Geothermal resource evaluation method, device and equipment based on machine learning

Abstract

The invention relates to the field of geothermal geology, in particular to a geothermal resource evaluation method, device and equipment based on machine learning. Comprising the following steps: acquiring geological data of a plurality of observation points selected in a first geothermal exploration area; selecting a second geothermal exploration area according to the geological data of a plurality of observation points in the first geothermal exploration area; acquiring a data set of a plurality of well drilling points in the second geothermal exploration area; training a geothermal parameter prediction model and a geothermal resource evaluation model according to the data sets of the plurality of well drilling points; and obtaining a geothermal resource evaluation result of the first geothermal exploration area by using the geothermal parameter prediction model and the geothermal resource evaluation model obtained by training according to the geological data of the plurality of observation points. According to the embodiment of the invention, the ability of predicting geothermal parameters and evaluating geothermal resources according to geological data is learned by using the machine learning model, the accuracy of geothermal resource exploration under the condition of lack of drilling data is improved, and the method has important significance for delineation and division of favorable areas of geothermal resource exploration.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of geothermal geology, and specifically to a geothermal resource evaluation method, device and equipment based on machine learning. Background Art

[0002] The current geothermal field reflects the energy balance between various dynamic processes inside the earth and is the basis for regional geothermal resource potential exploration and evaluation. The most reliable method for obtaining the current geothermal field is direct measurement, that is, direct measurement of underground temperature through drilling, so as to calculate geothermal distribution, geothermal gradient, geothermal heat flow and other geothermal parameters, construct the current regional geothermal field, and evaluate geothermal resources. However, due to the uneven distribution of drilling, not all geothermal exploration areas have sufficient deep drilling data for geothermal resource exploration.

[0003] In the related technology, for geothermal exploration areas lacking drilling data, most scholars directly estimate the deep temperature of the strata in the geothermal exploration areas through various indirect factors and subjective tendencies.

[0004] The above methods often deviate greatly from the actual situation due to incomplete understanding of the underground characteristics of the geothermal exploration area, and the accuracy is often unsatisfactory. Therefore, how to carry out geothermal resource exploration with high accuracy in the absence of drilling data is a key issue that needs to be urgently solved in the delineation and division of geothermal resource advantage areas. Summary of the invention

[0005] The purpose of the embodiments of this specification is to provide a geothermal resource evaluation method, device and equipment based on machine learning to improve the accuracy of geothermal resource exploration in the absence of drilling data.

[0006] On the one hand, an embodiment of the present specification proposes a geothermal resource evaluation method based on machine learning, the method comprising: obtaining geological data of multiple observation points selected in a first geothermal exploration area; selecting a second geothermal exploration area based on the geological data of multiple observation points in the first geothermal exploration area; obtaining a data set of multiple drilling points in the second geothermal exploration area, the data set of the multiple drilling points including the geological data of the drilling points, the geothermal parameters of the drilling points and the geothermal resource evaluation labels of the drilling points; training a geothermal parameter prediction model and a geothermal resource evaluation model based on the data set of the multiple drilling points; and obtaining a geothermal resource evaluation result of the first geothermal exploration area using the trained geothermal parameter prediction model and the geothermal resource evaluation model based on the geological data of the multiple observation points.

[0007] On the other hand, a geothermal resource evaluation device based on machine learning includes: a first acquisition module, used to acquire geological data of multiple observation points selected in a first geothermal exploration area; a selection module, used to select a second geothermal exploration area based on the geological data of multiple observation points in the first geothermal exploration area; a second acquisition module, used to acquire data sets of multiple drilling points in the second geothermal exploration area, the data sets of the multiple drilling points including the geological data of the drilling points, the geothermal parameters of the drilling points and the geothermal resource evaluation labels of the drilling points; a training module, used to train a geothermal parameter prediction model and a geothermal resource evaluation model based on the data sets of the multiple drilling points; and a third acquisition module, used to acquire the geothermal resource evaluation results of the first geothermal exploration area based on the geological data of the multiple observation points using the trained geothermal parameter prediction model and the geothermal resource evaluation model.

[0008] On the other hand, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the above-mentioned geothermal resource evaluation method based on machine learning.

