A magnetic field observation method and system suitable for high-density deployment

Abstract

The present invention provides a magnetic field observation method and system suitable for high-density deployment, and relates to the field of data processing technology. The method comprises: obtaining observation planning data, determining the number, interval and position distribution of observation nodes, setting multiple observation nodes to designated positions in a target area, collecting magnetic field data in the target area through the observation nodes, preliminarily screening the collected data according to background noise generated by equipment characteristics, performing secondary screening according to spatial distribution characteristics between observation nodes, aggregating observation data of different observation nodes according to spatial position association rules, cross-validating data consistency between observation nodes in the target area, optimizing through a precision improvement algorithm, obtaining a magnetic field distribution model, identifying magnetic field abnormal areas, performing multi-level refinement analysis on abnormal area data, and generating magnetic field distribution results and abnormal characteristic data of the target area. The present invention improves the magnetic field observation accuracy.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a magnetic field observation method and system suitable for high-density deployment. Background Art

[0002] Magnetic field observation technology has been widely used in geological exploration, space science and industrial applications. At present, existing technologies mostly use observation equipment based on vector magnetometers or scalar magnetometers to collect magnetic field data. A typical observation system deploys several observation nodes in a distributed manner, uses the magnetic sensor in each node to collect magnetic field data in the environment in real time, and transmits the data to the central processing unit through a communication module for centralized analysis. Such systems usually complete observations through fixed deployment points, but in high-density deployment scenarios, such as magnetic anomaly monitoring in complex geological structure areas, the balance between signal interference between nodes and data acquisition accuracy becomes a difficult problem to achieve.

[0003] However, existing technologies face the problem of difficulty in ensuring observation accuracy in high-density deployment. Specifically, in geological exploration scenarios, observation points usually need to be deployed in spaces with small adjacent intervals, and observation equipment is mostly fixed. When the distance between devices is too close, the magnetic field sensor may be interfered by the electromagnetic noise emitted by the surrounding observation equipment, resulting in deviations in the collected data. Such interference problems are particularly exacerbated under complex geological conditions, which may lead to misjudgment of the monitoring results of magnetic field anomalies, affecting subsequent analysis and decision-making. Summary of the invention

[0004] The purpose of the present invention is to provide a magnetic field observation method and system suitable for high-density deployment, aiming to solve the problems mentioned in the background technology.

[0005] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0006] In a first aspect, a magnetic field observation method suitable for high-density deployment is provided, the method comprising:

[0007] Acquiring observation planning data, wherein the observation planning data includes geographical boundaries of the target area, accuracy requirements of the magnetic field observation target, and deployment environment characteristics;

[0008] According to the observation planning data, determine the number, interval and location distribution of observation nodes and form an observation node layout plan;

[0009] According to the observation node deployment plan, multiple observation nodes are set to designated locations within the target area to form an observation node network;

[0010] According to the observation node network, magnetic field data in the target area is collected through the observation nodes to obtain collected data, wherein the collected data includes magnetic field intensity data and direction data;

[0011] Perform preliminary screening of the collected data based on the background noise generated by the characteristics of the equipment to generate preliminary screening data;

[0012] According to the spatial distribution characteristics between observation nodes, identify and mark abnormal data caused by mutual interference between nodes, perform secondary screening on the preliminary screening data, and generate observation data;

[0013] Aggregate the observation data of different observation nodes according to the spatial position association rules to generate a preliminary distribution data model of the regional magnetic field;

[0014] According to the preliminary distribution data model, the data consistency between the observation nodes in the target area is cross-validated to obtain the verification data model, and the verification data model is optimized by the precision improvement algorithm to obtain the magnetic field distribution model;

[0015] According to the magnetic field distribution model, the magnetic field abnormal area is identified, and the characteristic parameters of the abnormal area and its corresponding spatial position are recorded to obtain the abnormal area data;

[0016] The abnormal area data is subjected to multi-level refinement analysis to generate magnetic field distribution results and abnormal characteristic data of the target area. The magnetic field distribution results include a magnetic field distribution map in the target area and annotation data of the abnormal area.

[0017] Preferably, the collected data is preliminarily screened according to the background noise generated by the characteristics of the equipment to generate preliminary screening data, including:

[0018] Extracting equipment operation characteristics of each observation node, wherein the equipment operation characteristics include a frequency range and a fluctuation value of the noise;

[0019] Compare the collected data point by point according to the equipment operation characteristics, filter out the stable fluctuation mode data in the data, and match it with the noise characteristics corresponding to the equipment operation characteristics to obtain stable fluctuation data;

[0020] The stable fluctuation data in the collected data are eliminated step by step to obtain the preliminary screening data.

[0021] Preferably, according to the spatial distribution characteristics between the observation nodes, the abnormal data caused by the mutual interference between the nodes is identified and marked, and the preliminary screening data is screened again to generate the observation data, including:

[0022] Obtain the number, interval and location distribution of observation nodes;

[0023] By analyzing the data change trends of adjacent nodes, the abnormal fluctuations caused by the interference of adjacent nodes are identified and marked to obtain abnormal fluctuation data;

[0024] The abnormal fluctuation data is processed in segments, and the abnormal fluctuation part in the segment is removed to obtain the characteristic data of the target area;

[0025] Integrate the characteristic data of the target area to generate observation data.

[0026] Preferably, the observation data of different observation nodes are aggregated according to the spatial position association rule to generate a preliminary distribution data model of the regional magnetic field, including:

[0027] According to the deployment location of each observation node, the target area is divided into gridded spatial units, and each spatial unit corresponds to the observation data of one or more nodes;

[0028] Compare the observation data in each gridded spatial unit, identify and merge multiple observation data in the same unit, and obtain grid data;

[0029] By comparing the grid data trends between gridded spatial units, the data changes of adjacent units are smoothed to obtain smooth data;

[0030] The smoothed data are combined to generate a preliminary distribution data model of the target area, which reflects the overall characteristics of the magnetic field in the target area.

