vertical depth of the seawater layer. This refers to the burial depth within the sedimentary layer.

[0031] S32. Construct an equivalent model of an orthogonal magnetic dipole to decompose the high-conductivity target on the seabed into three mutually orthogonal magnetic moment components.

[0032] S321. Based on the geometric and electrical parameters of the high-conductivity target determined in step S1 and the results of the first-order field calculation in step S31, the high-conductivity target on the seabed is equivalent to an orthogonal magnetic dipole model.

[0033] S322. Calculate the equivalent volume of a high-conductivity target on the seabed. The initial induced magnetic moment of the target body is determined by combining the primary field strength. : ; ; In the formula, The equivalent diameter of the high-conductivity target on the seabed; The length of the high-conductivity target on the seabed; The conductivity of the target high conductivity on the seabed.

[0034] S323. Combining the target attitude angle with the initial magnetic moment, a coordinate transformation is performed to decompose it into three mutually orthogonal induced magnetic moment components: X, Y, and Z. Calculation formulas are then obtained. ; In the formula, , , These are the induced magnetic moment components of the target in the X, Y, and Z directions in the observation coordinate system, respectively. The tilt angle of the high-conductivity target on the seabed; The rotation angle of the high-conductivity target on the seabed; , , These are the X, Y, and Z components of the initial induced magnetic moment in the target body coordinate system, respectively.

[0035] S33. By coupling the primary field excitation with the equivalent model of the orthogonal magnetic dipole, the secondary field response data of the high-conductivity target on the seabed is obtained.

[0036] S331. Based on the orthogonal magnetic moment components from step S323, calculate the secondary field components generated by the magnetic dipoles in each direction at the observation point: ; In the formula, For the target secondary field X, Y, or Z component; To observe the induced magnetic moment components in the X, Y, or Z directions in the coordinate system; The unit position vector of the observation point relative to the center of the target; This is the straight-line distance from the observation point to the center of the target.

[0037] S332. Synthesize the three-dimensional quadratic field components to obtain the complete quadratic field response data of the target: ; In the formula, The horizontal coordinate of the observation point ,time The target total secondary field response; , , These are the X, Y, and Z components of the target secondary field, respectively.

[0038] The empirical exponential decay relationship expression mentioned in step S5 is as follows: ; In the formula, Burial depth within the sedimentary layer Normalized maximum response amplitude; For the target secondary field response Maximum absolute amplitude; For reference burial depth The corresponding target secondary field response maximum absolute amplitude; The burial depth attenuation coefficient is to be solved.

[0039] The empirical exponential decay relationship was solved using a numerical fitting method to obtain the burial depth decay coefficient. .

[0040] Step S6 specifically includes the following steps: S61. Based on the transient electromagnetic total field response data obtained in step S2, determine the core reference parameters for the symmetrical measurement point differential: extract the coordinates of the measurement line center or the projection position coordinates of the high-conductivity target on the seabed from step S2, and set them as the differential symmetry center, denoted as... Simultaneously, set the offset parameters for symmetrical measuring points and determine the symmetrical distance. , For symmetrical measuring points relative to the difference symmetry center The lateral offset distance.

[0041] S62, Based on the difference symmetry center symmetrical distance Select the response data of the symmetrical measuring point from the transient electromagnetic total field response data in step S2: select the measuring point to the right of the center of symmetry, with the horizontal coordinate of the measuring point being... Extract the transient total electromagnetic field response at the measuring point under a unified time channel, denoted as Simultaneously, a measuring point is selected to the left of the center of symmetry, with the horizontal coordinate of the measuring point being... Extract the transient total electromagnetic field response at the measuring point under a unified time channel, denoted as .

[0042] S63. Perform differential calculation on symmetrical measuring points to obtain the differential processing response of a single set of symmetrical measuring points: ; In the formula, Symmetrical distance ,time Differential processing response at symmetrical measurement points; S64. Traverse all effective symmetrical measurement point pairs along the entire survey line to form differential processing response data for the entire survey line.

[0043] Step S7 specifically includes the following steps: S71. Based on the seabed detection environment parameters determined in step S1, and according to the distribution law of the seabed transient electromagnetic detection background field, a simplified background field model is established, and the calculation formula is as follows: ; In the formula, Horizontal coordinate Background field response amplitude at that location; The background response amplitude; The scale of spatial variation of the background field; Background horizontal non-uniform intensity; The half-width of the survey area; This is for natural exponent calculations.

