A continuous gravity detection method, medium and system based on seafloor crawler

Abstract

The present invention provides a continuous gravity detection method, medium and system based on a seafloor crawler, which belongs to the field of geophysical technology, and includes: using a seafloor crawler equipped with an absolute gravimeter and a relative gravimeter to perform grid detection of a test area. First, the relative gravity data of the center point of each grid unit is obtained to establish a relative gravity matrix. Then, the Bayesian probability model and the Markov chain Monte Carlo method are used to determine the abnormal points, boundary points and transition points of the relative gravity data. These feature points are clustered to obtain multiple cluster centers and flying points. Next, the seafloor crawler is controlled to traverse the path of the cluster centers and flying points to obtain absolute gravity data. Finally, in combination with the relative gravity matrix, the Kriging interpolation method is used to calculate the absolute gravity matrix of the entire test area as the gravity detection result. The present invention solves the technical problem that the prior art is difficult to achieve the detection of high-precision seafloor gravity data.

Description

Technical Field

[0001] The present invention belongs to the field of geophysical technology, and in particular, relates to a continuous gravity detection method, medium and system based on a seafloor crawler. Background Art

[0002] Gravity exploration is a common method in geophysical exploration. By measuring and analyzing the earth's gravity field, information on underground density distribution can be obtained, providing important support for geological structure research and mineral exploration. In the marine environment, gravity exploration has unique application value and can be used to detect seabed geological structures and determine the location of oil and gas reservoirs. Compared with the terrestrial environment, marine gravity exploration faces many challenges, mainly including:

[0003] 1. Limited measurement conditions: Gravity measurements from ships or aircraft have low measurement accuracy and resolution, while the deployment cost of traditional seabed fixed measurement stations is high, making it difficult to achieve continuous measurement.

[0004] 2. Complex seabed environment: The seabed topography is undulating, with many mountains, trenches and other terrain obstacles, which bring great difficulties to measurement and data processing. At the same time, changes in environmental factors such as seabed currents, temperature and pressure will also interfere with gravity data.

[0005] 3. Complex data processing: Due to the complex seabed environment and weak gravity anomaly signals, it is difficult to directly extract valuable geological information from the raw data, and complex data processing and analysis are required.

[0006] In addition, the advantage of relative gravimeters is that they have a fast single measurement speed, but there is a data drift problem; the advantage of absolute gravimeters is that the data is stable, but there are problems with single measurement time-consuming and slow speed. In order to better solve these problems, some new gravity exploration technologies have begun to be applied to marine environments in recent years, such as unmanned underwater vehicles (AUVs) and underwater remotely operated vehicles (ROVs). These unmanned vehicles can move and operate autonomously on the seabed, and use the gravimeters they carry to obtain more detailed gravity data. However, existing unmanned vehicle technologies usually have problems such as short battery life and limited range of activities, making it difficult to achieve continuous exploration of the entire seabed area.

[0007] In other words, it is difficult to detect high-precision gravity data on the seabed with existing technology. Summary of the invention

[0008] In view of this, the present invention provides a continuous gravity detection method, medium and system based on a seabed crawler, which can solve the technical problem that the prior art is difficult to achieve detection of high-precision seabed gravity data.

[0009] The present invention is achieved in that:

[0010] A first aspect of the present invention provides a continuous gravity detection method based on a seafloor crawler, wherein gravity data detection is performed using a seafloor crawler equipped with an absolute gravimeter and a relative gravimeter, comprising the following steps:

[0011] S10, dividing the seabed gravity measurement area into grids to obtain a plurality of grid units arranged in rows and columns;

[0012] S20, controlling the seafloor crawler to traverse the center of each grid unit and obtain the relative gravity data of the center point of each grid unit, and establishing a relative gravity matrix of the area to be measured;

[0013] S30, calculating the similarity between each element in the relative gravity matrix and the preset reference relative gravity data as standard similarity, and calculating the similarity between adjacent elements as adjacent similarity;

[0014] S40, using the Bayesian probability model and Markov chain Monte Carlo method to determine the abnormal points, boundary points and transition points of the standard similarity and adjacent similarities respectively:

[0015] S50, clustering the abnormal points, boundary points and transition points to obtain a plurality of cluster centers and flying points;

[0016] S60, establishing a traversal path between a plurality of cluster centers and flying points, controlling the seafloor crawler to crawl each cluster center and flying point and obtaining absolute gravity data of each cluster center and flying point;

[0017] S70. Based on the acquired absolute gravity data points and in combination with the previous relative gravity matrix, the Kriging interpolation method is used to interpolate the entire area to be measured, and finally the absolute gravity matrix of the area to be measured is obtained.

[0018] On the basis of the above technical solution, the continuous gravity detection method based on the seafloor crawler of the present invention can also be improved as follows:

[0019] Wherein, the side length of the grid unit is 50 meters to 500 meters.

[0020] Furthermore, the method for traversing the center of each grid cell is a traveling salesman problem algorithm.

[0021] Furthermore, the relative gravity data includes the coordinates of the measurement point, the measurement time, the relative gravity reading and environmental parameters, including water depth, temperature and pressure.

[0022] Furthermore, the similarity is calculated by using a weighted method of multiple similarities, wherein the multiple similarities include: Euclidean distance, cosine similarity, and Pearson correlation coefficient.

[0023] Furthermore, a method combining the Bayesian probability model and the Markov Chain Monte Carlo is adopted, specifically: the standard similarity and the adjacent similarity are used as observation data, it is assumed that the gravity field follows a Gaussian process distribution as prior knowledge, and the posterior distribution is sampled through the Markov Chain Monte Carlo method.

[0024] Furthermore, the method for clustering the outliers, boundary points and transition points is a DBSCAN algorithm.

[0025] Specifically, the steps of the present invention are described in detail as follows:

[0026] S10. Finely grid the seabed gravity test area: First, determine the appropriate grid resolution based on the total area of ​​the test area, the expected characteristic scale of gravity anomalies, and the complexity of the seabed topography. Use an equally spaced rectangular coordinate grid system to divide the entire test area into multiple grid cells arranged in rows and columns. Taking into account the particularity of the seabed environment, the seabed topography data is combined in the division process to appropriately encrypt the grid for complex terrain areas (such as submarine mountains, trenches, etc.). At the same time, a grid numbering system is established to assign a unique identifier to each grid cell to facilitate subsequent data processing and analysis. Finally, a visual image of the grid division is generated, including a plan view and a three-dimensional stereogram, to intuitively display the grid distribution.

[0027] S20, control the seafloor crawler to collect comprehensive relative gravity data: Based on the grid division results in S10, plan the optimal measurement path of the seafloor crawler. Considering the seafloor topography, ocean current conditions and the technical parameters of the crawler (such as endurance, moving speed, etc.), design a traversal path that can cover the center points of all grid cells. At the center point of each grid cell, control the seafloor crawler to accurately position and maintain stability, and use a high-precision relative gravimeter to measure. Record the precise coordinates, measurement time, relative gravity readings and environmental parameters (such as water depth, temperature, pressure, etc.) of each measurement point. To ensure data quality, calibrate the instrument before and after each measurement, and return to the reference point regularly for drift correction throughout the measurement process. All collected data are transmitted and stored in real time to construct an initial data matrix containing spatial position and relative gravity values. Finally, preprocess the raw data, including outlier detection, drift correction and environmental factor compensation, to establish a high-quality relative gravity matrix of the area to be measured.