[0031] Preferably, according to the preliminary distribution data model, the data consistency between the observation nodes in the target area is cross-validated to obtain a verification data model, and the verification data model is optimized by a precision improvement algorithm to obtain a magnetic field distribution model, including:

[0032] Extract the associated range of the observation results of each observation node in the target area according to the preliminary distribution data model, and divide the data within the associated range into multiple verification unit data;

[0033] Compare the data from different observation nodes in each verification unit data, identify the difference values ​​and outliers, and record the difference values, outliers and source nodes;

[0034] According to the difference value, outliers and source nodes, the observation error caused by the position deviation between the observation nodes is adjusted, and the outlier data is smoothed to obtain the unit adjustment data;

[0035] Recombining unit adjustment data to generate a validation data model;

[0036] Based on the difference value and observation error of the observation nodes in the verification data model, the observation data of the nodes in the neighboring area are dynamically adjusted through the optimization of the global consistency error and used to generate the magnetic field distribution model; the optimization of the global consistency error is achieved through the following formula:

[0037] ,in, is the global consistency error loss function, For Node and nodes The observation error, , For Node and nodes The distance is the weight coefficient, is the number of nodes, For Node The set of neighbor nodes of is the regularization coefficient, For Node and nodes The difference value of , and Node and nodes Observation data.

[0038] Preferably, according to the magnetic field distribution model, the magnetic field abnormal area is identified, and the characteristic parameters of the abnormal area and its corresponding spatial position are recorded to obtain the abnormal area data, including:

[0039] According to the magnetic field distribution model, calculate the position The second-order gradient at , obtains the variation range data;

[0040] According to the variation range and local fluctuation characteristics, the significance score is calculated to measure the significance level of the magnetic field anomaly area. The calculation formula of the significance score of the anomaly area A is:

[0041] ,

[0042] in, is the significance score of abnormal region A, For location The second-order gradient of the magnetic field strength at , and is the magnetic field strength and The second-order partial derivative of For location ) of the magnetic field strength, is the average magnetic field strength in the target area, and is the integration variable;

[0043] According to the significance score, the area exceeding the preset significance score threshold is screened, the area boundary is marked and the magnetic field anomaly area is generated;

[0044] Calculate the change rate, local intensity peak and change direction of the magnetic field anomaly area, extract the characteristic data inside the area, and form the characteristic parameter data of the magnetic field anomaly area;

[0045] Associating each feature point in the feature parameter set with the spatial coordinates of the target area, recording the boundary information and center position of the magnetic field anomaly area, and generating spatial position data;

[0046] According to the characteristic parameter data and the spatial position data, abnormal region data is generated, and the abnormal region data is used to describe the specific distribution and characteristics of the magnetic field anomaly in the target region.

[0047] Preferably, a multi-level refinement analysis is performed on the abnormal area data to generate a magnetic field distribution result and abnormal characteristic data of the target area, wherein the magnetic field distribution result includes a magnetic field distribution map in the target area and annotated data of the abnormal area, including:

[0048] Obtain boundary information, center position and characteristic parameters in abnormal area data;

[0049] The abnormal area is divided into multiple sub-areas according to the boundary information, and the local change characteristics are calculated for each sub-area, and the intensity range, change direction and gradient value of the sub-area are extracted to generate the local change characteristics;

[0050] The change characteristics of each sub-region are hierarchically classified, and the local eigenvalues ​​are aggregated according to the multi-dimensional weights to generate multi-level refined features; the calculation formula of the eigenvalue is:

[0051] ,in, is the first The eigenvalues ​​of the layer aggregated features, is the first The characteristic values ​​of each feature, including intensity, direction, and rate of change, For the The weight coefficient of the feature, is the number of features;

[0052] By comparing the multi-level refined features with the surrounding non-abnormal area data, the correlation features inside and outside the area are analyzed, and the boundary of the abnormal area is optimized through the following dynamic adjustment formula to generate boundary data; the dynamic adjustment formula is:

[0053] ,in, is the optimized magnetic field strength, is the original magnetic field strength, is the threshold for boundary adjustment, where:

[0054] , is the maximum value of the magnetic field intensity in the abnormal area A, is the minimum value of the magnetic field intensity in the abnormal area A, is the mean of the second-order gradient in the abnormal area A, is the standard deviation of the second-order gradient in the abnormal area A, is the rate of change of magnetic field intensity in abnormal area A, is the peak value of the magnetic field intensity in the abnormal area A, is the difference in direction of magnetic field intensity in abnormal area A, , , is the direction of the magnetic field strength, is the magnetic field strength along The gradient of the direction, is the magnetic field strength along The gradient of the direction, is the inverse tangent function, , , , , , and is the weight coefficient, is the adjustment coefficient, is the area of ​​abnormal region A;

[0055] Integrate boundary data with the magnetic field distribution model to generate a magnetic field distribution map of the target area;

[0056] Mark the location, characteristic parameters and boundary changes of the abnormal area on the magnetic field distribution map to form a distribution result map;

[0057] According to the abnormal area data in the distribution result map, abnormal characteristic data is generated, including the boundary range, center position, intensity change and directional characteristics of each abnormal area.

[0058] In a second aspect, a magnetic field observation system suitable for high-density deployment is provided, the system comprising:

[0059] An observation planning module is used to obtain observation planning data, wherein the observation planning data includes the geographical boundaries of the target area, the accuracy requirements of the magnetic field observation target, and the deployment environment characteristics;

[0060] The layout plan module is used to determine the number, interval and location distribution of observation nodes according to the observation planning data, and form an observation node layout plan;

[0061] A node setting module is used to set multiple observation nodes to designated locations in the target area according to the observation node layout plan to form an observation node network;

[0062] A data acquisition module is used to acquire magnetic field data in a target area through an observation node network to obtain acquired data, wherein the acquired data includes magnetic field intensity data and direction data;

[0063] A preliminary screening module is used to perform preliminary screening on the collected data according to the background noise generated by the characteristics of the equipment and generate preliminary screening data;

[0064] The secondary screening module is used to identify and mark abnormal data caused by mutual interference between nodes according to the spatial distribution characteristics between observation nodes, perform secondary screening on the primary screening data, and generate observation data;

[0065] A model generation module is used to aggregate the observation data of different observation nodes according to the spatial position association rules to generate a preliminary distribution data model of the regional magnetic field;

[0066] The model optimization module is used to cross-validate the data consistency between the observation nodes in the target area according to the preliminary distribution data model to obtain the verification data model, and optimize the verification data model through the precision improvement algorithm to obtain the magnetic field distribution model;

[0067] The abnormality identification module is used to identify the abnormal magnetic field area according to the magnetic field distribution model, and record the characteristic parameters of the abnormal area and its corresponding spatial position to obtain the abnormal area data;

[0068] The abnormal analysis module is used to perform multi-level detailed analysis on the abnormal area data to generate the magnetic field distribution results and abnormal characteristic data of the target area. The magnetic field distribution results include the magnetic field distribution map in the target area and the annotation data of the abnormal area.

