Underwater geological interpretation method and system based on multi-source data integration algorithm

Abstract

The invention relates to the technical field of water conservancy and hydropower engineering and shipping engineering, and discloses an underwater geological interpretation method based on a multi-source data integration algorithm, which comprises the following steps: acquiring underwater geological data, resistivity data, topographic data and water level data; constructing a relational expression between the relative resistivity and a stratum threshold value; calculating water depth data and the distance between the water depth data and the thalweg; selecting an ensemble learning algorithm to predict relative resistivity so as to realize stratum prediction; calculating the gradient of the underwater terrain to judge the stratum condition; and integrating an integrated learning algorithm stratum prediction result and an underwater topographic slope judgment stratum result to realize underwater geological interpretation. According to the method, the underwater geological interpretation is efficiently realized, the overwater operation safety risk of geological workers is reduced, the result is more reliable, the cost is saved, and a brand new thought is provided for the underwater geological interpretation.

Description

Technical Field

[0001] The present invention relates to the technical fields of water conservancy and hydropower engineering and shipping engineering, and particularly relates to an underwater geological interpretation method and system based on a multi-source data integration algorithm. Background Art

[0002] River channel siltation affects the smooth flow of water in water conservancy and hydropower projects downstream, hinders the normal operation of navigation, and has a greater adverse impact on the operation of the project. In recent years, underwater dredging has become the main means to solve this problem. However, using conventional drilling methods, not only the project cost is high, but also the operators are affected by the water flow and the operation space is limited, with a high safety risk. Moreover, the drilling can only reflect the geological conditions at a single drilling position, with great limitations. Using geophysical prospecting means can improve the efficiency of geological exploration, but the cost is high, and the geophysical prospecting accuracy is greatly affected by the flow velocity, and the interpretation level is greatly affected by technical personnel. Therefore, researchers are required to continuously innovate ideas and optimize methods to develop a more efficient, accurate and adaptable underwater geological interpretation system based on various sources of data.

[0003] In recent years, based on multi-source data analysis, basic data obtained from multiple channels can be fused. By selecting appropriate machine learning methods, a high-precision proxy model can be obtained to achieve target value prediction. Using multi-source data to predict the results can overcome the large errors caused by single-source data and maximize the improvement of prediction accuracy. Research results on multi-source data, such as highway construction early warning based on road data and vehicle data; the application of the lithosphere density inversion method proposed through residual gravity anomaly, residual geoid anomaly, and residual terrain anomaly; the integrated algorithm can fully utilize the advantages of individual weak learners to improve prediction accuracy. By establishing many weak estimators and then integrating the evaluation results of these weak evaluations according to a certain integration rule, a better effect than a single weak estimator can be achieved. For example, Adaboost optimizes the BLS model based on the influencing factors for water source discrimination to achieve water source discrimination; the concrete mix design method invented by integrated learning based on the "Standard for Mix Design of Ordinary Concrete" standard and existing data. Although machine learning research based on multi-source data has been applied in more and more fields. However, there is no machine learning determination method for underwater geological conditions. Summary of the Invention

[0004] To overcome the deficiencies of the above-mentioned prior art, the present invention provides an underwater geological interpretation method and system based on a multi-source data integration algorithm. By revealing the geological conditions through limited boreholes based on exploration methods, obtaining the resistivity of multiple survey lines using geophysical prospecting methods, and obtaining water depth data and underwater terrain data using RTK, single-beam sounding systems, and multi-beam sounding systems; fusing the data obtained by the three methods and combining with ensemble learning to achieve the prediction of relative resistivity, and finally realizing high-precision underwater geological interpretation, greatly improving the prediction accuracy and prediction efficiency.