[0028] S30, conduct a comprehensive similarity analysis of relative gravity data: First, based on the regional geological background and the known gravity field model, determine a suitable preset reference relative gravity data set. For the relative gravity matrix obtained in S20, calculate the standard similarity between each element and the preset reference data. Use a variety of similarity calculation methods (such as Euclidean distance, cosine similarity, Pearson correlation coefficient, etc.), and obtain a comprehensive standard similarity index through weight combination. Next, calculate the similarity of relative gravity values ​​between adjacent grid cells, which is recorded as adjacent similarity. Considering the complexity of the seabed environment, adaptive weights are introduced when calculating adjacent similarity, and the influence of adjacent cells is dynamically adjusted according to the degree of terrain change. Construct a similarity heat map to intuitively display the spatial distribution of standard similarity and adjacent similarity. Finally, based on the two types of similarity indicators calculated, preliminarily identify potential gravity anomaly areas, gradient change areas, and areas of suspicious data quality, providing a basis for subsequent Bayesian analysis and clustering. At the same time, generate a similarity analysis report, including statistical summaries, anomaly indicators, and preliminary interpretations, to provide a basis for the next step of fine analysis and selection of absolute gravity measurement points.

[0029] S40. Using the Bayesian probability model and the Markov Chain Monte Carlo (MCMC) method, a spatial distribution model of the gravity field is constructed: the standard similarity (the similarity of each element to the preset reference relative gravity data) and the adjacent similarity (the similarity between adjacent elements) are used as observation data, and the posterior distribution is sampled by the MCMC method assuming that the gravity field follows a Gaussian process distribution as prior knowledge. The sampling results are analyzed to identify abnormal points (points that are significantly different from the surrounding areas), determine boundary points (areas where the gravity value changes sharply), identify transition points (areas where the gravity value changes gradually), and judge normal points (points that conform to the overall trend).

[0030] S50. Perform multi-level clustering analysis on the feature points determined in the previous step: First, use the DBSCAN algorithm to perform density clustering on the abnormal points to identify the gravity anomaly area; then apply the hierarchical clustering method to process the boundary points and accurately divide the boundary of the gravity anomaly; use K-means clustering on the transition points to determine the gravity gradient change area; mark the isolated points in the clustering results or the points that do not meet the clustering criteria as flying points. Finally, calculate the coordinates of the center point of each cluster to obtain multiple representative cluster centers and outlier flying points, which will serve as key positions for subsequent absolute gravity measurements.

[0031] S60. Based on the obtained cluster centers and flying points, optimize the measurement path and perform absolute gravity measurement: Use the improved traveling salesman problem (TSP) algorithm, combined with the seabed topography and potential obstacle information, to plan the optimal traversal path for the seabed crawler. Consider the measurement characteristics of the absolute gravimeter (such as the need for a long time to stabilize), and arrange the measurement sequence and time reasonably. Control the seabed crawler to move accurately to each measurement point to perform absolute gravity measurement. At each measurement point, record a complete data set, including precise coordinates, measurement time, environmental parameters (such as temperature, pressure), absolute gravity value and its uncertainty.

[0032] S70. Integrate absolute gravity measurement data and relative gravity matrix to generate a complete gravity field model: Take the acquired absolute gravity measurement points as known points and relative gravity data as auxiliary variables to construct a variogram model to describe spatial correlation. Apply the co-kriging interpolation method to comprehensively utilize absolute gravity and relative gravity data to perform high-precision interpolation of the entire area to be measured. Obtain a complete absolute gravity matrix, including the coordinates of each grid point and the estimated absolute gravity value. At the same time, calculate the uncertainty of the interpolation results, generate error estimation maps and confidence intervals. Finally, output a complete gravity data set, including the coordinates of each grid point, relative gravity value, estimated absolute gravity value and its uncertainty, and generate visualization results such as gravity anomaly maps and contour maps, providing comprehensive gravity field distribution information and accuracy evaluation for subsequent geological interpretation and resource assessment.

[0033] A second aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores program instructions, and when the program instructions are executed, they are used to execute the above-mentioned continuous gravity detection method based on a seafloor crawler.

[0034] The third aspect of the present invention provides a continuous gravity detection system based on a seafloor crawler, wherein a survey ship, a seafloor crawler, a relative gravimeter, an absolute gravimeter and a control chip are provided, wherein the survey ship is connected to the seafloor crawler via a towing cable, the seafloor crawler is provided with the relative gravimeter and the absolute gravimeter, the seafloor crawler is built with the control chip, and the control chip contains the above-mentioned computer-readable storage medium. The relative gravimeter uses one or more relative gravity gradiometers, and the absolute gravimeter uses one or more absolute gravity gradiometers.

[0035] The technical effects of a continuous gravity detection system based on a seabed crawler provided by the present invention are as follows: since it is separated from the survey ship platform, the gravimeter on the seabed crawler is not affected by the shaking of the hull, and the gravity measurement accuracy jumps from the milligal (mGal) level to the microgal (μGal) level or even the nanoga (nGal) level, and the measurement accuracy is improved by at least 2 orders of magnitude.

[0036] Moreover, since gravity is measured on the seabed, it is not affected by the low-pass filtering of the thick seawater layer, and the high-frequency information of the seabed gravity signal will not be lost, so the measurement data is more real and effective. Compared with conventional seabed gravimeters (such as INO manufactured by Canada's Sintrex), the gravity detection system based on the seabed crawler can perform continuous seabed operations, and does not need to be retracted and released once for each measuring point like conventional seabed gravimeters, which greatly improves the operation efficiency. Each observation value of the absolute gravimeter takes a long time, and generally requires the seabed crawler to be stationary for about 1 minute. In order to further improve the observation efficiency, the absolute gravimeter and relative gravimeter are integrated on the seabed crawler body. The relative gravimeter is used for relative gravity measurement in most of the operation time, and only a small amount of absolute gravity observation is performed in a few operation times for quality control, drift correction and absolute value correction, which can further greatly improve the efficiency of seabed gravity detection.

[0037] On the basis of the above technical solution, the continuous gravity detection system based on the seafloor crawler of the present invention can also be improved as follows:

[0038] Furthermore, the relative gravimeter adopts a relative gravity gradiometer, and the absolute gravimeter adopts an absolute gravity gradiometer.

[0039] Compared with the prior art, the continuous gravity detection method, medium and system based on the seafloor crawler provided by the present invention have the following beneficial effects:

[0040] 1. Achieved comprehensive exploration of the seabed area: This method adopts a grid-based measurement strategy and uses a seabed crawler to comprehensively traverse the entire area to be measured, obtaining relative gravity data covering the entire seabed. Compared with traditional fixed measuring stations or temporary measuring lines, this method greatly improves the spatial coverage and resolution of the exploration.