[0069] The above solution of the present invention includes at least the following beneficial effects:

[0070] Compared with the existing technology, the proposed method has significant advantages in high-density deployed magnetic field observation scenarios, especially in magnetic anomaly monitoring in complex geological structure areas, and effectively solves the accuracy and interference problems that are difficult to overcome by the existing technology.

[0071] First, by planning the geographic boundaries of the target area, the observation accuracy requirements, and the characteristics of the deployment environment, the number, spacing, and location distribution of observation nodes can be flexibly adjusted based on actual needs to form an optimized node deployment plan. This plan maximizes the coverage of the observation area while avoiding overly dense deployment between nodes, reducing the risk of signal interference caused by improper deployment.

[0072] Secondly, the accuracy of the observed data was greatly improved through multi-stage processing of the collected data. In the initial screening stage, the inherent interference of the node itself was eliminated by identifying and eliminating the background noise generated by the operating characteristics of the equipment; in the secondary screening stage, the abnormal fluctuations caused by node interference were accurately identified and eliminated by analyzing the spatial distribution characteristics and data change trends between nodes. The multi-level data screening strategy effectively reduced the noise in the observed data and laid a high-quality foundation for subsequent data analysis.

[0073] Thirdly, in the data processing stage, through multi-node data aggregation based on spatial position association rules, the observation data of different nodes are comprehensively analyzed to generate a preliminary distribution model reflecting the magnetic field characteristics of the target area. Subsequently, the accuracy improvement algorithm is used to dynamically optimize the data consistency and further generate a high-precision magnetic field distribution model. This adaptive optimization processing method can minimize the impact of node position deviation and signal interference in a complex interference environment, ensuring the accurate capture of magnetic field anomaly characteristics.

[0074] Finally, by analyzing and refining the abnormal areas of the magnetic field distribution model, the magnetic field abnormal areas can be efficiently identified and detailed characteristic data including boundary range, center position, intensity change and direction change can be generated. These data provide reliable support for magnetic anomaly monitoring under complex geological conditions and significantly improve the sensitivity and analysis accuracy of abnormal areas.

[0075] In summary, this method fully combines the characteristics of magnetic field observation technology. Through refined observation node deployment, multi-stage data processing and optimization, and in-depth analysis of abnormal areas, it solves the problems of serious interference between nodes and difficulty in ensuring observation accuracy in existing technologies. In high-density deployment scenarios, especially in magnetic anomaly monitoring in complex geological conditions, it has significant technical advantages and application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 It is a flowchart of a magnetic field observation method suitable for high-density deployment provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0077] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0078] like Figure 1 As shown, an embodiment of the present invention provides a magnetic field observation method suitable for high-density deployment, the method comprising:

[0079] Acquiring observation planning data, wherein the observation planning data includes geographical boundaries of the target area, accuracy requirements of the magnetic field observation target, and deployment environment characteristics;

[0080] According to the observation planning data, determine the number, interval and location distribution of observation nodes and form an observation node layout plan;

[0081] According to the observation node deployment plan, multiple observation nodes are set to designated locations within the target area to form an observation node network;

[0082] According to the observation node network, magnetic field data in the target area is collected through the observation nodes to obtain collected data, wherein the collected data includes magnetic field intensity data and direction data;

[0083] Perform preliminary screening of the collected data based on the background noise generated by the characteristics of the equipment to generate preliminary screening data;

[0084] According to the spatial distribution characteristics between observation nodes, identify and mark abnormal data caused by mutual interference between nodes, perform secondary screening on the preliminary screening data, and generate observation data;

[0085] Aggregate the observation data of different observation nodes according to the spatial position association rules to generate a preliminary distribution data model of the regional magnetic field;

[0086] According to the preliminary distribution data model, the data consistency between the observation nodes in the target area is cross-validated to obtain the verification data model, and the verification data model is optimized by the precision improvement algorithm to obtain the magnetic field distribution model;

[0087] According to the magnetic field distribution model, the magnetic field abnormal area is identified, and the characteristic parameters of the abnormal area and its corresponding spatial position are recorded to obtain the abnormal area data;

[0088] The abnormal area data is subjected to multi-level refinement analysis to generate magnetic field distribution results and abnormal characteristic data of the target area. The magnetic field distribution results include a magnetic field distribution map in the target area and annotation data of the abnormal area.

[0089] In an embodiment of the present invention, by acquiring observation planning data, determining the geographical boundaries of the target area, the accuracy requirements of the magnetic field observation, and the environmental characteristics, the layout plan of the observation nodes is optimized. On this basis, the magnetic field data is collected through multiple observation nodes to effectively cover the target area. In the preliminary screening stage, the purity of the original data is ensured by eliminating the background noise generated by the equipment characteristics. In this process, the interference data is secondary screened based on the spatial distribution characteristics between nodes to generate more accurate observation data. Subsequently, the data of different nodes are aggregated through spatial position association rules to generate a preliminary distribution model of the regional magnetic field. Further based on the model, the data consistency is cross-validated, and the verification model is optimized using the precision improvement algorithm, and finally a magnetic field distribution model is generated.

[0090] This method identifies significant magnetic field anomaly areas by comprehensively analyzing the magnetic field distribution model, and records the regional characteristic parameters and spatial location information. Through multi-level refinement analysis, results including magnetic field distribution maps and anomaly area annotation data are generated. Ultimately, accurate detection of magnetic field anomalies in high-density deployment scenarios is achieved, overcoming the problem of data distortion caused by node interference and background noise in traditional solutions.

[0091] In a preferred embodiment of the present invention, the collected data is preliminarily screened according to the background noise generated by the characteristics of the equipment to generate preliminary screening data, including:

[0092] Extracting equipment operation characteristics of each observation node, wherein the equipment operation characteristics include a frequency range and a fluctuation value of the noise;

[0093] Compare the collected data point by point according to the equipment operation characteristics, filter out the stable fluctuation mode data in the data, and match it with the noise characteristics corresponding to the equipment operation characteristics to obtain stable fluctuation data;

[0094] The stable fluctuation data in the collected data are eliminated step by step to obtain the preliminary screening data.