[0044] S72. Based on the actual conditions of the detected background, quantitatively determine the transverse non-uniform intensity of the background. And classify the background uniformity level. ; The quantized background horizontal non-uniform intensity Substitute the simplified background field model from step S71 to complete the model parameter calibration.

[0045] S73. Based on the transient electromagnetic total field response data from step S2, determine the maximum amplitude of the differential background. .

[0046] S731, From the electromagnetic total field response data of step S2 Separate the pure background response component .

[0047] S732. Traverse the background response data of the entire survey line, extract the maximum value as the maximum amplitude of the background before differential grading, and denot it as... .

[0048] S74. Based on the differential processing response data from step S6 and the simplified background field model from step S71, determine the maximum amplitude of the residual background after differential processing. .

[0049] S741. Substitute the simplified background field model calibrated in step S71 into the symmetric measurement point difference formula in step S6, and calculate the residual background response after difference. .

[0050] S742. Traverse the differential background residual response data of the entire survey line, extract the maximum value as the maximum amplitude of the differential background residual, denoted as... .

[0051] S75. Based on the differential processing response data from step S6 and the target secondary field response data from step S3, determine the maximum amplitude of the target anomaly after differential processing. .

[0052] S751, From the differential processing response data of step S6 Separate the pure target anomaly response component from .

[0053] S752. Traverse the differential target anomaly response data across the entire survey line, extract the maximum value as the maximum amplitude of the differential target anomaly, denoted as... .

[0054] In step S8, the residual ratio Background compression rate Contrast with target anomalies The calculation formula is as follows: ; ; .

[0055] The intelligent analysis agent described in step S9 includes a data access module, a feature construction module, a feature normalization module, a knowledge rule module, a machine learning discrimination module, a credibility assessment module, and a strategy generation module. The data access module receives transient electromagnetic response data of the center loop, target equivalent response calculation results, burial depth attenuation calculation results, and differential applicability evaluation results. The feature construction module extracts peak quantity, peak sign, response principal axis direction, component amplitude ratio, normalized maximum response amplitude, burial depth attenuation coefficient, residual ratio, background suppression rate, target anomaly contrast, and background lateral non-uniformity intensity. The feature normalization module eliminates dimensional differences and forms a unified fused feature vector. The knowledge rule module stores the judgment rules established based on transient electromagnetic response laws and engineering experience. The system comprises several modules: a rule-based inference module; a machine learning discrimination module; a target attitude category, burial depth level, differential applicability level, and target anomaly visibility level; a rule-based inference model, decision tree model, random forest model, support vector machine model, neural network model, large language model-assisted inference model, or a combination thereof; and a training sample for the machine learning discrimination module consisting of simulated or measured data under different target attitudes, burial depths, target electrical parameters, seawater depths, background lateral non-uniformity intensity, and noise levels. The sample labels include target attitude category, burial depth level, background applicability level, and target anomaly visibility level. A credibility assessment module is used to evaluate the reliability of the target interpretation results. A strategy generation module is used to provide suggestions for background suppression, survey line adjustment, time channel selection, or retesting.

[0056] And the feature vectors are fused , These represent the peak structure characteristics of the X, Y, and Z components, respectively. Indicates the response characteristics along the principal axis direction. This represents the normalized maximum response amplitude. This indicates the observed noise level or signal-to-noise ratio related parameters.

[0057] The specific execution logic of step S10 is as follows: When the target has abnormal contrast Greater than the preset contrast threshold and residual ratio When the residual value is less than the preset residual threshold, the intelligent analysis agent outputs results with high visibility of target anomalies and high reliability of symmetrical measurement point differential post-processing.

[0058] When the target has abnormal contrast Less than the preset contrast threshold or residual ratio When the residual value exceeds the preset residual threshold, the intelligent analysis agent outputs results with a strong background residual influence and provides suggestions such as reducing the weight of the differential results, using background modeling correction, adjusting the survey line layout, changing the time channel, or adding retest points.

[0059] It should be noted that the method described in this invention can be used for interpreting the transient electromagnetic response characteristics of underwater unexploded ordnance, underwater high-conductivity metal targets, seabed pipelines or other small-scale high-conductivity targets, evaluating target visibility, selecting background suppression processing parameters, and intelligent auxiliary decision-making.