[0041] 2. Fusion of relative gravity and absolute gravity data: This method first uses relative gravity data to identify key features of the gravity field, such as abnormal areas and gradient change areas, and then performs precise absolute gravity measurements on these areas. By fusing relative gravity and absolute gravity data, this method achieves a high-precision description of the entire seafloor gravity field.

[0042] 3. Application of advanced data analysis technology: This method uses the Bayesian probability model and MCMC sampling technology to extract rich gravity field feature information from relative gravity data; at the same time, it applies density clustering, hierarchical clustering and other algorithms to achieve accurate positioning of key measurement points. These data analysis technologies have greatly improved the ability to extract and analyze gravity field information.

[0043] 4. Good adaptability and robustness: This method fully considers the complexity of the seabed environment and introduces adaptive mechanisms in grid division, similarity calculation, cluster analysis and other links. It can dynamically adjust parameters and strategies according to actual conditions and maintain high data quality and analysis accuracy in complex seabed environments.

[0044] In summary, the present invention solves the technical problem that the prior art is difficult to achieve detection of high-precision seabed gravity data. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A flow chart of a continuous gravity detection method based on a seafloor crawler provided by the present invention;

[0046] Figure 2 A schematic diagram of the composition of a continuous gravity detection system based on a seafloor crawler provided by the present invention;

[0047] Figure 3 This is a schematic diagram of the composition of the wired transmission detection equipment of Example 2;

[0048] Figure 4 This is a schematic diagram of the composition of the wireless transmission detection equipment of Example 2. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0050] like Figure 1 As shown in FIG. 1 , it is a flow chart of a continuous gravity detection method based on a seafloor crawler provided by the present invention. The method uses a seafloor crawler equipped with an absolute gravimeter and a relative gravimeter to detect gravity data, and includes the following steps:

[0051] S10, dividing the seabed gravity measurement area into grids to obtain a plurality of grid units arranged in rows and columns;

[0052] S20, controlling the seafloor crawler to traverse the center of each grid unit and obtain the relative gravity data of the center point of each grid unit, and establishing a relative gravity matrix of the area to be measured;

[0053] S30, calculating the similarity between each element in the relative gravity matrix and the preset reference relative gravity data as standard similarity, and calculating the similarity between adjacent elements as adjacent similarity;

[0054] S40, using the Bayesian probability model and Markov chain Monte Carlo method to determine the abnormal points, boundary points and transition points of the standard similarity and adjacent similarities respectively:

[0055] S50, clustering the abnormal points, boundary points and transition points to obtain multiple cluster centers and flying points;

[0056] S60, establishing traversal paths between multiple cluster centers and flying points, controlling the seabed crawler to crawl each cluster center and flying point and obtaining absolute gravity data of each cluster center and flying point;

[0057] S70. Based on the acquired absolute gravity data points and in combination with the previous relative gravity matrix, the Kriging interpolation method is used to interpolate the entire area to be measured, and finally the absolute gravity matrix of the area to be measured is obtained.

[0058] The specific implementation methods of the above steps are described in detail below:

[0059] The specific implementation of step S10 is: first, determine the appropriate grid resolution according to the total area of ​​the area to be tested, the expected characteristic scale of gravity anomalies, and the complexity of the seabed terrain. Use an equidistant rectangular coordinate grid system to divide the entire area to be tested into multiple grid cells arranged in rows and columns. Considering the particularity of the seabed environment, the seabed terrain data is combined in the division process to appropriately encrypt the grid for complex terrain areas (such as seabed mountains, trenches, etc.). For example, for a total area of , the expected characteristic scale of gravity anomaly is , For areas with complex seabed terrain, you can choose The grid resolution is 10000. At the same time, a grid numbering system is established to assign a unique identifier to each grid unit to facilitate subsequent data processing and analysis. Finally, a visual image of the grid division is generated, including a plane map and a three-dimensional stereo map, to intuitively display the grid distribution. The main purpose of this step is to reasonably divide the area to be measured according to the actual situation, providing a basis for subsequent relative gravity data acquisition and absolute gravity measurement.

[0060] The specific implementation method of step S20 is: based on the grid division result in S10, plan the optimal measurement path of the seabed crawler. Considering the seabed topography, ocean current conditions and the technical parameters of the crawler (such as endurance, moving speed, etc.), design a traversal path that can cover the center points of all grid units. For example, an improved traveling salesman problem (TSP) algorithm can be used to determine the optimal path. At the center point of each grid unit, the seabed crawler is controlled to accurately position and maintain stability, and a high-precision relative gravimeter is used for measurement. The precise coordinates, measurement time, relative gravity readings and environmental parameters (such as water depth, temperature, pressure, etc.) of each measurement point are recorded. To ensure data quality, the instrument is calibrated before and after each measurement, and the reference point is returned to regularly for drift correction during the entire measurement process. All collected data are transmitted and stored in real time to construct an initial data matrix containing spatial position and relative gravity values. Finally, the raw data is preprocessed, including outlier detection, drift correction and environmental factor compensation, to establish a high-quality relative gravity matrix of the area to be measured. The main purpose of this step is to collect comprehensive relative gravity data of the area to be tested, providing basic data for subsequent similarity analysis and absolute gravity measurement.

[0061] The specific implementation of step S30 is: first, based on the regional geological background and the known gravity field model, determine a suitable preset reference relative gravity data set. For the relative gravity matrix obtained in S20, calculate the standard similarity between each element and the preset reference data. Use a variety of similarity calculation methods (such as Euclidean distance, cosine similarity, Pearson correlation coefficient, etc.), and obtain a comprehensive standard similarity index through weight combination. For example, the standard similarity can be defined as for:

[0062] ;

[0063] in, For elements The Euclidean distance between the reference data, is the cosine similarity, is the Pearson correlation coefficient, is the corresponding weight coefficient. Next, the similarity of relative gravity values ​​between adjacent grid cells is calculated, which is recorded as adjacent similarity. Considering the complexity of the seabed environment, adaptive weights are introduced when calculating adjacent similarity, and the influence of adjacent cells is dynamically adjusted according to the degree of terrain change. A similarity heat map is constructed to intuitively display the spatial distribution of standard similarity and adjacent similarity. Finally, based on the two types of similarity indicators calculated, potential gravity anomaly areas, gradient change areas, and areas of suspicious data quality are preliminarily identified, providing a basis for subsequent Bayesian analysis and clustering. The main purpose of this step is to conduct a comprehensive analysis of the acquired relative gravity data, and provide a basis for the subsequent extraction of gravity field features and selection of absolute gravity measurement points.