[0095] In the embodiment of the present invention, the equipment operation characteristics of the observation node, including the frequency range and fluctuation value of the noise, are extracted, and the collected data is compared point by point. This process identifies the stable fluctuation pattern data and matches it with the equipment operation noise characteristics, thereby screening out valid data. By using a step-by-step elimination method, the noise interference caused by the equipment characteristics is removed, and pure preliminary screening data is generated. Through this method, the data quality is effectively improved, ensuring the accuracy of subsequent magnetic field observation and analysis.

[0096] This method provides an adaptable screening mechanism for the operating environment of observation nodes. Especially in high-density deployment scenarios, the differences in operating characteristics between nodes are systematically quantified and eliminated, significantly improving the credibility of magnetic field acquisition data.

[0097] More specifically, the collected data is compared point by point according to the equipment operation characteristics, the stable fluctuation mode data in the data is screened out, and it is matched with the noise characteristics corresponding to the equipment operation characteristics to obtain the stable fluctuation data, which specifically includes:

[0098] The collected data is compared point by point according to the equipment operation characteristics. Point by point comparison means comparing the magnetic field data of each collection point with the pre-extracted equipment operation characteristics to identify and eliminate the noise data caused by the equipment's own characteristics. Equipment operation characteristics usually include the noise characteristics of the sensor in different frequency ranges and the fluctuation range generated by the equipment during operation.

[0099] Specifically: First, extract the operating characteristics of the equipment, such as the noise frequency range and fluctuation value when the equipment is in a stationary state or under normal operating conditions; then, compare each point in the collected data with the operating characteristics of the equipment one by one to determine whether the point meets the characteristics of the equipment noise.

[0100] Filter out the stable fluctuation mode data in the data. The stable fluctuation mode data refers to those data points that conform to the noise law of the equipment operation characteristics, usually manifested as fluctuations within a specific frequency range or amplitude. Through the screening process, the equipment noise can be distinguished from the actual magnetic field signal.

[0101] Specifically: on a point-by-point comparison basis, all data points that meet the equipment noise characteristics are extracted; these data points may exhibit stable frequency fluctuations or amplitude changes to form a data set.

[0102] It is matched with the noise characteristics corresponding to the equipment operation characteristics. The matching process is to establish a corresponding relationship between the screened stable fluctuation pattern and the equipment operation characteristics. The matching result is used to confirm that these fluctuation data are indeed generated by equipment noise, rather than the magnetic field signal itself.

[0103] Stable fluctuation data is obtained, and the noise data confirmed by matching is marked as stable fluctuation data. These data will be used as the target of noise removal and will eventually be removed from the collected data to provide pure preliminary screening data for subsequent data processing.

[0104] In a preferred embodiment of the present invention, according to the spatial distribution characteristics between the observation nodes, the abnormal data caused by the mutual interference between the nodes is identified and marked, and the preliminary screening data is screened twice to generate the observation data, including:

[0105] Obtain the number, interval and location distribution of observation nodes;

[0106] By analyzing the data change trends of adjacent nodes, the abnormal fluctuations caused by the interference of adjacent nodes are identified and marked to obtain abnormal fluctuation data;

[0107] The abnormal fluctuation data is processed in segments, and the abnormal fluctuation part in the segment is removed to obtain the characteristic data of the target area;

[0108] Integrate the characteristic data of the target area to generate observation data.

[0109] In the embodiment of the present invention, by analyzing the data change trend between adjacent nodes, the system identifies abnormal fluctuations caused by interference between nodes and marks such data. Subsequently, segmentation processing is performed on the marked abnormal fluctuation data, the abnormal part is removed, and characteristic data of the target area is generated. On this basis, the quality of the observation data is further improved by integrating the characteristic data.

[0110] This method effectively solves the problem of observation deviation caused by interference between nodes and ensures the accuracy of magnetic field data under high-density deployment. By eliminating and integrating abnormal fluctuations, the impact of noise on the overall observation results is significantly reduced, providing a high-quality data foundation for the subsequent generation of magnetic field distribution models.

[0111] More specifically, by analyzing the data change trend of adjacent nodes, the abnormal fluctuations caused by the interference of adjacent nodes are identified and marked to obtain abnormal fluctuation data, including:

[0112] The data change trend of adjacent nodes refers to the characteristics of the change in magnetic field strength and direction collected by different nodes within the same time range. These trends are usually manifested as consistency or difference in fluctuation amplitude, frequency and phase. By analyzing these trends, abnormal fluctuations can be identified.

[0113] If the data changes of some nodes deviate significantly from the normal correlation with other nodes, it can be judged that the data of these nodes are interfered. The interference may come from electromagnetic signals of other nodes or external noise.

[0114] For the identified abnormal fluctuations, the system will mark them as abnormal fluctuation data, which will be specially processed or eliminated in subsequent processing.

[0115] More specifically, the abnormal fluctuation data is processed in segments, and the abnormal fluctuation part in the segment is removed to obtain the characteristic data of the target area, which specifically includes:

[0116] Abnormal fluctuation data may be continuous or discontinuous, and segmentation is to divide these abnormal fluctuations into multiple independent segments. For example, by segmenting by time or space, the abnormal fluctuation part of the data can be separated.

[0117] In each segment, the parts with abnormal fluctuations are eliminated. For example, if the amplitude or frequency of a certain segment of data fluctuations exceeds the set normal range, these points are eliminated; only those data segments that conform to the normal fluctuation pattern are retained.

[0118] The remaining data set is the characteristic data of the target area, which eliminates the influence of abnormal fluctuations and retains the real magnetic field signal in the area.

[0119] More specifically, the characteristic data of the target area are integrated to generate observation data, including:

[0120] Integration refers to the comprehensive processing of all node data in the target area to generate overall observation data. The integration process includes spatial data fusion, combining the geographical location information of each node to map the data to the spatial grid of the target area; temporal data fusion, aligning and interpolating time series data to ensure that the data of all nodes can be synchronized.

[0121] Generate observation data, and the integrated data becomes observation data, which reflects the intensity and direction distribution of the magnetic field in the target area, and provides a high-quality data basis for the subsequent generation of magnetic field distribution models.

[0122] In a preferred embodiment of the present invention, the observation data of different observation nodes are aggregated according to the spatial position association rule to generate a preliminary distribution data model of the regional magnetic field, including:

[0123] According to the deployment location of each observation node, the target area is divided into gridded spatial units, and each spatial unit corresponds to the observation data of one or more nodes;

[0124] Compare the observation data in each gridded spatial unit, identify and merge multiple observation data in the same unit, and obtain grid data;

[0125] By comparing the grid data trends between gridded spatial units, the data changes of adjacent units are smoothed to obtain smooth data;

[0126] The smoothed data are combined to generate a preliminary distribution data model of the target area, which reflects the overall characteristics of the magnetic field in the target area.