[0060] Example This embodiment uses an underwater unexploded ordnance target as an example for illustration.

[0061] Table 1 Observation System and Target Parameters

[0062] First, according to... Figure 2 The observation system shown performs a point-by-point scan of the center loop to obtain the response at each measuring point. Then press as follows Figure 3 As shown, the unexploded ordnance target is equivalent to three mutually orthogonal magnetic dipole components, and the target tilt angle is changed. Slew angle The structure of the secondary field response was analyzed. The results show that the tilt angle mainly affects the peak distribution pattern of the response, the rotation angle mainly controls the direction of the principal axis of the response, and the vertical component has a strong indicative effect on attitude changes.

[0063] Secondly, the maximum amplitude of the response under different burial depths was extracted and normalized, and the results are shown in Table 2.

[0064] Table 2. Statistics of normalized maximum response amplitude under different burial depths

[0065] As shown in Table 2, when the burial depth increases from 0.5m to 2.0m, the normalized maximum response amplitude decreases from 1.00 to approximately 0.10, a decrease of about one order of magnitude. Substituting into... Figure 4 The empirical exponential decay relationship shown can be obtained This indicates that, under the observation parameters of this embodiment, for every 1m increase in burial depth, the peak response decreases to approximately 21% of its original value.

[0066] Finally, based on, as Figure 5 The symmetrical measurement point differential approach shown here performs post-processing on the data after completing the point-by-point scanning, and utilizes methods such as... Figure 6 The indicators shown evaluate the applicability of the differential method. Lateral non-uniform intensity is set. The values ​​of 0, 0.1, 0.2, and 0.3 correspond to horizontal layered, weakly non-uniform, moderately non-uniform, and strongly non-uniform backgrounds, respectively. The quantitative evaluation results are shown in Table 3.

[0067] Table 3. Evaluation results of the applicability of differential methods under different background non-uniform intensity.

[0068] Table 3 shows that under a horizontally layered background, the residual ratio is close to 0, and the background suppression rate is close to 100%, indicating that the symmetrical measurement point difference can effectively weaken the smooth common-mode background; with As the ratio increases, the residual ratio gradually rises, while the background suppression rate and target anomaly contrast decrease. When When the target anomaly contrast drops to 0.72, it indicates that the background residue may exceed the target differential anomaly, and the reliability of differential post-processing decreases significantly at this point.

[0069] After obtaining the quantitative results shown in Tables 2 and 3, the target response peak structure, normalized maximum response amplitude, burial depth attenuation coefficient, residual ratio, background suppression rate, and target anomaly contrast are input as follows: Figure 7 The intelligent analysis agent shown first determines the target's attitude change characteristics based on the peak distribution pattern and the principal axis direction of the response, and then determines them based on Anorm and Determine the target burial depth level, and finally based on , and Determine the reliability of the differential background suppression results.

[0070] For example, when hour, , , The agent determines that the background has weak lateral non-uniformity, the target anomaly visibility is high, and the post-processing results of symmetrical measurement point difference have high reliability; when hour, , , The agent determined that background residue significantly impacts target anomaly identification and recommended reducing the weight of differential results in target interpretation, while combining background modeling correction or adding retest points for comprehensive judgment. Thus, the agent can transform the aforementioned physical response characteristics and differential applicability indicators into intelligent interpretation results and processing suggestions tailored to practical detection tasks.

[0071] The above embodiments demonstrate that the present invention can extract the secondary field response characteristics of underwater unexploded ordnance, quantitatively describe the attenuation effect of burial depth on the response amplitude, and evaluate the applicability of the symmetrical measuring point differential background suppression method under different lateral non-uniform backgrounds, thus proving the effectiveness of the present invention.