[0064] The specific implementation of step S40 is: taking the standard similarity (the similarity between each element and the preset reference relative gravity data) and the adjacent similarity (the similarity between adjacent elements) as observation data, assuming that the gravity field follows the Gaussian process distribution as prior knowledge, and sampling the posterior distribution through the Markov Chain Monte Carlo (MCMC) method. For example, the Gaussian process model can be defined as:

[0065] ;

[0066] in, is the mean function, is the covariance kernel function. The MCMC method is used to infer the parameters of the model and obtain the posterior distribution samples. The sampling results are analyzed to identify outliers (points that are significantly different from the surrounding areas), determine boundary points (areas where gravity values ​​change sharply), identify transition points (areas where gravity values ​​change gradually), and judge normal points (points that conform to the overall trend). For example, the standard similarity threshold of outliers can be set to ,The adjacent similarity threshold of the boundary points is ,The adjacent similarity threshold of the transition point is The main purpose of this step is to construct a spatial distribution model of the gravity field and identify gravity anomaly areas with different characteristics, providing support for subsequent cluster analysis and selection of absolute gravity measurement points.

[0067] The specific implementation of step S50 is: first use the DBSCAN algorithm to perform density clustering on the abnormal points and identify the gravity anomaly area. The core parameters of the DBSCAN algorithm include the neighborhood radius and the minimum number of samples , these two parameters can be reasonably set according to the characteristics of the data. For example, , Then, the hierarchical clustering method is applied to process the boundary points and accurately divide the boundary of gravity anomaly. The termination condition of hierarchical clustering can be set as the average adjacent similarity between clusters after clustering is less than . K-means clustering is used for the transition points to determine the gravity gradient change area. The number of clusters of K-means clustering It can be determined by the silhouette coefficient or elbow rule, for example . Isolated points in the clustering results or points that do not meet the clustering criteria are marked as flying points. Finally, the coordinates of the center point of each cluster are calculated to obtain multiple representative cluster centers and outlier flying points, which will serve as key positions for subsequent absolute gravity measurements. The main purpose of this step is to conduct an in-depth analysis of the previously identified feature points, determine representative cluster centers and outlier points, and provide guidance for subsequent absolute gravity measurements.

[0068] The specific implementation method of step S60 is: using the improved traveling salesman problem (TSP) algorithm, combined with the seabed terrain and potential obstacle information, to plan the optimal traversal path for the seabed crawler. For example, on the basis of the standard TSP algorithm, a penalty term can be introduced to avoid complex terrain areas. At the same time, considering the measurement characteristics of the absolute gravimeter (such as the need for a long time to stabilize), the measurement sequence and time are reasonably arranged. The seabed crawler is controlled to move accurately to each measurement point to perform absolute gravity measurement. At each measurement point, a complete data set is recorded, including precise coordinates, measurement time, environmental parameters (such as temperature, pressure), absolute gravity value and its uncertainty. The main purpose of this step is to accurately measure the key absolute gravity measurement points based on the aforementioned clustering results and obtain a complete absolute gravity data set.

[0069] The specific implementation of step S70 is: taking the acquired absolute gravity measurement points as known points and the relative gravity data as auxiliary variables, constructing a variogram model to describe spatial correlation. For example, an exponential variogram model can be used:

[0070] ;

[0071] in, is the variance function, For distance, is the base value, is the peak value, is the scale parameter. The co-kriging interpolation method is applied to comprehensively utilize the absolute gravity and relative gravity data to perform high-precision interpolation on the entire area to be measured. A complete absolute gravity matrix is ​​obtained, which contains the coordinates of each grid point and the estimated absolute gravity value. At the same time, the uncertainty of the interpolation result is calculated, and the error estimation map and confidence interval are generated. Finally, a complete gravity data set is output, including the coordinates of each grid point, the relative gravity value, the estimated absolute gravity value and its uncertainty, and visual results such as gravity anomaly maps and contour maps are generated, providing comprehensive gravity field distribution information and accuracy evaluation for subsequent geological interpretation and resource assessment. The main purpose of this step is to comprehensively utilize the absolute gravity and relative gravity data obtained above, and generate a complete gravity field model through high-precision interpolation methods to provide support for further geological interpretation and resource exploration.

[0072] In summary, this continuous gravity detection method based on seafloor crawlers achieves fine gravity exploration of the target area through seven steps. Among them, step S10 determines a reasonable grid division scheme, laying the foundation for subsequent data collection and processing; step S20 conducts comprehensive relative gravity data collection and obtains the initial data matrix of the area to be measured; step S30 conducts in-depth similarity analysis on the relative gravity data and identifies potential gravity anomaly areas; step S40 uses the Bayesian probability model and MCMC method to construct a spatial distribution model of the gravity field and extract gravity anomaly points with different characteristics; step S50 performs cluster analysis on the above-mentioned feature points and determines representative cluster centers and outliers; step S60 plans the optimal path of absolute gravity measurement based on the clustering results and obtains the absolute gravity data of key measurement points; finally, step S70 fuses the absolute gravity data and relative gravity data to generate a complete absolute gravity field model, which provides comprehensive support for subsequent geological interpretation and resource assessment. The whole method makes full use of the flexibility of the seafloor crawler and the measurement capability of the high-precision gravimeter, and realizes the detailed exploration of the seafloor gravity field through systematic data processing and analysis technology.

[0073] Specifically, the principle of the present invention is:

[0074] 1. Gridded full coverage: This method first divides the entire area to be measured into grid cells arranged in rows and columns, and uses a seafloor crawler to traverse and measure the center of each grid cell. This gridding method can not only ensure full coverage of the entire area, but also appropriately encrypt the grid in complex terrain areas according to actual conditions, further improving the measurement resolution. Compared with traditional temporary survey lines or fixed survey stations, the gridded measurement strategy greatly enhances the spatial continuity of exploration and the uniformity of data.

[0075] 2. Fusion of relative gravity and absolute gravity: This method uses a measurement method that combines relative gravity meters and absolute gravity meters. First, the relative gravity meter is used to quickly obtain the relative gravity data of the entire area to be measured to establish a preliminary gravity field distribution. Then, based on the analysis of the relative gravity data, key features such as gravity anomaly areas and gradient change areas are accurately located, and then the absolute gravity meter is used to perform precise measurements on these areas. This fusion of relative gravity and absolute gravity data gives full play to the respective advantages of the two gravity meters and improves the overall measurement accuracy and efficiency.

[0076] 3. Data analysis based on probability model: This method uses Bayesian probability model and MCMC sampling technology to conduct in-depth analysis of relative gravity data. First, it is assumed that the gravity field obeys Gaussian process distribution, and then the relative gravity data is used as the observed quantity, and the posterior distribution is sampled by MCMC method. In this way, not only can rich gravity field feature information such as anomalies, boundary points, transition points, etc. be extracted from the relative gravity data, but also the confidence of these features can be quantitatively evaluated. This analysis method based on probability model greatly enhances the ability to describe the distribution of gravity field in complex seabed environment.

[0077] 4. Multi-scale cluster analysis: This method uses multiple algorithms such as density clustering, hierarchical clustering, and K-means clustering for in-depth analysis of the gravity field features identified above. Through these clustering methods, not only can the boundaries of the abnormal area be accurately demarcated, but also the gravity gradient change area can be determined, and even some outliers can be identified. These representative cluster centers and outliers will serve as key locations for subsequent absolute gravity measurements. Multi-scale cluster analysis not only fully mines the characteristic information in relative gravity data, but also provides guidance for efficient absolute gravity measurement.