[0009] Finally, an embodiment of the present invention also provides a computer storage medium on which computer instructions are stored. When the computer instructions are executed by a processor of a computer device, the above-mentioned geothermal resource evaluation method based on machine learning is executed.

[0010] It can be seen from the technical solutions provided by the above embodiments of this specification that the embodiments of this specification obtain geological data of multiple observation points selected in the first geothermal exploration area; select the second geothermal exploration area according to the geological data of multiple observation points in the first geothermal exploration area; obtain data sets of multiple drilling points in the second geothermal exploration area, and the data sets of the multiple drilling points include geological data of the drilling points, geothermal parameters of the drilling points, and geothermal resource evaluation labels of the drilling points; train geothermal parameter prediction models and geothermal resource evaluation models according to the data sets of the multiple drilling points; and obtain geothermal resource evaluation results of the first geothermal exploration area using the trained geothermal parameter prediction models and geothermal resource evaluation models according to the geological data of the multiple observation points. Compared with the existing methods, the learning and induction capabilities of the machine learning model for data can better and more objectively evaluate the geothermal resources of the geothermal exploration area that lacks drilling data, thereby improving the accuracy of geothermal exploration under the condition of lacking drilling data. The technical solutions of the embodiments of this specification are of great significance for the delineation and division of favorable areas for geothermal resource exploration under the condition of lacking drilling data. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of this specification or the prior art, the drawings required for use in the embodiments or the prior art description are briefly introduced below.

[0012] Figure 1 is a flow chart of a geothermal resource evaluation method based on machine learning provided in an embodiment of this specification;

[0013] Figure 2 It is a topographic map of region A provided in the embodiments of this specification;

[0014] Figure 3 It is the drilling distribution in area B provided in the embodiment of this specification;

[0015] Figure 4 It is a structural unit map of region B provided in the embodiments of this specification;

[0016] Figure 5 It is a pattern diagram of predicting geothermal parameters using a geothermal parameter prediction model provided in an embodiment of this specification;

[0017] Figure 6 It is a pattern diagram for predicting the quality of geothermal resources by using the geothermal resource evaluation model provided in the embodiment of this specification;

[0018] Figure 7 It is the geothermal resource advantage area in region A predicted by the comprehensive use of the geothermal parameter prediction model and the geothermal resource evaluation model provided in the embodiments of this specification;

[0019] Figure 8 It is a schematic diagram of the structure of a geothermal resource evaluation device based on machine learning provided in an embodiment of this specification;

[0020] Fig. 9 It is a schematic diagram of the structural composition of a computer device provided in an embodiment of this specification. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of this specification to clearly and completely describe the technical solutions in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0022] Figure 1 The following is a flow chart of the geothermal resource evaluation method. When implemented, the method includes the following steps:

[0023] S101: Acquire geological data of a plurality of observation points selected in a first geothermal exploration area.

[0024] In some embodiments, the first geothermal exploration area is an exploration area for which geothermal resource evaluation is to be conducted. The first geothermal exploration area refers to a geothermal exploration area that lacks drilling information, has no geothermal exploration data, or has a low degree of geothermal exploration but has good geothermal resources. Figure 2 As shown in Region A, Region A has an active volcanic system and there are many volcanoes in the region. At the same time, there are a large number of hot spring outcrops and a small number of geothermal drillings in Region A. Region A shows good geothermal resources, but due to the scarce drilling data, the distribution of the geothermal resource advantage area in Region A has always lacked a clear definition. The observation points are several points selected in the first geothermal exploration area according to the spatial distribution. For example, multiple points can be evenly selected as observation points in the first geothermal exploration area. The geological data of the multiple observation points can be obtained so as to evaluate the geothermal resources in the first geothermal exploration area using the geological data of the multiple observation points.

[0025] The geological data of the observation point can be collected by drilling at the observation point.

[0026] Based on the selected multiple observation points, geological data of each observation point that has a certain correlation with geothermal resource exploration and evaluation can be collected. The geological data at least includes: Curie depth, Moho depth, aeromagnetic anomaly data, Bouguer gravity anomaly data, altitude data, longitude and latitude, and distance from other faults.