[0072] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for extracting transient electromagnetic features of highly conductive seabed targets and for differential background suppression decision-making, characterized in that: Includes the following steps: S1. Set the seabed exploration environment parameters, high-conductivity target parameters, and transient electromagnetic observation parameters of the center loop, and determine the exploration background consisting of the seawater layer and sediment layer; S2. Based on the transient electromagnetic observation parameters of the center loop determined in S1, the transmitting coil and receiving coil deployed at the same point are moved synchronously using the point-by-point scanning method of the center loop to obtain the transient electromagnetic total field response data of each measuring point in the measuring line or measuring area. S3. Based on the transient electromagnetic observation parameters of the center loop determined in S1, perform the primary field calculation of the circular loop. Combined with the orthogonal magnetic dipole equivalent model, decompose the high-conductivity target on the seabed into three mutually orthogonal magnetic moment components, and calculate the secondary field response data of the high-conductivity target on the seabed. S4. Based on the secondary field response data of the high-conductivity target on the seabed obtained in S3, extract the peak distribution pattern, principal axis direction and component amplitude characteristics of the response of the high-conductivity target on the seabed under different tilt and rotation angles to form a target attitude response feature set. S5. Based on the secondary field response data of the high-conductivity target on the seabed obtained in S3, extract the maximum amplitude of the target response under different burial depth conditions and complete the normalization process, establish the empirical exponential decay relationship between the peak value of the target response and the burial depth, and determine the burial depth attenuation coefficient. S6. Based on the transient electromagnetic total field response data obtained in S2, select measurement points that are symmetrical about the center of the measurement line or the target projection position, perform differential post-processing of symmetrical measurement points, and obtain differential processed response data. S7. Based on the seabed exploration environment parameters determined in S1, a simplified background field model is constructed, and the background lateral non-uniform intensity is quantified. Combined with the differential processing response data obtained in S6, the maximum amplitude of the background before differential processing, the maximum amplitude of the residual background after differential processing, and the maximum amplitude of the target anomaly after differential processing are determined. S8. Based on the various amplitude parameters obtained in S7, calculate the residual ratio, background suppression rate and target anomaly contrast, quantitatively evaluate the applicability of differential post-processing for symmetrical measurement points, and form a set of differential applicability evaluation indicators. S9. Construct an intelligent analysis agent for seabed transient electromagnetic detection. The target attitude response feature set obtained in S4, the burial depth attenuation coefficient obtained in S5, the differential applicability evaluation index set obtained in S8, the center loop transient electromagnetic observation parameters determined in S1, and the detection environment noise level are fused from multiple sources to generate a unified fused feature vector. S10, based on the fused feature vector obtained from S9, uses an intelligent analysis agent to perform rule-based reasoning, machine learning discrimination, and credibility assessment, outputting the attitude discrimination result of the high-conductivity target on the seabed, burial depth level, target anomaly visibility level, differential processing credibility, background suppression strategy, and observation parameter adjustment suggestions.

2. The method for extracting transient electromagnetic features of high-conductivity seabed targets and making differential background suppression decisions according to claim 1, characterized in that: In step S1, the electrical and geometric parameters of the high-conductivity seabed target to be detected are defined, and the operating parameters and deployment rules of the center loop observation system are clarified.

3. The method for extracting transient electromagnetic features of high-conductivity seabed targets and making differential background suppression decisions according to claim 1, characterized in that: In step S2, transient electromagnetic response values ​​are acquired at each measuring point using a uniform time channel. The measured transient electromagnetic total field response is decomposed into two parts: background response and target anomaly response. The expression for the transient electromagnetic total field response is as follows: ； In the formula, Horizontal coordinate place, time The transient electromagnetic total field response; This represents the background response generated jointly by the seawater layer and the sedimentary layer. This is an abnormal response generated by sensing high-conductivity targets on the seabed.