[0005] According to one aspect of the specification of the present invention, there is provided an underwater geological interpretation method based on a multi-source data integration algorithm, including: Obtain underwater geological data, resistivity data, terrain data, and water level data; Based on the obtained underwater geological data and resistivity data, construct a relational expression between relative resistivity and formation threshold values, where the formation threshold values are critical values for different rock formations of silty alluvial layers, boulder alluvial layers, and hard rocks; Based on the obtained terrain data and water level data, calculate the water depth data and the distance from the thalweg line; Select a trained ensemble learning model to predict the relative resistivity and achieve formation prediction; the ensemble learning model uses the water depth, formation depth, and distance from the thalweg line as known quantities and the relative resistivity as the predicted value; Based on the predicted relative resistivity and combined with the relational expression between relative resistivity and formation threshold values, achieve formation prediction.

[0006] As a further technical solution, after obtaining the formation prediction result, it further includes: Calculate the underwater terrain slope and review the predicted formation according to the calculated underwater terrain slope.

[0007] As a further technical solution, obtaining underwater geological data, resistivity data, terrain data, and water level data includes: The underwater geological data is obtained using drilling equipment; the resistivity data is obtained using the high-density electrical method, including apparent resistivity; the terrain data and water level data are obtained using real-time kinematic positioning, single-beam sounding systems, or multi-beam sounding systems.

[0008] As a further technical solution, constructing the relational expression between relative resistivity and formation threshold values includes: Perform threshold division on the underwater geological data and apparent resistivity to construct a relational expression between relative resistivity and formation threshold values. Among them, according to the underwater geological data, the underwater terrain is divided into silty alluvial layers, boulder alluvial layers, and hard rock layers, and the relative resistivity is obtained by converting the apparent resistivity, and the corresponding relative resistivity and formation threshold values for different rock formations of silty alluvial layers, boulder alluvial layers, and hard rocks are obtained.

[0009] As a further technical solution, calculating water depth data and the distance from the thalweg line, including: Calculating water depth data: determined according to the difference between the water surface elevation measured by real-time differential positioning and the underwater topography obtained by a single-beam sounding system or a multi-beam sounding system; Calculating the distance from the thalweg line: based on the measured three-dimensional coordinate data of the topography, draw a cross-section perpendicular to the water flow direction, determine the thalweg points on the cross-section perpendicular to the water flow direction and form a complete thalweg line; introduce a geophysical exploration line based on the formed thalweg line, and the distance between the measuring point and the thalweg line is controlled by the two nearest thalweg points to the measuring point, and the straight-line distance perpendicular to the two points is the distance between the measuring point and the thalweg line.

[0010] As a further technical solution, obtaining the water depths of geophysical exploration points and borehole points, including: Obtain the vector elevation data of multiple points closest to the target point and assign different weights, and the weights are determined by the Euclidean distance from the target point. The coordinates of the target point are obtained by multiplying the coordinates of multiple points from the target point by the weights.

[0011] As a further technical solution, selecting a trained ensemble learning model to predict relative resistivity for formation prediction, including: Through the ensemble learning model, select different base learners to train the obtained multi-source data including water depth, formation depth, distance from the thalweg line, and relative resistivity, and select the three base learners with high scores as the base learners of the surrogate model; According to the selected base learners combined with the meta-learner, construct a surrogate model as the trained ensemble learning model; Use the trained ensemble learning model to predict relative resistivity, and according to the predicted value combined with the relationship between relative resistivity and formation threshold, obtain the predicted formation.

[0012] According to one aspect of the specification of the present invention, there is provided an underwater geological interpretation system based on a multi-source data integration algorithm, including: A first main module for obtaining underwater geological data, resistivity data, topographic data, and water level data; A second main module for constructing a relationship between relative resistivity and formation threshold based on the obtained underwater geological data and resistivity data, and the formation threshold is the critical value for different rock formations such as alluvial deposits, boulder alluvial deposits, and hard rocks; A third main module for calculating water depth data and the distance from the thalweg line based on the obtained topographic data and water level data; A fourth main module for selecting a trained ensemble learning model to predict relative resistivity to achieve formation prediction; the ensemble learning model uses water depth, formation depth, and distance from the thalweg line as known quantities and relative resistivity as the predicted value; The fifth main module is used to realize formation prediction based on the predicted relative resistivity and in combination with the relationship between the relative resistivity and the formation threshold.