[0027] Among them, the Curie point depth indicates the depth at which ferromagnetic minerals in crustal rocks become paramagnetic due to the temperature reaching the Curie point. The Curie point depth is usually closely related to the geothermal gradient; the Moho surface depth indicates the depth of the interface between the crust and the mantle. The Moho surface is a sudden change boundary, which marks a change in chemical composition and crystal structure; the aeromagnetic anomaly data indicates the relationship between the anisotropy vector of regional rock magnetic susceptibility and the earth's magnetic field; the Bouguer gravity anomaly data indicates the gravity difference data obtained by subtracting the latitude correction, altitude correction, Bouguer correction, and terrain correction based on the observation results of the gravimeter, and then subtracting the normal gravity value; the altitude data indicates the surface morphology information; the longitude and latitude indicate the location information; the distance to other faults indicates the closest distance to the seismogenic fault. These geological data can be specifically selected from a large amount of basic geological data to select data that are considered to have a significant correlation with the quality of geothermal resources and some data that may be correlated with the quality of geothermal resources. The more types of geological data that can be obtained, the higher the accuracy of the model obtained by the later training model. These geological data can be obtained through literature research and public related databases.

[0028] S102: Selecting a second geothermal exploration area based on geological data of multiple observation points in the first geothermal exploration area.

[0029] In some embodiments, an area where the number of drilling points is greater than or equal to a set threshold and the similarity between the corresponding geological background and the geological background of the first geothermal exploration area meets the similarity condition can be selected as the second geothermal exploration area. The number of drilling points is greater than or equal to the set threshold, so that the second geothermal exploration area has sufficient drilling point data. The similarity between the corresponding geological background and the geological background of the first geothermal exploration area meets the similarity condition, so that the geological background of the second geothermal exploration area is similar to that of the first geothermal exploration area. In this way, the training effect of the geothermal parameter prediction model and the geothermal resource evaluation model can be improved, thereby improving the accuracy of the geothermal resource evaluation of the first geothermal exploration area. For example, a geothermal exploration area set can be provided. The geothermal exploration area set includes multiple geothermal exploration areas. Each geothermal exploration area corresponds to the number of drilling points and has geological data. Then, one or more candidate geothermal exploration areas with the number of drilling points greater than or equal to the set threshold can be selected from the geothermal exploration area set. For each candidate geothermal exploration area, the similarity between the geological background of the candidate geothermal exploration area and the first geothermal exploration area can be calculated based on the geological data of the candidate geothermal exploration area and the first geothermal exploration area. A candidate geothermal exploration area with a similarity greater than or equal to a set threshold can be selected from the one or more candidate geothermal exploration areas as the second geothermal exploration area.

[0030] In some embodiments, the geological background refers to the geological conditions of the region, including strata, folds, faults, geochemistry, geophysics, etc., reflecting the geological evolution history, geological structural characteristics, rock distribution and geological action of the region, as well as the influences that contribute to the geological environment and distribution of mineral resources in the region. Among them, strata refer to a layer or a group of rock layers in the region with certain unified characteristics and properties and clearly distinguished from the upper and lower layers; folds refer to the rock layers that make up the crust of the region, which are subjected to strong tectonic stress and are bent in a wavy manner without losing their continuity; faults refer to the structure of the crust of the region being fractured by stress, and the rocks on both sides of the fracture surface undergoing significant relative displacement; geochemistry refers to the composition of elements and their isotopes in the geological body of the region, as well as the balance and cycle of elements, and the laws of change in time and space; geophysics refers to the distribution and changes of various physical fields in the region, as well as the medium structure, material composition, formation and evolution of geology and their laws. Therefore, the similarity between geological backgrounds refers to the similarity in the above-mentioned strata, folds, faults, geochemistry, geophysics, etc. More specifically, it refers to the similarity of information such as geological evolution history, geological structural characteristics, stratigraphic distribution and geological action. The second geothermal exploration area refers to an area where the number of drilling points is greater than or equal to the set threshold, that is, an area with sufficient drilling data, a high degree of geothermal exploration, and a similarity with the geological background of the first geothermal exploration area greater than or equal to the set threshold. For example, area B has similar information to area A in terms of geological evolution history, geological structural characteristics, stratigraphic distribution and geological action, and area B has a high degree of geothermal exploration and rich relevant geothermal data. Figure 3 As shown. Among them, Figure 3 Each solid circle in the figure represents a geothermal drilling site in area B.