4. The method for extracting transient electromagnetic features of high-conductivity seabed targets and making differential background suppression decisions according to claim 3, characterized in that: Step S3 specifically includes the following steps: S31. Based on the center loop observation parameters determined in step S1, and according to the principle of transient electromagnetic excitation of the center loop, calculate the primary magnetic induction intensity generated by the circular transmitting coil at the target position, and obtain the primary field distribution at the target position: ； In the formula, The magnetic induction intensity at the target location; Permeability of free space; This refers to the number of turns of the transmitting coil; For transmitting current; The radius of the transmitting coil; The horizontal radial distance of the target position relative to the center of the transmitting coil; The vertical burial depth of the target location relative to the plane of the transmitting coil, and , The vertical depth of the seawater layer. The depth within the sedimentary layer; S32. Construct an equivalent model of orthogonal magnetic dipoles to decompose the high-conductivity seabed target into three mutually orthogonal magnetic moment components. S321. Based on the geometric and electrical parameters of the high-conductivity target determined in step S1 and the results of the first field calculation in step S31, the high-conductivity target on the seabed is equivalent to an orthogonal magnetic dipole model. S322. Calculate the equivalent volume of a high-conductivity target on the seabed. The initial induced magnetic moment of the target body is determined by combining the primary field strength. : ； ； In the formula, The equivalent diameter of the high-conductivity target on the seabed; The length of the high-conductivity target on the seabed; The conductivity of the target high conductivity on the seabed; S323. Combining the target attitude angle with the initial magnetic moment, a coordinate transformation is performed to decompose it into three mutually orthogonal induced magnetic moment components: X, Y, and Z. Calculation formulas are then obtained. ； In the formula, , , These are the induced magnetic moment components of the target in the X, Y, and Z directions in the observation coordinate system, respectively. The tilt angle of the high-conductivity target on the seabed; The rotation angle of the high-conductivity target on the seabed; , , These are the X, Y, and Z components of the initial induced magnetic moment in the target body coordinate system, respectively. S33. Couple the primary field excitation with the equivalent model of the orthogonal magnetic dipole to obtain the secondary field response data of the high-conductivity target on the seabed. S331. Based on the orthogonal magnetic moment components from step S323, calculate the secondary field components generated by the magnetic dipoles in each direction at the observation point: ； In the formula, For the target secondary field X, Y, or Z component; To observe the induced magnetic moment components in the X, Y, or Z directions in the coordinate system; The unit position vector of the observation point relative to the center of the target; The straight-line distance from the observation point to the center of the target; S332. Synthesize the three-dimensional quadratic field components to obtain the complete quadratic field response data of the target: ； In the formula, The horizontal coordinate of the observation point ,time The target total secondary field response; , , These are the X, Y, and Z components of the target secondary field, respectively.

5. The method for extracting transient electromagnetic features of high-conductivity seabed targets and making differential background suppression decisions according to claim 4, characterized in that: The empirical exponential decay relationship expression mentioned in step S5 is as follows: ； In the formula, Burial depth within the sedimentary layer Normalized maximum response amplitude; For the target secondary field response Maximum absolute amplitude; For reference burial depth The corresponding target secondary field response maximum absolute amplitude; The burial depth attenuation coefficient is to be solved. The empirical exponential decay relationship was solved using a numerical fitting method to obtain the burial depth decay coefficient. .

6. The method for extracting transient electromagnetic features of high-conductivity seabed targets and making differential background suppression decisions according to claim 5, characterized in that: Step S6 specifically includes the following steps: S61. Based on the transient electromagnetic total field response data obtained in step S2, determine the core reference parameters for the symmetrical measurement point differential: extract the coordinates of the measurement line center or the projection position coordinates of the high-conductivity target on the seabed from step S2, and set them as the differential symmetry center, denoted as... Simultaneously, set the offset parameters for symmetrical measuring points and determine the symmetrical distance. , For symmetrical measuring points relative to the difference symmetry center The lateral offset distance; S62, Based on the difference symmetry center symmetrical distance Select the response data of the symmetrical measuring point from the transient electromagnetic total field response data in step S2: select the measuring point to the right of the center of symmetry, with the horizontal coordinate of the measuring point being... Extract the transient total electromagnetic field response at the measuring point under a unified time channel, denoted as Simultaneously, a measuring point is selected to the left of the center of symmetry, with the horizontal coordinate of the measuring point being... Extract the transient total electromagnetic field response at the measuring point under a unified time channel, denoted as ; S63. Perform differential calculation on symmetrical measuring points to obtain the differential processing response of a single set of symmetrical measuring points: ； In the formula, Symmetrical distance ,time Differential processing response at symmetrical measurement points; S64. Traverse all effective symmetrical measurement point pairs along the entire survey line to form differential processing response data for the entire survey line.