[0013] According to one aspect of the specification of the present invention, there is provided an electronic device including a memory and a processor. The memory stores program instructions executed by the processor, and the processor calls the program instructions to execute the steps of the underwater geological interpretation method based on the multi-source data integration algorithm.

[0014] According to one aspect of the specification of the present invention, there is provided a non-transitory computer-readable storage medium. The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions cause the computer to execute the steps of the underwater geological interpretation method based on the multi-source data integration algorithm.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. In view of the problem that dredging is required in water conservancy and hydropower projects and waterway projects, and it is difficult to predict the underwater geological conditions comprehensively, at low cost and with high precision due to the influence of water bodies, the present invention proposes an underwater geological interpretation method based on a multi-source data integration algorithm. Specifically, first, underwater geological data are obtained through on-water drilling, resistivity data are obtained through geophysical prospecting, and topographic data and water level data are obtained through measurement; then, the relationship between the relative resistivity and the formation threshold is constructed to realize the division of the critical values of different rock formations such as the alluvial layer, boulder alluvial layer, and hard rock; then, the water depth data and the distance from the thalweg are calculated through the measured topography; then, the integrated learning algorithm is selected to use the water depth, formation depth, and distance from the thalweg as known quantities and the relative resistivity as the predicted value for training, and then the formation determination is realized through the set relationship between the relative resistivity and the formation threshold.

[0016] 2. The present invention also calculates the underwater topographic slope, and makes a formation determination according to the relationship between the underwater topographic slope and the underwater angle of repose of the gravel layer. Finally, the underwater geological interpretation is realized by integrating the formation prediction results of the integrated algorithm and the formation determination results of the underwater topographic slope.

[0017] 3. The present invention efficiently realizes the underwater geological determination through multi-source data fusion, reducing the construction risk of geological personnel. Therefore, the present invention provides a solid technical support for underwater geological interpretation and contributes to reducing the operation risk of geophysical prospecting personnel. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings used in the description of the embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative efforts.

[0019] Figure 1 It is a flowchart of an underwater geological interpretation method based on a multi-source data integration algorithm provided by an embodiment of the present invention.

[0020] Figure 2 It is a flowchart of an on-water drilling in an embodiment of the present invention.

[0021] Figure 3 It is a flowchart of a multi-beam sounding system measurement in an embodiment of the present invention.

[0022] Figure 4 It is a schematic diagram of the principle of a geophysical exploration method in an embodiment of the present invention.

[0023] Figure 5 It is a layout diagram of geophysical exploration lines in an embodiment of the present invention.

[0024] Figure 6 It is a flowchart of an ensemble learning in an embodiment of the present invention. Specific Embodiments

[0025] In view of the problems that dredging is required in water conservancy and hydropower projects and waterway projects, and it is difficult to predict the underwater geological conditions comprehensively, at low cost and with high precision due to the influence of water bodies, the present invention provides an underwater geological interpretation method based on a multi-source data integration algorithm. First, underwater geological data, resistivity data, terrain data and water level data are acquired; then, a relational expression between relative resistivity and formation threshold is constructed; subsequently, water depth data and the distance from the thalweg are calculated; finally, an ensemble learning algorithm is selected to predict the relative resistivity to achieve formation prediction and realize underwater geological interpretation. The present invention not only efficiently realizes underwater geological interpretation, reduces the safety risks of geological workers' on-water operations, but also has more reliable results and cost savings, providing a new idea for underwater geological interpretation.

[0026] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention. In addition, the technical features in each embodiment or a single embodiment provided by the present invention can be combined with each other arbitrarily to form a new technical solution, and this combination is not restricted by the order of steps and / or the pattern of structural composition, but must be based on the fact that those of ordinary skill in the art can implement it. When the combination of technical solutions results in contradictions or cannot be implemented, it should be considered that this combination of technical solutions does not exist and is not within the protection scope required by the present invention.

[0027] The following combines the attached Figures 1-6A further detailed description of the present invention will be given below.