[0078] 5. Interpolation fusion based on co-kriging: Finally, this method integrates the absolute gravity measurement data obtained with the relative gravity matrix established in the early stage through the co-kriging interpolation method. Co-kriging interpolation can make full use of the correlation between the two types of data to achieve high-precision gravity field interpolation for the entire seabed area. At the same time, the uncertainty of the gravity value of each grid point can also be calculated, providing a reliable accuracy assessment for subsequent geological interpretation and resource assessment.

[0079] 6. Since it is separated from the survey ship platform, the gravimeter on the seabed crawler is not affected by the shaking of the ship, and the gravity measurement accuracy has jumped from the milligal (mGal) level to the microgal (μGal) level or even the nanogal (nGal) level, and the measurement accuracy has increased by at least 2 orders of magnitude. Moreover, since gravity is measured on the seabed, it is not affected by the low-pass filtering of the thick seawater layer, and the high-frequency information of the seabed gravity signal will not be lost, and the measurement data is more real and effective. Compared with conventional seabed gravimeters, the gravity detection system based on the seabed crawler can perform continuous seabed operations, and does not need to be retracted and released once for each measuring point like conventional seabed gravimeters, which greatly improves the operating efficiency.

[0080] In summary, the continuous gravity detection method of the present invention fully combines the flexibility of the seabed crawler and the measurement capability of the high-precision gravimeter, and realizes the detailed exploration and continuous monitoring of the gravity field in the complex seabed environment through key technologies such as comprehensive grid coverage, fusion of relative gravity and absolute gravity, data analysis based on probability model, multi-scale clustering and collaborative Kriging interpolation.

[0081] In order to better understand and implement the present invention, a specific embodiment 1 of the method of the present invention is provided below. The steps of this embodiment 1 are described in detail as follows:

[0082] First, in step S10, the area to be tested needs to be divided into grids. Specifically, the following method can be used:

[0083] Determining the grid resolution : According to the total area of ​​the area to be measured , Expected gravity anomaly characteristic scale The grid resolution can be reasonably determined based on the complexity of the seafloor topography. , the expected characteristic scale of gravity anomaly is , For areas with complex seabed terrain, you can choose The grid resolution.

[0084] Establishing grid coordinate system: using rectangular coordinate system ,in ,and .

[0085] Combine seafloor topography data to encrypt the grid: For complex seafloor topography areas (such as seamounts, trenches, etc.), appropriately reduce the grid size and improve the resolution. Adaptive grid refinement strategy can be used to dynamically adjust the grid size according to the degree of terrain change.

[0086] Establish a grid numbering system: For each grid cell Assigning unique identifiers ,in ; ; ; .

[0087] Generate mesh visualization: Output mesh partitioning visualization results including plane diagrams and three-dimensional stereo diagrams to intuitively display the mesh distribution.

[0088] Through the above steps, a grid system covering the area to be measured is obtained, which provides a basis for subsequent relative gravity data collection and absolute gravity measurement.

[0089] Next is step S20, which is to control the seafloor crawler to collect comprehensive relative gravity data. The specific implementation method is as follows:

[0090] Planning the optimal measurement path of the seafloor crawler: Based on the above grid division results, the improved traveling salesman problem (TSP) algorithm is used to design a traversal path that can cover the center points of all grid cells. The TSP problem can be described as:

[0091] ;

[0092] in, ; ; ;

[0093] in, is the total number of grid cells, For grid cells The distance between is a binary decision variable, indicating whether the edge is passed on the path. This optimization problem takes into account factors such as seabed topography, ocean current conditions, and crawler technical parameters, and can obtain an optimal measurement path that can cover the center points of all grid cells.

[0094] Relative gravity measurement is performed at the center of each grid cell: the seafloor crawler is controlled to accurately position and maintain stability, using a high-precision relative gravimeter Take measurements. Record the exact coordinates of each measurement point , measurement time , relative gravity readings and environmental parameters (such as water depth ,temperature ,pressure wait).

[0095] Preprocess the raw data: including outlier detection, drift correction, and environmental factor compensation. Outlier detection can be based on statistics (such as Drift correction can be performed by periodically returning to the reference point for calibration; compensation for environmental factors requires the establishment of a suitable physical model.

[0096] Construct relative gravity data matrix: Organize the preprocessed data into a matrix containing spatial positions and relative gravity values G r =[ g r ( x ij , t ij ) ] N x × N y ,in Respectively represent The center coordinates of each grid cell and the measurement time.

[0097] Through the above steps, a high-quality relative gravity data matrix of the area to be measured is obtained. , which provides basic data for subsequent similarity analysis and absolute gravity measurement.

[0098] Next is step S30, which is to conduct a comprehensive similarity analysis on the relative gravity data. The specific implementation method is as follows:

[0099] Determine the preset reference relative gravity dataset :Based on the regional geological background and known gravity field model, a suitable reference relative gravity dataset is selected.

[0100] Calculate standard similarity: for Each element in , calculate its difference with the reference data The standard similarity between A variety of similarity measurement methods can be used, such as Euclidean distance, cosine similarity and Pearson correlation coefficient, and a comprehensive standard similarity index can be obtained through weighted combination:

[0101] ;

[0102] in, is the Euclidean distance, is the cosine similarity, is the Pearson correlation coefficient, is the corresponding weight coefficient.

[0103] Calculate adjacent similarity: For adjacent grid cells and or , calculate the similarity of relative gravity values ​​between them Taking into account the complexity of the seabed environment, adaptive weights are introduced when calculating adjacent similarities to dynamically adjust the influence of adjacent units according to the degree of terrain change:

[0104] ;

[0105] ;

[0106] in, , is the standard deviation of the gravity difference between adjacent cells.

[0107] Construct similarity heat map: based on the calculated standard similarity and adjacent similarity ,The visualization results of the spatial distribution of similarity are generated, which can intuitively display the gravity anomaly areas, gradient change areas and areas with suspicious data quality.

[0108] Analyze similarity results: Combined with the above analysis, preliminarily identify potential gravity anomaly areas, gradient change areas, and areas of suspicious data quality, providing a basis for subsequent Bayesian analysis and clustering. At the same time, generate a similarity analysis report, including statistical summary, anomaly indicators and preliminary explanations.

[0109] Through the above similarity analysis steps, we have a deeper understanding of the acquired relative gravity data, which provides an important basis for the subsequent gravity field feature extraction and absolute gravity measurement point selection.

[0110] Next is step S40, which uses the Bayesian probability model and the Markov Chain Monte Carlo (MCMC) method to construct a gravity field spatial distribution model. The specific implementation method is as follows:

[0111] Define the Gaussian process model: Assume the gravity field It obeys the Gaussian process distribution, that is, ,in is the mean function, is the covariance kernel function. Commonly used covariance kernel functions include exponential kernel, Gaussian kernel, Mattern kernel, etc.