[0031] S103: Acquire data sets of multiple drilling points in the second geothermal exploration area, where the data sets of the multiple drilling points include geological data of the drilling points, geothermal parameters of the drilling points, and geothermal resource evaluation labels of the drilling points.

[0032] In some embodiments, a drilling point refers to a location in the second geothermal exploration area where a well is drilled and has drilling data. Geological data, geothermal parameters, and geothermal resource evaluation labels related to these drilling points can be collected. These geological data, geothermal parameters, and geothermal resource evaluation labels are digitized to establish a data set; the data set contains multiple groups of data, each group of data corresponds to an observation point, and each group of data includes geological data, geothermal parameters, and geothermal resource evaluation labels; the geological data at least includes: Curie depth, Moho depth, aeromagnetic anomaly data, Bouguer gravity anomaly data, altitude data, longitude and latitude, and distance from other faults; the geothermal parameters at least include: geothermal gradient, geothermal heat flow, and 150°C burial depth; the geothermal resource evaluation label can be set to 0 and 1, indicating the quality of geothermal resources, respectively, or it can be set to multiple more detailed and accurate levels, which will not be repeated here.

[0033] Among them, the geothermal gradient is a parameter that represents the degree of uneven temperature distribution inside the earth. Generally, the deeper the burial depth, the higher the temperature value; the geothermal heat flow represents the heat transmitted from the interior of the earth to the surface; the burial depth of 150°C represents the depth at which the ground temperature reaches 150°C. These geothermal parameters are usually data closely related to the quality of geothermal resources selected from drilling data. The more types of geothermal parameters that can be obtained, the higher the model accuracy obtained by the later training model. The second geothermal exploration area usually has sufficient geothermal parameters to characterize the deep temperature characteristics of the region, and then evaluate the quality of geothermal resources. However, the above-mentioned first geothermal exploration area is mostly limited by the distribution of drilling wells and lacks sufficient geothermal parameters, so it is impossible to directly evaluate the quality of geothermal resources through geothermal parameters.

[0034] In some embodiments, the process of digitizing the above geological data, geothermal parameters and geothermal resource evaluation labels is specifically divided into two categories: the first category is for data that can be expressed in the form of data itself. For example, the longitude and latitude in geological data, Curie depth, Moho depth, distance from other faults and geothermal gradient, geothermal heat flow and 150°C burial depth in geothermal parameters need to be collected and uniformly converted into Arabic numeral units under common units, such as converting the longitude and latitude minutes and seconds notation into decimal notation (128°3'39", 42°3'12" into 128.0608, 42.0533); and the data is unified into common international units, such as converting the Curie depth, Moho depth, distance from other faults, 150°C burial depth and other length units into m as units). The second category is for data that is difficult to express in the form of data. For example, terrain and geothermal resource quality labels are usually divided into two or more types. For example, the terrain of the study area is divided into plains, hills, mountains, etc., and then numbers are used to represent each type, with 1 representing plains, 2 representing hills, 3 representing mountains, etc. Another example is to divide the geothermal resource quality labels into good and bad geothermal resources, with 0 representing good geothermal resources and 1 representing bad geothermal resources.

[0035] S104: Training a geothermal parameter prediction model and a geothermal resource evaluation model according to the data sets of the multiple drilling points.

[0036] In some embodiments, a geothermal parameter prediction model can be trained based on the geological data and geothermal parameters in the data sets of the multiple drilling points. For example, for a drilling point, the geological data in the data set of the drilling point can be input into a pre-constructed geothermal parameter prediction model, and the model parameters of the geothermal parameter prediction model can be adjusted based on the output of the geothermal parameter prediction model and the geothermal parameters in the data set of the drilling point. A geothermal resource evaluation model can be trained based on the geothermal parameters and geothermal resource evaluation labels in the data sets of the multiple drilling points. For example, for a drilling point, the geothermal parameters in the data set of the drilling point can be input into a pre-constructed geothermal resource evaluation model, and the model parameters of the geothermal resource evaluation model can be adjusted based on the output of the geothermal resource evaluation model and the geothermal resource evaluation labels in the data set of the drilling point.