[0127] In the embodiment of the present invention, the target area is divided into gridded spatial units by using spatial position association rules, and the observation data in each unit are compared and merged to generate gridded data. On this basis, the system further optimizes the data consistency between adjacent units by smoothing the trend of grid data. Finally, all smoothed data are combined to generate a preliminary distribution data model reflecting the overall characteristics of the magnetic field in the target area.

[0128] This method effectively improves the spatial resolution of data in the region by dividing the grid space units. By comparing and smoothing the trends between adjacent units, it avoids the problem of overall model distortion caused by local fluctuations in traditional methods, providing a reliable basis for subsequent data verification and distribution optimization.

[0129] More specifically, the target area is divided into gridded spatial units according to the deployment locations of each observation node.

[0130] Gridding spatial unit division is to divide the target area into several small grids, each of which contains the collected data of one or more observation nodes, including:

[0131] The division is based on the geographical location and distribution rules of the observation nodes. Through gridding, the data distribution of the target area can be processed more finely.

[0132] Get the geographic coordinate information of all observation nodes;

[0133] Determine the grid division scale based on the distance between nodes and the size of the target area;

[0134] Assign the nodes to the corresponding grids, and each grid cell corresponds to one or more nodes.

[0135] More specifically, the observation data within each gridded spatial unit are compared, and multiple observation data within the same unit are identified and merged, including:

[0136] The purpose of comparison is to analyze the consistency of observation data of each node in the grid. The merging process combines the data of multiple nodes in the grid into a representative value.

[0137] Compare the data in each grid and calculate the difference between the node data;

[0138] Eliminate abnormal data whose difference value exceeds the threshold;

[0139] The remaining data are combined by averaging or weighted averaging to generate gridded data.

[0140] More specifically, by comparing the grid data trends between gridded spatial cells, the data changes of adjacent cells are smoothed, including:

[0141] The purpose of smoothing is to eliminate the discontinuity of data between adjacent grids and ensure the coherence and consistency of the overall data.

[0142] Compare and analyze the data change trends of adjacent grids;

[0143] If the data of adjacent grids vary too much, the grid data is adjusted through interpolation algorithms or weighted smoothing methods;

[0144] Generate smoothed mesh data.

[0145] More specifically, the smoothed data are combined to generate a preliminary distribution data model for the target area, including:

[0146] Finally, the smoothed data of all grids are combined to form a preliminary data model reflecting the distribution of magnetic field intensity and direction in the target area.

[0147] Combine the smoothed grid data into complete data of the target area according to the geographical location;

[0148] The preliminary distribution model reflects the overall characteristics of the magnetic field in the target area and provides a basis for subsequent analysis.

[0149] In a preferred embodiment of the present invention, according to the preliminary distribution data model, the data consistency between the observation nodes in the target area is cross-validated to obtain a verification data model, and the verification data model is optimized by the precision improvement algorithm to obtain a magnetic field distribution model, including:

[0150] Extract the associated range of the observation results of each observation node in the target area according to the preliminary distribution data model, and divide the data within the associated range into multiple verification unit data;

[0151] Compare the data from different observation nodes in each verification unit data, identify the difference values ​​and outliers, and record the difference values, outliers and source nodes;

[0152] According to the difference value, outliers and source nodes, the observation error caused by the position deviation between the observation nodes is adjusted, and the outlier data is smoothed to obtain the unit adjustment data;

[0153] Recombining unit adjustment data to generate a validation data model;

[0154] Based on the difference value and observation error of the observation nodes in the verification data model, the observation data of the nodes in the neighboring area are dynamically adjusted through the optimization of the global consistency error and used to generate the magnetic field distribution model; the optimization of the global consistency error is achieved through the following formula:

[0155] ,in, is the global consistency error loss function, For Node and nodes The observation error, , For Node and nodes The distance is the weight coefficient, is the number of nodes, For Node The set of neighbor nodes of is the regularization coefficient, For Node and nodes The difference value of , and Node and nodes Observation data.

[0156] In an embodiment of the present invention, the data consistency between the observation nodes is cross-validated according to the preliminary distribution data model. By calculating the difference value and observation error between the nodes, the error data caused by the node position deviation is identified and adjusted, and the abnormal point data is smoothed. The system recombines the adjusted data to generate an optimized verification data model. On this basis, the observation data of the adjacent area is dynamically adjusted through the global consistency error optimization model, and a high-precision magnetic field distribution model is generated based on the optimized data.

[0157] This method realizes the dynamic adjustment of the consistency of node observation data, effectively overcoming the problem of inaccurate model caused by observation errors between nodes. Through global consistency optimization, a high-precision description of magnetic field distribution in complex scenarios is achieved.

[0158] More specifically, the unit adjustment data is recombined to generate a verification data model, including:

[0159] Get the adjusted observation data within each validation unit;

[0160] Regroup the unit data according to their spatial positions;

[0161] Check whether the boundary data between each unit is consistent to ensure there is no mutation;

[0162] If the boundary data are inconsistent, they are adjusted through smoothing or interpolation methods;

[0163] Finally, a verification data model is generated, which can reflect the consistency characteristics of the observation data in the target area and provide a basis for the subsequent optimization of the magnetic field distribution model.

[0164] In a preferred embodiment of the present invention, the magnetic field abnormal area is identified according to the magnetic field distribution model, and the characteristic parameters of the abnormal area and its corresponding spatial position are recorded to obtain the abnormal area data, including:

[0165] According to the magnetic field distribution model, calculate the position The second-order gradient at , obtains the variation range data;

[0166] According to the variation range and local fluctuation characteristics, the significance score is calculated to measure the significance level of the magnetic field anomaly area. The calculation formula of the significance score of the anomaly area A is:

[0167] ,

[0168] in, is the significance score of abnormal region A, For location The second-order gradient of the magnetic field strength at , and is the magnetic field strength and The second-order partial derivative of For location ) of the magnetic field strength, is the average magnetic field strength in the target area, and is the integration variable;

[0169] According to the significance score, the area exceeding the preset significance score threshold is screened, the area boundary is marked and the magnetic field anomaly area is generated;

[0170] Calculate the change rate, local intensity peak and change direction of the magnetic field anomaly area, extract the characteristic data inside the area, and form the characteristic parameter data of the magnetic field anomaly area;

[0171] Associating each feature point in the feature parameter set with the spatial coordinates of the target area, recording the boundary information and center position of the magnetic field anomaly area, and generating spatial position data;

[0172] According to the characteristic parameter data and the spatial position data, abnormal region data is generated, and the abnormal region data is used to describe the specific distribution and characteristics of the magnetic field anomaly in the target region.