[0112] Establishing Bayesian probability model: transforming standard similarity and adjacent similarity As observation data, g=[g( x ij ) ] N x × N y As the hidden variable to be solved. According to the Bayesian formula, the posterior probability distribution can be obtained:

[0113] ;

[0114] in, is the likelihood function, is a Gaussian process prior.

[0115] Use MCMC methods for posterior sampling: Apply Markov chain Monte Carlo algorithms, such as Gibbs sampling or Metropolis-Hastings algorithm, to the posterior distribution Sampling is performed to obtain a set of samples that obey the distribution .

[0116] Analyze sampling results: Based on MCMC sampling, identify outliers (points that are significantly different from the surrounding area), determine boundary points (areas where gravity values ​​change sharply), identify transition points (areas where gravity values ​​change gradually), and judge normal points (points that conform to the overall trend). For example, the standard similarity threshold of outliers can be set to ,The adjacent similarity threshold of the boundary points is ,The adjacent similarity threshold of the transition point is .

[0117] By constructing a Bayesian probability model and using the MCMC method for posterior sampling, gravity anomaly points with different characteristics in the measured area can be identified, providing support for subsequent clustering analysis and selection of absolute gravity measurement points.

[0118] Next is step S50, that is, performing multi-level cluster analysis on the aforementioned identified feature points. The specific implementation method is as follows:

[0119] Density clustering of outliers: Use the DBSCAN algorithm to cluster outliers to identify gravity anomaly areas. The core parameters of the DBSCAN algorithm include the neighborhood radius and the minimum number of samples ,These two parameters can be set reasonably according to the characteristics of the data.

[0120] ;

[0121] in, Represents the clustering results, i.e., different gravity anomaly areas.

[0122] Hierarchical clustering of boundary points: Apply hierarchical clustering method to process boundary points to accurately divide the boundary of gravity anomaly. The termination condition of hierarchical clustering can be set as the average adjacent similarity between clusters after clustering is less than .

[0123] ;

[0124] in, Represents the result of hierarchical clustering, i.e., different gravity anomaly boundary areas.

[0125] K-means clustering of transition points: Use K-means algorithm to cluster transition points to determine the area where gravity gradient changes. It can be determined by methods such as silhouette coefficient or elbow rule, for example .

[0126] ;

[0127] ;

[0128] in, Represents the result of K-means clustering, namely different gravity gradient transition areas.

[0129] Mark outliers: Mark isolated points in the clustering results or points that do not meet the clustering criteria as outliers.

[0130] ;

[0131] Calculate the cluster center: For each cluster , calculate its centroid as the representative cluster center coordinates .

[0132] Through the above multi-level cluster analysis, gravity anomaly areas, boundary areas and gradient transition areas with different characteristics in the measured area were determined, and some outliers were identified. These cluster centers and outliers will serve as key locations for subsequent absolute gravity measurements.

[0133] Next is step S60, that is, based on the above clustering results, optimizing the measurement path and performing absolute gravity measurement. The specific implementation method is as follows:

[0134] Planning the absolute gravity measurement path: Using the improved traveling salesman problem (TSP) algorithm, combined with the seabed topography and potential obstacle information, the optimal traversal path is planned for the seabed crawler. The TSP problem can be described as:

[0135] ;

[0136] in, ; ; ;

[0137] in, is the cluster center and The distance between is the seafloor topography elevation difference, is the weight coefficient of the terrain factor. This optimization problem can obtain an optimal path covering all key measurement points.

[0138] Perform absolute gravity measurement: Control the seafloor crawler to move precisely to each measurement point (cluster center and flying point), using an absolute gravimeter At each measuring point , record the complete data set, including the exact coordinates , measurement time , environmental parameters (such as temperature ,pressure ), absolute gravity value and its uncertainty .

[0139] Through the above steps, the absolute gravity data of key measurement points were obtained, providing necessary information for the subsequent construction of the gravity field model.

[0140] Finally, step S70 is to comprehensively utilize the absolute gravity measurement data and the relative gravity matrix to generate a complete gravity field model. The specific implementation method is as follows:

[0141] Construct the variogram model: take the absolute gravity measurement point as the known point, , relative gravity data As auxiliary variables, a variogram model describing spatial correlation is established:

[0142] ;

[0143] in, is the variance function, For distance, is the base value, is the peak value, These model parameters can be determined by fitting absolute gravity measurement data.

[0144] Application of co-kriging interpolation: Based on the constructed variogram model, the co-kriging interpolation method is used to perform high-precision interpolation of the entire measured area to obtain a complete absolute gravity matrix :

[0145] ;

[0146] in, is the interpolation weight coefficient, which is determined by the variogram model.

[0147] Calculate the uncertainty of the interpolation result: Calculate the absolute gravity estimate for each grid point according to the Kriging variance formula Variance , which generates error estimate plots and confidence intervals.

[0148] Output gravity field results: Organize the complete gravity data set, including the coordinates of each grid point , relative gravity value , estimated absolute gravity and its uncertainty At the same time, visual results such as gravity anomaly maps and contour maps are generated to provide comprehensive gravity field distribution information and accuracy assessment for subsequent geological interpretation and resource assessment.

[0149] Through the above steps, a complete gravity field model of the area to be measured is finally obtained, including absolute gravity values, relative gravity values ​​and corresponding uncertainty information, which provides comprehensive support for further geological interpretation and resource exploration.

[0150] In summary, this continuous gravity detection method based on seafloor crawlers achieves fine gravity exploration of the target area through seven steps. Among them, step S10 determines a reasonable grid division scheme, laying the foundation for subsequent data collection and processing; step S20 conducts comprehensive relative gravity data collection and obtains the initial data matrix of the area to be measured; step S30 conducts in-depth similarity analysis on the relative gravity data and identifies potential gravity anomaly areas; step S40 uses the Bayesian probability model and MCMC method to construct a spatial distribution model of the gravity field and extract gravity anomaly points with different characteristics; step S50 performs cluster analysis on the above-mentioned feature points and determines representative cluster centers and outliers; step S60 plans the optimal path of absolute gravity measurement based on the clustering results and obtains the absolute gravity data of key measurement points; finally, step S70 fuses the absolute gravity data and relative gravity data to generate a complete absolute gravity field model, which provides comprehensive support for subsequent geological interpretation and resource assessment. The whole method makes full use of the flexibility of the seafloor crawler and the measurement capability of the high-precision gravimeter, and realizes the detailed exploration of the seafloor gravity field through systematic data processing and analysis technology.

[0151] The following is an example 2 of a practical application scenario of the present invention: In June 2024, a marine geological survey agency plans to conduct a large-scale seabed gravity detection mission in a certain sea area. The water depth in the area is about 800-1200m, and the area is about 100 square kilometers. The main purpose of this detection is to understand the geological structure characteristics of the area in detail and provide basic data support for potential oil and gas resource exploration.