[0037] In some embodiments, the second geothermal exploration area can be divided into a training area and a verification area, the training area includes multiple first drilling points, and the verification area includes multiple second drilling points; the geothermal parameter prediction model can be trained based on the geological data and geothermal parameters in the data sets of the multiple first drilling points; the geothermal resource evaluation model can be trained based on the geothermal parameters and geothermal resource evaluation labels in the data sets of the multiple first drilling points.

[0038] In some embodiments, the second geothermal exploration area can be divided into a training area and a verification area according to the geological unit division information of the second geothermal exploration area, such as Figure 4 As shown. You can also use methods such as the holdout method and the K-fold cross-validation method for division. The holdout method divides the data set into two mutually exclusive sets. In order to keep the consistency of the data distribution as much as possible and avoid introducing additional deviations due to differences in distribution, stratified sampling can be used. At the same time, due to the randomness of the division, the results of a single holdout method are often not stable enough. You can use several random divisions and repeat the experiment to take the average value. The K-fold cross-validation method divides the data set into K mutually exclusive subsets of the same size. Similarly, it keeps the consistency of the data distribution as much as possible and uses stratified sampling to obtain these subsets. The idea of ​​the K-fold cross-validation method is: each time, the union of K-1 subsets is used as the training set, and the remaining subset is used as the validation set. In this way, there are K types of training set / test set divisions, so that K training and testing can be performed, and finally the mean of the K test results is returned. For the division ratio of the training set and the validation set, it is generally set to 2:1~4:1. When the total data volume is small, the ratio is reduced, and vice versa. Increase the ratio.

[0039] In some embodiments, the geothermal parameter prediction model adopts a multi-input and multi-output machine learning model, and the input features of the geothermal parameter prediction model include at least: Curie depth, Moho depth, aeromagnetic anomaly data, Bouguer gravity anomaly data, altitude data, longitude and latitude, and distance from other faults; the geothermal parameter prediction model can be used to predict the geothermal parameters of the training area, and the deviation between the obtained prediction results and the measured results is used to train the geothermal parameter prediction model, wherein the predicted geothermal parameters include at least: geothermal gradient, geothermal heat flow, and 150°C burial depth. The geothermal resource evaluation model adopts a multi-input and single-output machine learning model, and the input features of the geothermal resource evaluation model include at least: geothermal gradient, geothermal heat flow, and 150°C burial depth; the geothermal resource evaluation model can be used to predict the quality of geothermal resources in the training area, and the deviation between the obtained prediction results and the measured results is used to train the geothermal resource evaluation model.

[0040] When any of the machine learning models in the above geothermal parameter prediction model and geothermal resource evaluation model is trained, the learning rate update process is as follows:

[0041] During the training process, gradient descent is used to update the machine learning model parameters according to the gradient descent algorithm of the Newton-Leibniz formula. The calculation formula is:

[0042]

[0043] Where ωj represents the jth parameter of the machine learning model, ωj' is the corresponding updated parameter; λ represents the learning rate of the machine learning model, which is used to determine the convergence rate of the machine learning model; It represents the partial derivative of the loss function F(ωj) with respect to ωj when the deviation between the predicted result and the actual result of the machine learning model is known. The deviation between the predicted result and the actual result can be calculated by the loss function; the loss function is usually defined based on the sum of squared errors between the predicted result and the actual result or the absolute value of the error, logarithmic loss, and other methods. The machine learning model adjusts the values ​​of each parameter in the model by the deviation between the predicted result and the actual result, so that the deviation becomes smaller and smaller. The machine learning model used in the geothermal parameter prediction model and the geothermal resource evaluation model can specifically adopt models such as logistic regression, KNN, decision tree, DNN, support vector machine, XGBoost, and random forest.