[0028] An embodiment of the present invention discloses an underwater geological interpretation method based on a multi-source data integration algorithm. Referring to Figure 1 , the underwater geological interpretation method based on the multi-source data integration algorithm includes the following steps: S100. Obtain underwater geological data, resistivity data, terrain data, and water level data through different systems.

[0029] The underwater geological data refers to dividing the underwater terrain into sedimentary alluvial layers, boulder alluvial layers, and hard rock layers by comprehensively considering the fragmentation degree, weathering degree, and structural plane conditions of rock cores at different depths.

[0030] The underwater geological data is obtained by using underwater drilling equipment. Different drill bits are used according to different geological conditions. During the drilling process, different diameter casings are used for the water surface to the river bottom and the alluvial layer according to the actual situation.

[0031] The resistivity data is obtained by high-density electrical method. An artificial electric field is formed by supplying power to the ground through electrodes. The electrodes are arranged in an array with an electrode spacing of 5 - 15m, and the Wenner device is used to collect data. According to the detection depth requirements, each array collects data at the maximum detection depth. By scanning and measuring the artificial electric fields at different parts of the ground surface, apparent resistivity distribution data is obtained.

[0032] The terrain data and water level data are obtained through various methods such as RTK, single-beam sounding system, and multi-beam sounding system. For underwater surveys in areas with poor navigation conditions and shoal areas, a single-beam sounding system is used to obtain the three-dimensional data of the plane coordinates and elevation of the measurement points in the same coordinate system; for the deep-water areas of the reservoir, a multi-beam sounding system is used to conduct a fine three-dimensional scan of the underwater terrain to obtain high-density three-dimensional vector data of the plane and elevation of the underwater terrain; the water level data is obtained by measuring with RTK.

[0033] Specifically, it includes the following steps: S110. Obtain geological data on the distribution of different rock layers at the drilling points according to underwater drilling: Referring to Figure 2, the surface soil such as the silt layer can be obtained by a soil sampler with a larger diameter. For the alluvial layer with boulders and gravels, diamond or composite bit drilling should be adopted, and a larger-size casing should be used for isolation to ensure core sampling in the borehole. Finally, the borehole is terminated at a specified elevation with a smaller-size borehole diameter. A casing with a diameter of not less than ϕ273mm is used for isolation from the water surface to the river bottom, and casings with diameters of ϕ219mm, ϕ173mm, and ϕ146mm can be used for the alluvial layer; the final borehole diameters are ϕ110mm, ϕ91mm, and ϕ75mm. After core sampling and through ventilation and drying treatments, the stratification of different-depth strata is realized: the cores obtained by the drill rig are grouped and numbered, and the underwater rock masses obtained after ventilation and drying are divided into three types: silt alluvial layer, boulder alluvial layer, and hard rock layer according to the degree of fragmentation, combined with the degree of weathering, the difficulty of drilling, and the core recovery rate. Table 1 shows the depth conditions of different borehole strata.

[0034] Table 1 Depth Conditions of Different Borehole Strata

[0035] S120. Obtain resistivity data through the high-density resistivity method: Refer to Figure 4 and Figure 5 , arrange the array with an electrode spacing of 10m and collect data using the Wenner device. According to the detection depth requirements, each array collects data at the maximum detection depth. By scanning and measuring the artificial electric field at different parts of the ground surface, apparent resistivity distribution data are obtained. During the measurement, the electrode spacing AM = MN = NB = AB / 3 is one electrode spacing, and A, M, N, and B move synchronously to the right to obtain the profile line of the first-layer depth; then AM, MN, and NB increase by one electrode spacing, and A, M, N, and B move synchronously to the right to obtain the profile line of the second-layer depth.

[0036] S130. Obtain topographic data and water level data: The water surface line measurement and geophysical prospecting positioning can be carried out using real-time kinematic (RTK) positioning. For areas with poor navigation conditions and shoal areas, a single-beam sounding system is used for underwater measurement to obtain the three-dimensional data of the plane coordinates and elevations of the measurement points in the same coordinate system; for the deep-water area of the reservoir, a multi-beam sounding system is used to conduct a fine three-dimensional underwater topographic scan to obtain high-density three-dimensional vector data of the underwater topography in terms of plane and elevation.