[0152] Detection equipment configuration such as Figure 3-4 As shown:

[0153] 1. Survey vessel: The "Ocean" comprehensive geophysical survey vessel was used, with a total length of 103m, a beam of 18.8m, a depth of 8.5m, and a displacement of 5800t.

[0154] 2. Seabed crawler: A-400 large crawler underwater robot produced by Boya Gongdao (Beijing) Robotics Technology Co., Ltd. is used as the seabed crawler, with a weight of 3.5t, a length of 3.2m, a width of 2.1m and a height of 1.8m.

[0155] 3. Relative gravimeter: CG-6 relative gravimeter with an accuracy of 1μGal is used.

[0156] 4. Absolute gravimeter: A10-L absolute gravimeter with an accuracy of 10μGal is used.

[0157] 5. Control chip: It uses a self-developed chip based on the ARM Cortex-A72 architecture, with a main frequency of 2.5GHz and built-in 4GB RAM.

[0158] 6. Staffing: 1 captain, 2 drivers, 2 mechanical engineers, 3 geophysicists, 2 data analysts, a total of 10 staff members.

[0159] Mission preparation phase:

[0160] On June 1, 2024, the "Ocean" survey ship set out from a port in Guangzhou and is expected to arrive at the scheduled survey area in 6 hours. Before departure, the staff conducted a comprehensive inspection and calibration of all equipment.

[0161] 1. Equipment inspection:

[0162] a) Seabed crawler inspection:

[0163] Mechanical system: Confirm that the tracks, motors, sealed cabins, etc. are intact.

[0164] Power system: When the battery pack is fully charged, the endurance is more than 72 hours.

[0165] Communications System: Wired and acoustic communications systems with the survey vessel tested normal.

[0166] b) Gravity meter inspection:

[0167] CG-6 relative gravimeter: 24-hour continuous observation was carried out at the dock, and the drift rate was less than 0.02mGal / day.

[0168] A10-L absolute gravimeter: Comparative measurements were carried out with the national gravity benchmark, and the error was within ±3μGal.

[0169] c) Control chip test:

[0170] Run the self-test program and all modules are working normally.

[0171] The simulated data processing capability test can complete the Kriging interpolation operation of 100,000 grid points within 1 second.

[0172] 2. Mission planning:

[0173] a) Grid division: The survey area of ​​100 square kilometers is divided into 100x100 grids, and the area of ​​each grid unit is 100x100sqm.

[0174] b) Detection path: An "S"-shaped path is used for traversal, and the total distance is expected to be approximately 110 km.

[0175] c) Schedule:

[0176] Relative gravity measurement: Each grid point stays for 10 seconds, and 100 sampling data are obtained and averaged.

[0177] Absolute gravity measurement: performed every 2 hours, each measurement lasting 3 minutes.

[0178] d) Data transmission: Data is transmitted in real time via composite cables and backed up to local storage every 5 minutes.

[0179] Implementation process:

[0180] At 14:00 on June 1, 2024, the "Ocean" arrived at the designated sea area.

[0181] 1. Equipment deployment (14:00-15:30):

[0182] a) Sea condition assessment: Wind force 3, wave height 0.5m, suitable for detection operations.

[0183] b) Launching of crawler:

[0184] At 14:15, the lifting of the seabed crawler began.

[0185] At 14:30, the crawler entered the water and slowly descended.

[0186] At 15:25, the crawler reached the seabed (water depth 1050m) and landed stably.

[0187] 2. Initial calibration (15:30-16:00):

[0188] a) Absolute gravity measurement:

[0189] Measurement time: 15:30-15:33

[0190] Measurement results: 979123456.32μGal, standard deviation ±8μGal

[0191] b) Relative gravity measurement:

[0192] Measurement time: 15:33-15:34

[0193] Measurement result: 3012.56mGal (relative to the dock reference point)

[0194] c) Data comparison: The difference between the absolute gravity value and the relative gravity value is within the expected range, and the equipment is working normally.

[0195] 3. Continuous detection phase (16:00-14:00 the next day, a total of 22 hours):

[0196] a) Relative gravity measurement:

[0197] Starting at 16:00, the crawler began "S"-shaped path detection at a speed of 0.5m / s.

[0198] The measurement was performed every 100 m with a 10-second pause, for a total of 10,000 measurement points.

[0199] An average of 100 sampling data were obtained for each measurement point.

[0200] b) Absolute gravity measurement:

[0201] This was done every 2 hours, for a total of 11 measurements.

[0202] Each measurement lasted 3 minutes, with the crawler completely stationary.

[0203] c) Data records are shown in Table 1:

[0204] Table 1 Data record table

[0205] time Position (X, Y) Relative gravity value (mGal) Absolute gravity value (μGal) 16:00 (0,0) 3012.56 979123456.32 18:00 (20,35) 3013.12 979123458.45 20:00 (42,68) 3014.78 979123461.23 22:00 (61,24) 3015.33 979123463.67 00:00 (83,51) 3016.89 979123466.12 02:00 (5,79) 3018.45 979123469.34 04:00 (27,96) 3019.78 979123472.56 06:00 (49,13) 3021.23 979123475.78 08:00 (71,40) 3022.67 979123478.90 10:00 (92,67) 3024.12 979123482.13 12:00 (14,95) 3025.56 979123485.45 14:00 (99,99) 3027.01 979123488.67

[0206] 4. Data processing and analysis (detection is carried out simultaneously):

[0207] a) Relative gravity matrix construction: Use the control chip to process data in real time and construct a 100x100 relative gravity matrix. The matrix element value range is: 3012.56mGal to 3027.01mGal.

[0208] b) Similarity calculation:

[0209] Standard similarity: 3012.56 mGal of the initial measurement point (0,0) was selected as the reference value.

[0210] Adjacent similarity: Calculate the similarity between each point and its four adjacent points above, below, left, and right.

[0211] c) Outlier Identification:

[0212] Use Bayesian probability models and Markov Chain Monte Carlo methods.

[0213] Set thresholds: standard similarity difference > 1.5 mGal, adjacent similarity difference > 0.5 mGal.

[0214] Identification results: 37 abnormal points, 12 boundary points, and 8 transition points were found.

[0215] d) Cluster analysis: K-means clustering algorithm was used. Clustering results: 5 main cluster centers and 3 flying points were formed, as shown in Table 2 and Table 3.

[0216] Table 2 Cluster center coordinates:

[0217] Cluster number X coordinate Y coordinate Relative gravity value (mGal) 1 25 30 3015.67 2 60 55 3019.23 3 40 80 3022.45 4 85 20 3017.89 5 75 70 3024.12

[0218] Table 3 Flying point coordinates:

[0219] Flying point number X coordinate Y coordinate Relative gravity value (mGal) 1 10 90 3026.78 2 95 5 3013.34 3 50 50 3020.56

[0220] 5. Precise absolute gravity measurement (14:00-18:00):

[0221] a) Path planning: Based on the clustering results, the shortest path traversing the five cluster centers and three flying points is planned.

[0222] b) Measurement process:

[0223] Absolute gravity measurements were performed at each point for 5 minutes.