[0173] In an embodiment of the present invention, the magnetic field distribution model is used to generate the variation range data by calculating the second-order gradient of the position in the target area. Combined with the local fluctuation characteristics, the significance score of each position is calculated, and the area with the significance score exceeding the threshold is screened out to form the boundary of the magnetic field anomaly area. The rate of change, intensity peak and direction inside the abnormal area are extracted to generate a detailed set of feature parameters. Subsequently, the feature points are associated with the spatial coordinates of the target area, the boundary information and the center position are recorded, and the complete abnormal area data is generated.

[0174] This method accurately locates the magnetic field anomaly area through significance score threshold screening and generates detailed data describing the distribution and characteristics of the anomaly area, effectively improving the detection sensitivity and boundary accuracy of the anomaly area.

[0175] More specifically, each feature point in the feature parameter set is associated with the spatial coordinates of the target area, the boundary information and center position of the magnetic field anomaly area are recorded, and spatial position data is generated, including:

[0176] Get the spatial coordinates of each feature point;

[0177] Bind characteristic data, such as intensity and direction, to spatial coordinates to form a complete data record;

[0178] Using these data, mark the boundaries and center locations of the abnormal areas;

[0179] Finally, spatial location data including boundary information and center position is generated.

[0180] In a preferred embodiment of the present invention, a multi-level refinement analysis is performed on the abnormal area data to generate a magnetic field distribution result and abnormal characteristic data of the target area, wherein the magnetic field distribution result includes a magnetic field distribution map in the target area and annotated data of the abnormal area, including:

[0181] Obtain boundary information, center position and characteristic parameters in abnormal area data;

[0182] The abnormal area is divided into multiple sub-areas according to the boundary information, and the local change characteristics are calculated for each sub-area, and the intensity range, change direction and gradient value of the sub-area are extracted to generate the local change characteristics;

[0183] The change characteristics of each sub-region are hierarchically classified, and the local eigenvalues ​​are aggregated according to the multi-dimensional weights to generate multi-level refined features; the calculation formula of the eigenvalue is:

[0184] ,in, is the first The eigenvalues ​​of the layer aggregated features, is the first The characteristic values ​​of each feature, including intensity, direction, and rate of change, For the The weight coefficient of the feature, is the number of features;

[0185] By comparing the multi-level refined features with the surrounding non-abnormal area data, the correlation features inside and outside the area are analyzed, and the boundary of the abnormal area is optimized through the following dynamic adjustment formula to generate boundary data; the dynamic adjustment formula is:

[0186] ,in, is the optimized magnetic field strength, is the original magnetic field strength, is the threshold for boundary adjustment, where:

[0187] , is the maximum value of the magnetic field intensity in the abnormal area A, is the minimum value of the magnetic field intensity in the abnormal area A, is the mean of the second-order gradient in the abnormal area A, is the standard deviation of the second-order gradient in the abnormal area A, is the rate of change of magnetic field intensity in abnormal area A, is the peak value of the magnetic field intensity in the abnormal area A, is the difference in direction of magnetic field intensity in abnormal area A, , , is the direction of the magnetic field strength, is the magnetic field strength along The gradient of the direction, is the magnetic field strength along The gradient of the direction, is the inverse tangent function, , , , , , and is the weight coefficient, is the adjustment coefficient, is the area of ​​abnormal region A;

[0188] Integrate boundary data with the magnetic field distribution model to generate a magnetic field distribution map of the target area;

[0189] Mark the location, characteristic parameters and boundary changes of the abnormal area on the magnetic field distribution map to form a distribution result map;

[0190] According to the abnormal area data in the distribution result map, abnormal characteristic data is generated, including the boundary range, center position, intensity change and directional characteristics of each abnormal area.

[0191] In an embodiment of the present invention, boundary information, center position and characteristic parameters are extracted from the abnormal region data, and the abnormal region is divided into multiple sub-regions according to the boundary information. Local change features are calculated for each sub-region, including intensity range, direction change and gradient value, to generate local change features. Sub-region features are aggregated through multi-dimensional weights to generate multi-level refined features. Subsequently, the system analyzes the correlation features inside and outside the region by comparing the multi-level refined features with the surrounding non-abnormal region data, and dynamically optimizes the boundary position.

[0192] The optimized boundary data is integrated with the magnetic field distribution model to generate a magnetic field distribution map of the target area, and the location, characteristic parameters and boundary changes of the abnormal area are marked on the distribution map. Finally, the characteristic data of the abnormal area is generated, including detailed information such as boundary range, center position, intensity change and direction change, providing accurate data support for subsequent analysis.

[0193] More specifically, the boundary data is integrated with the magnetic field distribution model to generate a magnetic field distribution map of the target area, including:

[0194] Superimpose the boundary data of the abnormal area onto the magnetic field distribution model of the target area;

[0195] Mark the location, extent and intensity of abnormal areas in the distribution model;

[0196] The magnetic field distribution map can intuitively display the magnetic field distribution characteristics and abnormal areas in the target area.

[0197] More specifically, the location, characteristic parameters and boundary changes of the abnormal area are marked on the magnetic field distribution map to form a distribution result map, which specifically includes:

[0198] Based on the magnetic field distribution map, detailed information of the abnormal area is marked, including:

[0199] The geographical location of the anomaly area;

[0200] Characteristic parameters of abnormal areas, such as intensity peak, rate of change, and direction;

[0201] Changes in the boundaries of the abnormal area, such as the direction and magnitude of expansion or contraction.

[0202] The distribution result map is the final visualization result of the magnetic field distribution, providing a clear basis for subsequent decision-making.

[0203] More specifically, based on the abnormal area data in the distribution result map, abnormal characteristic data is generated, including the boundary range, center position, intensity change and directional characteristics of each abnormal area, specifically including:

[0204] Extract the boundary range of each abnormal area in the distribution result map;

[0205] Mark the center of the abnormal area;

[0206] Analyze the intensity changes in abnormal areas, such as local peaks and gradient distribution;

[0207] Calculate directional characteristics, such as directional range and main direction.