[0037] The water depth and underwater elevation data of the geophysical prospecting measurement points and borehole points are obtained by acquiring the nearest 5 vector elevation data near the predicted points, and the weights are determined by the Euclidean distance S j from the target point. The greater the distance, the smaller the weight, and the weight . Therefore, the coordinates of the target point are obtained by multiplying the coordinates of the target point Y by the product of the weights. .

[0038] Refer toFigure 3 The use of a multi-beam bathymetric system to obtain underwater terrain must follow the conventional RTK operation process. After the RTK base station is set up, two or more control points are measured together, and the mobile station receiver is connected to the multi-beam bathymetric system. During the measurement process, the data of the navigation measurement system is output to the real-time acquisition system to record the position and motion posture of the vessel. At the same time, the navigation measurement system is set with parameters, network settings, and output settings.

[0039] The multi-beam bathymetric system uses a real-time data acquisition workstation system to monitor the data acquisition process, guide the ship to operate according to the arranged survey line, use software to control the working state of the sonar, and adjust the sonar's beam angle, range, central beam direction, transmission power, etc. according to the real-time reservoir environment. The ship attitude and transducer installation correction, including time delay correction, roll installation deflection correction, pitch installation deflection correction, bow installation deflection correction, etc., select flat underwater terrain areas and undulating areas respectively, and repeat multiple round-trip measurements. Under the premise of fully investigating the water environment of the reservoir survey area, the size of the beam angle is determined according to the average water depth to ensure that the overlap between the survey lines is not less than 10%.

[0040] Use data acquisition software to collect multi-beam field data. Use the real-time data acquisition workstation system to monitor the data collection process and guide the ship along the arranged survey line. According to changes in the reservoir environment, adjust the sonar's beam angle, range, central beam direction, transmission power, etc., and finally complete the field data collection work.

[0041] After the attitude correction parameters, sound velocity profile data, and tide level data are imported into the project using a multi-beam bathymetric system, the xtf format data obtained in the field is converted into HDCS format data and then the data is merged. A water depth data surface is established. Based on the merged water depth data, a water depth data surface is established, that is, a gridded underwater terrain data elevation model.

[0042] The water depth data collected in the field contains noise and wrong points caused by various error factors. These data are eliminated in the data post-processing process. Edit each survey line, filter the remaining large jump points manually combined with the parameter automatic method, and then filter the points outside the surface according to the water depth data surface.

[0043] S200, underwater geological conditions and resistivity threshold R Divide and form the result of division x Since the apparent resistivity fluctuates from single digits to thousands of digits, the relative resistivity is obtained by reducing the apparent resistivity by 100 times. , obtain different water drilling holes to reveal silt alluvial layer, boulder alluvial layer, hard rock layer and determine the corresponding threshold for different rock layers , , denoted as Pairwise combine n groups of different result data to construct a second-order determinant. After deleting the data with the largest deviation from 0 in the second-order determinant, perform a weighted average on the remaining thresholds to obtain the corresponding thresholds for different rock layers of the final silt alluvial layer, boulder alluvial layer, and hard rock layer, and round them to 1 and 5, as shown in Table 2.

[0044] It should be noted that the threshold here can be regarded as the measurement boundary value or the resistivity at the boundary between two rock layers.

[0045] Table 2 Corresponding thresholds determined for different rock layers

[0046] S300. Calculate the water depth data at the measuring point position and the distance from the thalweg. The water depth data H 1 is determined by the difference between the water surface elevation measured by RTK and the underwater topography obtained by single-beam sounding or multi-beam sounding. The distance from the thalweg L Import the measured three-dimensional coordinate data of the topography into the mapping software, draw a cross-section perpendicular to the water flow direction through the river direction. The lowest point on the cross-section is the thalweg point. Connect the thalweg points of all cross-sections to form a complete thalweg line. Draw the geophysical exploration line into the mapping software. The distance between the measuring point and the thalweg line is controlled by the two thalweg line control points closest to the measuring point. The straight-line distance perpendicular to the line determined by these two points is L the distance between the measuring point and the thalweg line.