[0044] In some embodiments, the learning rate can be set to a fixed value, or strategies such as learning rate decay, adaptive learning rate, and polynomial learning rate can be used. The idea of ​​learning rate decay is to gradually reduce the learning rate as training progresses to improve the stability and generalization of model training. Adaptive learning rate is a type of strategy that can automatically adjust the learning rate. It can dynamically adjust the learning rate according to the gradient information of the parameters, thereby improving the training speed and performance of the model. The idea of ​​polynomial decay is to decay the learning rate through a polynomial function, thereby gradually reducing the learning rate during training. It can help the model converge better in the later stages of training and improve the generalization ability of the model.

[0045] In some embodiments, the trained geothermal parameter prediction model and geothermal resource evaluation model can be tested based on the data sets of the multiple second drilling points to obtain geothermal parameters output by the geothermal parameter prediction model and geothermal resource evaluation results output by the geothermal resource evaluation model; abnormal second drilling points can be eliminated from the multiple second drilling points based on the output geothermal parameters; abnormal second drilling points can continue to be eliminated from the remaining second drilling points based on the output geothermal resource evaluation results; and the trained geothermal parameter prediction model and geothermal resource evaluation model can be optimized based on the data sets of the remaining second drilling points.

[0046] By removing abnormal second drilling points from multiple second drilling points twice, the influence of abnormal geological data and geothermal parameters on model training is effectively avoided. These abnormal geological data and geothermal parameters can make the model unable to accurately fit the real data distribution, affecting the effect of the actual application of the model. Therefore, removing abnormal second drilling points can make the model fit the real data distribution more accurately, and then obtain more refined geothermal resource evaluation results in the first geothermal exploration area.

[0047] In some embodiments, the process of testing the trained geothermal parameter prediction model includes:

[0048] The geological data of the data sets of the plurality of second drilling points may be input as input features into the trained geothermal parameter prediction model, wherein the geological data at least includes: Curie point depth, Moho surface depth, aeromagnetic anomaly data, Bouguer gravity anomaly data, altitude data, longitude and latitude, and distances from other faults;

[0049] The geothermal parameter prediction model can predict the geothermal parameters of each second drilling point in the output verification area. The predicted geothermal parameters include at least: geothermal gradient, geothermal heat flow and 150°C burial depth;

[0050] The performance index of the geothermal parameter prediction model can be calculated based on the predicted geothermal parameters and the measured geothermal parameters.

[0051] The process of testing the trained geothermal resource evaluation model includes:

[0052] The geothermal parameters of the data sets of the plurality of second drilling points may be input as input features into the trained geothermal resource evaluation model, wherein the geothermal parameter data at least includes: geothermal gradient, geothermal heat flow and 150°C burial depth;

[0053] The geothermal resource evaluation model can predict and output the geothermal resource evaluation results of each second drilling point in the verification area;

[0054] The performance index of the geothermal resource evaluation model can be calculated based on the predicted geothermal resource evaluation results and the measured geothermal resource evaluation labels.

[0055] In some embodiments, the performance index may be mean absolute error (MAE), mean square error (MSE), R-square, precision, recall, F1-score, etc. Among them, mean absolute error represents the average value of the absolute error between the predicted value and the measured value; mean square error represents the ratio of the sum of squares of the deviations between the predicted value and the measured value to the number of observations; the fit coefficient represents the ratio of the sum of squares of the difference between the predicted value and the measured value mean and the sum of squares of the difference between the measured value and the measured value mean; precision represents the probability of samples that are actually positive among all samples predicted to be positive; recall represents the probability of samples that are predicted to be positive among samples that are actually positive; and F1-score represents the harmonic mean of precision and recall.

[0056] In some embodiments, the process of eliminating abnormal second drilling points from the plurality of second drilling points according to the output geothermal parameters includes:

[0057] The geothermal parameters predicted by the geothermal parameter prediction model can be compared with the measured geothermal parameters;

[0058] If the difference between the predicted geothermal parameter data and the measured geothermal parameter value is greater than or equal to the set threshold, it can be considered that the second drilling point corresponding to the geological data input into the geothermal parameter prediction model is abnormal;

[0059] Abnormal second drilling points may be eliminated from the data set of the plurality of second drilling points.