[0224] Relative gravity measurements were also performed to verify consistency.

[0225] c) The measurement results are shown in Table 4:

[0226] Table 4 Measurement results

[0227] Location Type Coordinates (X,Y) Absolute gravity value (μGal) Relative gravity value (mGal) Cluster Center 1 (25,30) 979123466.78 3015.67 Cluster Center 2 (60,55) 979123472.34 3019.23 Cluster Center 3 (40,80) 979123477.89 3022.45 Cluster Center 4 (85,20) 979123469.56 3017.89 Cluster Center 5 (75,70) 979123479.23 3024.12 Flying point 1 (10,90) 979123484.67 3026.78 Flying Point 2 (95,5) 979123461.23 3013.34 Flying point 3 (50,50) 979123474.56 3020.56

[0228] 6. Data integration and interpolation (18:00-20:00):

[0229] a) Data correction: Use the absolute gravity measurement results to perform drift correction on the relative gravity data. Correction formula: ;in, is the corrected gravity value, is the relative gravity measurement value, is the absolute gravity measurement value, is the relative gravity value of the reference point.

[0230] b) Kriging interpolation:

[0231] The entire 100x100 grid is interpolated using ordinary kriging.

[0232] Variogram model: A spherical model is used, and the parameters are obtained by least squares fitting.

[0233] Interpolation result: Get a 100x100 absolute gravity matrix with a resolution of 100m.

[0234] 7. Result analysis and report generation (20:00-22:00):

[0235] a) Gravity anomaly map: Generate contour map based on the interpolation results, with an interval of 0.5 mGal.

[0236] b) Geological interpretation:

[0237] A significant low gravity anomaly area was found, located between (30,60) and (70,80), which may indicate the existence of a sedimentary basin.

[0238] There is a high gravity anomaly near (80,20), which may be a basement uplift or a magmatic body.

[0239] c) Accuracy assessment:

[0240] Relative gravity measurement average accuracy: ±0.01mGal

[0241] Average accuracy of absolute gravity measurement: ±5μGal

[0242] Estimated accuracy of interpolated grid data: ±0.05mGal

[0243] d) Report generation: Automatically generate preliminary survey reports including method description, data statistics, anomaly distribution and preliminary geological interpretation.

[0244] Equipment recovery and mission completion:

[0245] 1. Equipment recycling (22:00-23:30, June 2, 2024):

[0246] a) At 22:00, the recovery command was sent and the seafloor crawler began to surface.

[0247] b) At 23:15, the crawler reached the sea surface and began to be hoisted.

[0248] c) At 23:30, the crawler returned safely to the deck and began cleaning and inspection.

[0249] 2. Equipment inspection and data backup (23:30-01:00 the next day):

[0250] a) Mechanical inspection: The crawler is in good appearance, the sealing performance is normal, and the track wear is within an acceptable range.

[0251] b) Electronic equipment inspection: The gravimeter, control chip, etc. are working normally and no abnormalities were found.

[0252] c) Data backup: All original data and processing results are backed up to three independent storage devices.

[0253] 3. Return flight (from 01:00 on June 3, 2024):

[0254] The survey ship "Ocean" set off to return to a port in Guangzhou and is expected to arrive in 6 hours.

[0255] Mission summary:

[0256] The seabed gravity detection mission in a certain sea area was successfully completed. By using advanced seabed crawlers and high-precision gravimeters, combined with innovative data processing methods, high-resolution gravity field data within the area of ​​100 square kilometers were obtained.

[0257] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A continuous gravity detection method based on a seafloor crawler, characterized in that: The detection of gravity data using a seafloor crawler equipped with an absolute gravimeter and a relative gravimeter includes the following steps: S10, dividing the seabed gravity measurement area into grids to obtain a plurality of grid units arranged in rows and columns; S20, controlling the seafloor crawler to traverse the center of each grid unit and obtain the relative gravity data of the center point of each grid unit, and establish a relative gravity matrix of the area to be measured; S30, calculating the similarity between each element in the relative gravity matrix and the preset reference relative gravity data as standard similarity, and calculating the similarity between adjacent elements as adjacent similarity; S40, using the Bayesian probability model and Markov chain Monte Carlo method to determine the abnormal points, boundary points and transition points of the standard similarity and adjacent similarities respectively: S50, clustering the abnormal points, boundary points and transition points to obtain a plurality of cluster centers and flying points; S60, establishing a traversal path between a plurality of cluster centers and flying points, controlling the seafloor crawler to crawl each cluster center and flying point and obtaining absolute gravity data of each cluster center and flying point; S70, based on the acquired absolute gravity data points and in combination with the previous relative gravity matrix, the Kriging interpolation method is used to interpolate the entire area to be measured, and finally the absolute gravity matrix of the area to be measured is obtained; The transition point refers to an area where the gravity value changes gradually, and the flying point refers to marking an isolated point in the clustering result or a point that does not meet the clustering standard as a flying point.

2. The continuous gravity detection method based on a seafloor crawler according to claim 1 is characterized in that: The side length of the grid unit is 50 meters to 500 meters.

3. The continuous gravity detection method based on a seafloor crawler according to claim 2 is characterized in that: The method of traversing the center of each grid cell is the traveling salesman problem algorithm.

4. The continuous gravity detection method based on a seafloor crawler according to claim 3 is characterized in that: The relative gravity data includes the coordinates of the measurement point, the measurement time, the relative gravity reading, and environmental parameters, including water depth, temperature, and pressure.

5. The continuous gravity detection method based on a seafloor crawler according to claim 4 is characterized in that: The similarity is calculated by using a weighted multi-similarity method, wherein the multi-similarity includes: Euclidean distance, cosine similarity, and Pearson correlation coefficient.

6. The continuous gravity detection method based on a seafloor crawler according to claim 5 is characterized in that: The method combining the Bayesian probability model and the Markov Chain Monte Carlo is adopted, specifically: the standard similarity and the adjacent similarity are used as observation data, it is assumed that the gravity field follows the Gaussian process distribution as prior knowledge, and the posterior distribution is sampled by the Markov Chain Monte Carlo method.

7. The continuous gravity detection method based on a seafloor crawler according to claim 6 is characterized in that: The method for clustering the outliers, boundary points and transition points is the DBSCAN algorithm.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program instructions, which, when executed, are used to execute the continuous gravity detection method based on a seafloor crawler as described in any one of claims 1 to 7.

9. A continuous gravity detection system based on a seafloor crawler, characterized in that: A survey vessel, a seabed crawler, a relative gravimeter, an absolute gravimeter and a control chip, wherein the survey vessel is connected to the seabed crawler via a towing cable or wirelessly, the seabed crawler is provided with the relative gravimeter and the absolute gravimeter, the seabed crawler is built with the control chip, and the control chip contains the computer-readable storage medium according to claim 8.

10. A continuous gravity detection system based on a seafloor crawler according to claim 9, characterized in that: The relative gravimeter uses one or more relative gravity gradiometers, and the absolute gravimeter uses one or more absolute gravity gradiometers.

Images