[0047] S400. Select an ensemble learning model to predict the relative resistivity to achieve formation prediction. The ensemble learning model uses the water depth, formation depth, and distance from the thalweg as known quantities and the relative resistivity as the predicted value.

[0048] Refer to Figure 6 , select different base learners to train on the obtained water depth H 1. Formation depth H 2. Distance from the thalweg L , relative resistivity R ’ and other multi-source data, and select the three base learners with high scores as the base learners of the surrogate model. According to the obtained water depth H 1. Formation depth H 2. Distance from the thalweg L and other multi-source data as independent variables, measure the relative resistivity R'As the dependent variable, GBDT, ET, RF, ADA, XGB, and SVR are selected as the base models, and logical regression and linear regression are used as the meta-models for optimization. A proxy model based on Stacking is constructed. Taking MAE as the standard to measure the model accuracy, Bayesian Optimization is performed on different selected base learner pairs. Selecting MAE as the standard to measure the model accuracy, the three models with the highest optimized prediction rate are implemented. In this embodiment, GBDT, ET, and RF, three different base learners, are finally selected, and the Linear Regression model is selected as the meta-model. The calculation results and MAE are shown in Tables 3 and 4.

[0049] Table 3 Errors corresponding to different base learners and meta-models after Bayesian optimization of parameters

[0050] Table 4 Errors of calculation results using Stacking and optimized base learners and meta-models

[0051] Furthermore, the formation is predicted through the proxy model and the selected threshold. The formation prediction is realized through the corresponding relationship between the predicted resistivity result and the threshold, which is the regression analysis result of the silty alluvial layer, boulder alluvial layer, and hard rock layer at the predicted location based on ensemble learning.

[0052] To evaluate the prediction accuracy, the following steps are also included: S500. Calculate the underwater terrain slope to determine the formation situation: Further, through the obtained terrain vector data, slope analysis is performed on the terrain around the prediction point. The slope analysis selects measurement points within a 10m diameter range. Through linear interpolation, contour lines are obtained, and the comprehensive slope perpendicular to the contour lines within the selected range is calculated. If the slope is equal to or greater than the underwater angle of repose of the gravel layer, which is 32°, it is determined as bedrock, serving as the determination of whether the underwater geological surface is a hard rock layer.

[0053] S600. Calculate the underwater terrain slope to determine the formation situation: The underwater geological interpretation is realized based on the geological interpretation results predicted by the ensemble learning model and the underwater terrain slope.

[0054] In the embodiment of the present invention, the MAE of the underwater geological interpretation method based on the multi-source data integration algorithm is 0.56.

[0055] It can be seen that for the underwater geological interpretation method of the multi-source data integration algorithm of the present invention, the method determined by the relative resistivity and topographic slope based on multi-source data using the integration algorithm has an error within 1, and can completely perform different underground geological interpretations, showing good accuracy and efficiency, saving work investment, reducing the construction risks of geological personnel. The present invention conforms to geophysical exploration research means and enhances the scientificity and credibility of design decisions. It is very helpful for the geological interpretation of water conservancy and hydropower projects and dredging projects involved in waterway projects. Through precise design and interpretation, the present invention improves the accuracy and efficiency of geological interpretation and provides strong technical support for underwater geological determination.

[0056] The implementation basis of each embodiment of the present invention is achieved through programmed processing by a device with a processor function. Therefore, in engineering practice, the technical solutions and functions of each embodiment of the present invention are encapsulated into various modules. Based on this actual situation, on the basis of the above embodiments, an embodiment of the present invention provides an underwater geological interpretation system based on a multi-source data integration algorithm, and this system is used to execute the underwater geological interpretation method based on the multi-source data integration algorithm in the above method embodiments.