[0208] An embodiment of the present invention further provides a magnetic field observation system suitable for high-density deployment, the system comprising:

[0209] An observation planning module is used to obtain observation planning data, wherein the observation planning data includes the geographical boundaries of the target area, the accuracy requirements of the magnetic field observation target, and the deployment environment characteristics;

[0210] The layout plan module is used to determine the number, interval and location distribution of observation nodes according to the observation planning data, and form an observation node layout plan;

[0211] A node setting module is used to set multiple observation nodes to designated locations in the target area according to the observation node layout plan to form an observation node network;

[0212] A data acquisition module is used to acquire magnetic field data in a target area through an observation node network to obtain acquired data, wherein the acquired data includes magnetic field intensity data and direction data;

[0213] A preliminary screening module is used to perform preliminary screening on the collected data according to the background noise generated by the characteristics of the equipment and generate preliminary screening data;

[0214] The secondary screening module is used to identify and mark abnormal data caused by mutual interference between nodes according to the spatial distribution characteristics between observation nodes, perform secondary screening on the primary screening data, and generate observation data;

[0215] A model generation module is used to aggregate the observation data of different observation nodes according to the spatial position association rules to generate a preliminary distribution data model of the regional magnetic field;

[0216] The model optimization module is used to cross-validate the data consistency between the observation nodes in the target area according to the preliminary distribution data model to obtain the verification data model, and optimize the verification data model through the precision improvement algorithm to obtain the magnetic field distribution model;

[0217] The abnormality identification module is used to identify the abnormal magnetic field area according to the magnetic field distribution model, and record the characteristic parameters of the abnormal area and its corresponding spatial position to obtain the abnormal area data;

[0218] The abnormal analysis module is used to perform multi-level detailed analysis on the abnormal area data to generate the magnetic field distribution results and abnormal characteristic data of the target area. The magnetic field distribution results include the magnetic field distribution map in the target area and the annotation data of the abnormal area.

[0219] It should be noted that the system is a system corresponding to the above method, and all implementation methods in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0220] The embodiment of the present invention further provides a computing device, including: a processor, a memory storing a computer program, and when the computer program is executed by the processor, the method described above is executed. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0221] The embodiment of the present invention further provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute the above method. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0222] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A magnetic field observation method suitable for high-density deployment, characterized in that: The method comprises: Acquiring observation planning data, wherein the observation planning data includes geographical boundaries of the target area, accuracy requirements of the magnetic field observation target, and deployment environment characteristics; According to the observation planning data, determine the number, interval and location distribution of observation nodes and form an observation node layout plan; According to the observation node deployment plan, multiple observation nodes are set to designated locations within the target area to form an observation node network; According to the observation node network, magnetic field data in the target area is collected through the observation nodes to obtain collected data, wherein the collected data includes magnetic field intensity data and direction data; Perform preliminary screening of the collected data based on the background noise generated by the characteristics of the equipment to generate preliminary screening data; According to the spatial distribution characteristics between observation nodes, abnormal data caused by mutual interference between nodes are identified and marked, and the preliminary screening data are screened again to generate observation data; Aggregate the observation data of different observation nodes according to the spatial position association rules to generate a preliminary distribution data model of the regional magnetic field; According to the preliminary distribution data model, the data consistency between the observation nodes in the target area is cross-validated to obtain the verification data model, and the verification data model is optimized by the precision improvement algorithm to obtain the magnetic field distribution model; According to the magnetic field distribution model, the magnetic field abnormal area is identified, and the characteristic parameters of the abnormal area and its corresponding spatial position are recorded to obtain the abnormal area data; The abnormal area data is subjected to multi-level refinement analysis to generate magnetic field distribution results and abnormal characteristic data of the target area. The magnetic field distribution results include a magnetic field distribution map in the target area and annotation data of the abnormal area.

2. A magnetic field observation method suitable for high-density deployment according to claim 1, characterized in that: According to the background noise generated by the characteristics of the equipment, the collected data is preliminarily screened to generate preliminary screening data, including: Extracting equipment operation characteristics of each observation node, wherein the equipment operation characteristics include a frequency range and a fluctuation value of the noise; Compare the collected data point by point according to the equipment operation characteristics, filter out the stable fluctuation mode data in the data, and match it with the noise characteristics corresponding to the equipment operation characteristics to obtain stable fluctuation data; The stable fluctuation data in the collected data are eliminated step by step to obtain the preliminary screening data.

3. A magnetic field observation method suitable for high-density deployment according to claim 2, characterized in that: According to the spatial distribution characteristics between observation nodes, identify and mark abnormal data caused by mutual interference between nodes, perform secondary screening on the preliminary screening data, and generate observation data, including: Obtain the number, interval and location distribution of observation nodes; By analyzing the data change trends of adjacent nodes, the abnormal fluctuations caused by the interference of adjacent nodes are identified and marked to obtain abnormal fluctuation data; The abnormal fluctuation data is processed in segments, and the abnormal fluctuation part in the segment is removed to obtain the characteristic data of the target area; Integrate the characteristic data of the target area to generate observation data.

4. A magnetic field observation method suitable for high-density deployment according to claim 3, characterized in that: According to the spatial position association rules, the observation data of different observation nodes are aggregated to generate a preliminary distribution data model of the regional magnetic field, including: According to the deployment location of each observation node, the target area is divided into gridded spatial units, and each spatial unit corresponds to the observation data of one or more nodes; Compare the observation data in each gridded spatial unit, identify and merge multiple observation data in the same unit, and obtain grid data; By comparing the trends of grid data between gridded spatial units, the data changes of adjacent units are smoothed to obtain smooth data; The smoothed data are combined to generate a preliminary distribution data model of the target area, which reflects the overall characteristics of the magnetic field in the target area.