7. The method for extracting transient electromagnetic features and differential background suppression of high-conductivity targets on the seabed according to claim 6, characterized in that: Step S7 specifically includes the following steps: S71. Based on the seabed detection environment parameters determined in step S1, and according to the distribution law of the seabed transient electromagnetic detection background field, a simplified background field model is established, and the calculation formula is as follows: ； In the formula, Horizontal coordinate Background field response amplitude at that location; The background response amplitude; The scale of spatial variation of the background field; Background horizontal non-uniform intensity; The half-width of the survey area; For natural index calculations; S72. Based on the actual conditions of the detected background, quantitatively determine the transverse non-uniform intensity of the background. And classify the background uniformity level. ; The quantized background horizontal non-uniform intensity Substitute the simplified background field model from step S71 to complete the model parameter calibration; S73. Based on the transient electromagnetic total field response data from step S2, determine the maximum amplitude of the differential background. ; S731, From the electromagnetic total field response data of step S2 Separate the pure background response component ; S732. Traverse the background response data of the entire survey line, extract the maximum value as the maximum amplitude of the background before differential grading, and denot it as... ; S74. Based on the differential processing response data from step S6 and the simplified background field model from step S71, determine the maximum amplitude of the residual background after differential processing. ; S741. Substitute the simplified background field model calibrated in step S71 into the symmetric measurement point difference formula in step S6, and calculate the residual background response after difference. ; S742. Traverse the differential background residual response data of the entire survey line, extract the maximum value as the maximum amplitude of the differential background residual, denoted as... ; S75. Based on the differential processing response data from step S6 and the target secondary field response data from step S3, determine the maximum amplitude of the target anomaly after differential processing. ; S751, From the differential processing response data of step S6 Separate the pure target anomaly response component from ; S752. Traverse the differential target anomaly response data across the entire survey line, extract the maximum value as the maximum amplitude of the differential target anomaly, denoted as... .

8. The method for extracting transient electromagnetic features of high-conductivity seabed targets and making differential background suppression decisions according to claim 7, characterized in that: In step S8, the residual ratio Background compression rate Contrast with target anomalies The calculation formula is as follows: ； ； 。 9. The method for extracting transient electromagnetic features and differential background suppression of high-conductivity targets on the seabed according to claim 8, characterized in that: The intelligent analysis agent described in step S9 includes a data access module, a feature construction module, a feature normalization module, a knowledge rule module, a machine learning discrimination module, a credibility assessment module, and a strategy generation module; The data access module is used to receive transient electromagnetic response data of the center loop, calculation results of the target equivalent response, calculation results of burial depth attenuation, and differential applicability evaluation results. The feature construction module is used to extract the number of peaks, peak sign, response principal axis direction, component amplitude ratio, normalized maximum response amplitude, burial depth attenuation coefficient, residual ratio, background suppression rate, target anomaly contrast, and background lateral non-uniformity intensity. The feature normalization module is used to eliminate dimensional differences and form a unified fused feature vector; The knowledge rules module is used to store judgment rules based on transient electromagnetic response laws and engineering experience; The machine learning discrimination module outputs the target attitude category, burial depth level, differential applicability level, and target anomaly visibility level. The machine learning discrimination module employs rule-based reasoning models, decision tree models, random forest models, support vector machine models, neural network models, large language models to assist in reasoning, or combinations thereof. Furthermore, the training samples for the machine learning discrimination module consist of simulated or measured data under different target attitudes, burial depths, target electrical parameters, seawater depths, background lateral non-uniformity intensity, and noise levels. The sample labels include the target attitude category, burial depth level, background applicability level, and target anomaly visibility level. The credibility assessment module is used to evaluate the reliability of the target interpretation results; The strategy generation module is used to provide suggestions on background suppression, measurement line adjustment, time channel selection, or retesting. And the feature vectors are fused , These represent the peak structure characteristics of the X, Y, and Z components, respectively. Indicates the response characteristics along the principal axis direction. This represents the normalized maximum response amplitude. This indicates the observed noise level or signal-to-noise ratio related parameters.

10. The method for extracting transient electromagnetic features of high-conductivity seabed targets and making differential background suppression decisions according to claim 9, characterized in that: The specific execution logic of step S10 is as follows: When the target has abnormal contrast Greater than the preset contrast threshold and residual ratio When the residual value is less than the preset residual threshold, the intelligent analysis agent outputs results with high visibility of target anomalies and high reliability of symmetrical measurement point differential post-processing. When the target has abnormal contrast Less than the preset contrast threshold or residual ratio When the residual value exceeds the preset residual threshold, the intelligent analysis agent outputs results with a strong background residual influence and provides suggestions such as reducing the weight of the differential results, using background modeling correction, adjusting the survey line layout, changing the time channel, or adding retest points.

Images