[0057] This system includes: a first main module for acquiring underwater geological data, resistivity data, topographic data, and water level data; a second main module for constructing a relative resistivity and formation threshold relationship based on the acquired underwater geological data and resistivity data, where the formation threshold is the critical value for different rock layers such as alluvial deposits, boulder alluvial deposits, and hard rocks; a third main module for calculating water depth data and the distance from the thalweg based on the acquired topographic data and water level data; a fourth main module for selecting a trained integrated learning model to predict the relative resistivity and achieve formation prediction; the integrated learning model uses water depth, formation depth, and the distance from the thalweg as known quantities and the relative resistivity as the predicted value; a fifth main module for achieving formation prediction based on the predicted relative resistivity and combining the relative resistivity and formation threshold relationship.

[0058] The underwater geological interpretation system based on the multi-source data integration algorithm provided by the embodiment of the present invention, aiming at the problem of the deficiency that underwater geological conditions are difficult to predict comprehensively, at low cost, and with high accuracy due to the influence of water bodies in water conservancy and hydropower projects and waterway projects that require dredging, adopts the foregoing several modules and efficiently realizes the underwater geological determination situation through multi-source data fusion, reducing the construction risks of geological personnel.

[0059] It should be noted that the system embodiments provided by the present invention are used not only to implement the methods in the above method embodiments, but also to implement the methods in other method embodiments provided by the present invention. The difference is only in setting corresponding functional modules, and its principle is basically the same as that of the above system embodiments provided by the present invention. As long as those skilled in the art, on the basis of the above system embodiments, refer to the specific technical solutions in other method embodiments, obtain corresponding technical means by combining technical features, and the technical solutions constituted by these technical means, and improve the modules in the above system embodiments on the premise of ensuring the practicability of the technical solutions, corresponding system-like embodiments can be obtained to implement the methods in other method-like embodiments.

[0060] Based on the same inventive concept as the foregoing embodiments, an embodiment of the present invention further provides an electronic device, including a memory and a processor. The memory stores program instructions executed by the processor, and the processor invokes the program instructions to execute the steps of the underwater geological interpretation method based on the multi-source data integration algorithm.

[0061] Based on the same inventive concept as the foregoing embodiments, an embodiment of the present invention further provides a non-transitory computer-readable storage medium. The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions cause the computer to execute the steps of the underwater geological interpretation method based on the multi-source data integration algorithm.

[0062] In summary of the above embodiments, the present invention obtains underwater geological data, resistivity data, terrain data, and water level data, constructs a relative resistivity and formation threshold relationship formula, calculates the water depth data and the distance from the thalweg line, and selects an integrated learning algorithm to predict the relative resistivity to achieve formation prediction. And it also determines the formation situation by calculating the underwater terrain slope, and comprehensively realizes underwater geological interpretation by integrating the formation prediction results of the integrated learning algorithm and the formation determination results of the underwater terrain slope. The present invention not only efficiently realizes underwater geological interpretation, reduces the safety risk of geological workers' underwater operations, but also has more reliable results and saves costs, providing a new idea for underwater geological interpretation.

[0063] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the technical solutions of the embodiments of the present invention.

Claims

1. An underwater geological interpretation method based on a multi-source data integration algorithm, characterized in that include: Obtain underwater geological data, resistivity data, topographic data and water level data; Based on the acquired underwater geological data and resistivity data, a relationship between relative resistivity and formation threshold is constructed, wherein the formation threshold is a critical value for different rock layers including silt alluvial layer, boulder alluvial layer and hard rock; Calculate water depth data and distance to the deep well line based on the acquired terrain data and water level data; Selecting the trained integrated learning model to predict relative resistivity to achieve formation prediction; the integrated learning model uses water depth, formation depth and distance from the deep line as known quantities, and the relative resistivity is a predicted value; Based on the predicted relative resistivity and the relationship between relative resistivity and formation threshold, formation prediction is achieved.

2. The underwater geological interpretation method based on the multi-source data integration algorithm according to claim 1, characterized in that, After obtaining the formation prediction results, it also includes: Calculate the underwater terrain slope and review the predicted strata based on the calculated underwater terrain slope.