5. A magnetic field observation method suitable for high-density deployment according to claim 4, characterized in that: According to the preliminary distribution data model, the data consistency between the observation nodes in the target area is cross-validated to obtain the verification data model. The verification data model is optimized through the precision improvement algorithm to obtain the magnetic field distribution model, including: Extract the associated range of the observation results of each observation node in the target area according to the preliminary distribution data model, and divide the data within the associated range into multiple verification unit data; Compare the data from different observation nodes in each verification unit data, identify the difference values ​​and outliers, and record the difference values, outliers and source nodes; According to the difference value, outliers and source nodes, the observation error caused by the position deviation between the observation nodes is adjusted, and the outlier data is smoothed to obtain the unit adjustment data; Recombining unit adjustment data to generate a validation data model; Based on the difference value and observation error of the observation nodes in the verification data model, the observation data of the nodes in the neighboring area are dynamically adjusted through the optimization of the global consistency error and used to generate the magnetic field distribution model; the optimization of the global consistency error is achieved through the following formula: ,in, is the global consistency error loss function, For Node and nodes The observation error, , For Node and nodes The distance is the weight coefficient, is the number of nodes, For Node The set of neighbor nodes of is the regularization coefficient, For Node and nodes The difference value of , and Node and nodes Observation data.

6. A magnetic field observation method suitable for high-density deployment according to claim 5, characterized in that: According to the magnetic field distribution model, the magnetic field abnormal area is identified, and the characteristic parameters of the abnormal area and its corresponding spatial position are recorded to obtain the abnormal area data, including: According to the magnetic field distribution model, calculate the position The second-order gradient at , obtains the variation range data; According to the variation range and local fluctuation characteristics, the significance score is calculated to measure the significance level of the magnetic field anomaly area. The calculation formula of the significance score of the anomaly area A is: ， in, is the significance score of abnormal region A, For location The second-order gradient of the magnetic field strength at , and is the magnetic field strength and The second-order partial derivative of For location The magnetic field strength, is the average magnetic field strength in the target area, and is the integral variable; According to the significance score, the area exceeding the preset significance score threshold is screened, the area boundary is marked and the magnetic field anomaly area is generated; Calculate the change rate, local intensity peak and change direction of the magnetic field anomaly area, extract the characteristic data inside the area, and form the characteristic parameter data of the magnetic field anomaly area; Associating each feature point in the feature parameter set with the spatial coordinates of the target area, recording the boundary information and center position of the magnetic field anomaly area, and generating spatial position data; According to the characteristic parameter data and the spatial position data, abnormal region data is generated, and the abnormal region data is used to describe the specific distribution and characteristics of the magnetic field anomaly in the target region.

7. A magnetic field observation method suitable for high-density deployment according to claim 6, characterized in that: Perform multi-level refinement analysis on the abnormal area data to generate the magnetic field distribution results and abnormal characteristic data of the target area. The magnetic field distribution results include the magnetic field distribution map in the target area and the annotation data of the abnormal area, including: Obtain boundary information, center position and characteristic parameters in abnormal area data; The abnormal area is divided into multiple sub-areas according to the boundary information, and the local change characteristics are calculated for each sub-area, and the intensity range, change direction and gradient value of the sub-area are extracted to generate the local change characteristics; The change characteristics of each sub-region are hierarchically classified, and the local eigenvalues ​​are aggregated according to the multi-dimensional weights to generate multi-level refined features; the calculation formula of the eigenvalue is: ,in, is the first The eigenvalues ​​of the layer aggregated features, is the first The characteristic values ​​of each feature, including intensity, direction, and rate of change, For the The weight coefficient of the feature, is the number of features; By comparing the multi-level refined features with the surrounding non-abnormal area data, the correlation features inside and outside the area are analyzed, and the boundary of the abnormal area is optimized through the following dynamic adjustment formula to generate boundary data; the dynamic adjustment formula is: ,in, is the optimized magnetic field strength, is the original magnetic field strength, is the threshold for boundary adjustment, where: , is the maximum value of the magnetic field intensity in the abnormal area A, is the minimum value of the magnetic field intensity in the abnormal area A, is the mean of the second-order gradient in the abnormal area A, is the standard deviation of the second-order gradient in the abnormal area A, is the rate of change of magnetic field intensity in abnormal area A, is the peak value of the magnetic field intensity in the abnormal area A, is the difference in direction of magnetic field intensity in abnormal area A, , , is the direction of the magnetic field strength, is the magnetic field strength along The gradient of the direction, is the magnetic field strength along The gradient of the direction, is the inverse tangent function, , , , , , and is the weight coefficient, is the adjustment coefficient, is the area of ​​abnormal region A; Integrate boundary data with the magnetic field distribution model to generate a magnetic field distribution map of the target area; Mark the location, characteristic parameters and boundary changes of the abnormal area on the magnetic field distribution map to form a distribution result map; According to the abnormal area data in the distribution result map, abnormal characteristic data is generated, including the boundary range, center position, intensity change and directional characteristics of each abnormal area.

8. A magnetic field observation system suitable for high-density deployment, characterized in that: Applied to the method according to any one of claims 1 to 7, the system comprises: An observation planning module is used to obtain observation planning data, wherein the observation planning data includes the geographical boundaries of the target area, the accuracy requirements of the magnetic field observation target, and the deployment environment characteristics; The layout plan module is used to determine the number, interval and location distribution of observation nodes according to the observation planning data, and form an observation node layout plan; A node setting module is used to set multiple observation nodes to designated locations in the target area according to the observation node layout plan to form an observation node network; A data acquisition module is used to acquire magnetic field data in a target area through an observation node network to obtain acquired data, wherein the acquired data includes magnetic field intensity data and direction data; A preliminary screening module is used to perform preliminary screening on the collected data according to the background noise generated by the characteristics of the equipment and generate preliminary screening data; The secondary screening module is used to identify and mark abnormal data caused by mutual interference between nodes according to the spatial distribution characteristics between observation nodes, perform secondary screening on the primary screening data, and generate observation data; A model generation module is used to aggregate the observation data of different observation nodes according to the spatial position association rules to generate a preliminary distribution data model of the regional magnetic field; The model optimization module is used to cross-validate the data consistency between the observation nodes in the target area according to the preliminary distribution data model to obtain the verification data model, and optimize the verification data model through the precision improvement algorithm to obtain the magnetic field distribution model; The abnormality identification module is used to identify the abnormal magnetic field area according to the magnetic field distribution model, and record the characteristic parameters of the abnormal area and its corresponding spatial position to obtain the abnormal area data; The abnormal analysis module is used to perform multi-level detailed analysis on the abnormal area data to generate the magnetic field distribution results and abnormal characteristic data of the target area. The magnetic field distribution results include the magnetic field distribution map in the target area and the annotation data of the abnormal area.

9. A computing device, characterized in that include: one or more processors; A storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the method as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program, which, when executed by a processor, implements the method according to any one of claims 1 to 7.

Images