3. The underwater geological interpretation method based on the multi-source data integration algorithm according to claim 1, characterized in that Acquire underwater geological data, resistivity data, topographic data and water level data, including: The underwater geological data is obtained by drilling equipment; the resistivity data is obtained by high-density electrical method, including apparent resistivity; the terrain data and water level data are obtained by real-time differential positioning, single-beam sounding system or multi-beam sounding system.

4. The underwater geological interpretation method based on the multi-source data integration algorithm according to claim 3, characterized in that, Construct the relationship between relative resistivity and formation threshold, including: The underwater geological data and apparent resistivity are divided into thresholds, and a relationship between relative resistivity and formation threshold is constructed, wherein the underwater terrain is divided into siltation alluvial layer, boulder alluvial layer and hard rock layer according to the underwater geological data, and the relative resistivity is obtained according to the apparent resistivity conversion, and the relative resistivity and formation threshold corresponding to different rock layers of the siltation alluvial layer, boulder alluvial layer and hard rock layer are obtained.

5. The underwater geological interpretation method based on the multi-source data integration algorithm according to claim 3, characterized in that Calculate water depth data and distance to the deep line, including: Calculate water depth data: Determine the difference between the water surface elevation measured by real-time differential positioning and the underwater topography obtained by the single-beam bathymetry system or the multi-beam bathymetry system; Calculate the distance to the abyss line: Draw a cross section perpendicular to the water flow direction based on the measured three-dimensional coordinate data of the terrain, determine the abyss points on the cross section perpendicular to the water flow direction and form a complete abyss line; introduce the physical detection line based on the formed abyss line, and the distance between the measuring point and the abyss line is controlled by the two abyss points closest to the measuring point. The straight-line distance determined perpendicular to these two points is the distance between the measuring point and the abyss line.

6. The underwater geological interpretation method based on the multi-source data integration algorithm according to claim 1, wherein Obtaining water depth at geophysical detection points and drilling points, including: The vector elevation data of multiple points closest to the target point are obtained and assigned different weights. The weight is determined by the Euclidean distance from the target point. The coordinates of the target point are obtained by multiplying the coordinates of multiple points from the target point by the weight.

7. The underwater geological interpretation method based on the multi-source data integration algorithm according to claim 1, characterized in that Select the trained ensemble learning model to predict relative resistivity to achieve formation prediction, including: Different base learners are selected through the integrated learning model to train the multi-source data including water depth, formation depth, distance to the deep line and relative resistivity, and three base learners with high scores are selected as the base learners of the proxy model; According to the selected base learner combined with the meta learner, a proxy model is constructed as a trained ensemble learning model; The trained integrated learning model is used to predict relative resistivity, and the predicted formation is obtained based on the predicted value combined with the relationship between relative resistivity and formation threshold.

8. An underwater geological interpretation system based on a multi-source data integration algorithm, characterized in that, include: The first main module is used to obtain underwater geological data, resistivity data, terrain data, and water level data; The second main module is used to construct a relational expression between relative resistivity and formation threshold based on the obtained underwater geological data and resistivity data, where the formation threshold is the critical value for different rock layers of silty alluvial layer, boulder alluvial layer, and hard rock; The third main module is used to calculate water depth data and the distance from the thalweg based on the obtained terrain data and water level data; The fourth main module is used to select a trained ensemble learning model to predict relative resistivity and achieve formation prediction; the ensemble learning model uses water depth, formation depth, and the distance from the thalweg as known quantities and relative resistivity as the predicted value; The fifth main module is used to achieve formation prediction based on the predicted relative resistivity and in combination with the relational expression between relative resistivity and formation threshold.

9. An electronic device, characterized in that, It includes a memory and a processor. The memory stores program instructions executed by the processor, and the processor calls the program instructions to execute the steps of the underwater geological interpretation method based on the multi-source data integration algorithm according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions cause the computer to execute the steps of the underwater geological interpretation method based on the multi-source data integration algorithm according to any one of claims 1 to 7.

Images