[0060] The process of continuing to eliminate abnormal second drilling points from the remaining second drilling points according to the output geothermal resource evaluation results includes:

[0061] The geothermal resource evaluation results predicted by the geothermal resource evaluation model can be compared with the measured geothermal resource evaluation labels;

[0062] When the difference between the predicted geothermal resource evaluation result and the measured geothermal resource evaluation label value is greater than or equal to the set threshold, it can be considered that the second drilling point corresponding to the geothermal parameter input into the geothermal resource evaluation model is abnormal;

[0063] Abnormal second drilling points may continue to be eliminated from the remaining data sets of multiple second drilling points;

[0064] In some embodiments, the process of optimizing the trained geothermal parameter prediction model based on the remaining data set of the second drilling point includes:

[0065] According to the geological data in the remaining data set of the second drilling point, the geothermal parameters in the remaining data set of the second drilling point can be predicted using the trained geothermal parameter prediction model;

[0066] The deviation between the predicted geothermal parameters and the measured geothermal parameters can be used to adjust the geothermal parameter prediction model;

[0067] When the performance index of the geothermal parameter prediction model reaches a certain threshold, the adjustment is ended and an optimized geothermal parameter prediction model can be obtained.

[0068] The process of optimizing the trained geothermal resource evaluation model based on the remaining data set of the second drilling point includes:

[0069] According to the geothermal parameters in the remaining data set of the second drilling point, the geothermal resource evaluation model obtained by training can be used to predict the geothermal resource evaluation labels in the remaining data set of the second drilling point;

[0070] The geothermal resource evaluation model can be adjusted using the deviation between the predicted geothermal resource evaluation result and the measured geothermal resource evaluation label;

[0071] When the performance index of the geothermal resource evaluation model reaches a certain threshold, the adjustment is ended and an optimized geothermal resource evaluation model can be obtained.

[0072] S105: According to the geological data of the plurality of observation points, using the trained geothermal parameter prediction model and geothermal resource evaluation model, obtain a geothermal resource evaluation result of the first geothermal exploration area.

[0073] In some embodiments, the geothermal resource evaluation results of the first geothermal exploration area can be obtained directly based on the geological data of the multiple observation points using the optimized geothermal parameter prediction model and the optimized geothermal resource evaluation model. For example, the geological data of the observation point can be input as input features into the optimized geothermal parameter prediction model, and the geothermal influencing factor data at least include: Curie depth, Moho depth, aeromagnetic anomaly data, Bouguer gravity anomaly data, altitude data, longitude and latitude, and distance from other faults. The geothermal parameter prediction model can predict and output the geothermal parameters of each observation point, and the predicted geothermal parameters at least include: geothermal gradient, geothermal heat flow, and 150°C burial depth. The geothermal parameters of each observation point predicted and output by the geothermal parameter prediction model can be input as input features into the optimized geothermal resource evaluation model; the geothermal resource evaluation model can predict and output the geothermal resource evaluation results of each observation point. According to the geothermal resource evaluation results of each observation point, the geothermal resource evaluation results of the first geothermal exploration area can be obtained. For example, the geothermal resource evaluation results of the first geothermal exploration area can be obtained by taking the average of the geothermal resource evaluation results of each observation point, or by taking the median, weighted average or other methods of the geothermal resource evaluation results of each observation point.

[0074] In some embodiments, if there is a small amount of drilling data in the first geothermal exploration area, multiple observation points can be divided into multiple first observation points and multiple second observation points according to whether there is drilling data, the first observation points corresponding to geological data, geothermal parameters and geothermal resource evaluation labels, and the second observation points only correspond to geological data; the data related to the first observation points and the second observation points can be digitized; the first observation points corresponding to abnormal geological data and abnormal geothermal parameters are eliminated from the multiple first observation points; the geothermal parameter prediction model and the geothermal resource evaluation model are optimized again based on the geological data, geothermal parameters and geothermal resource evaluation labels of the remaining multiple first observation points; according to the geological data of the multiple second observation points, the geothermal parameter prediction model and the geothermal resource evaluation model after optimization are used to obtain the geothermal resource evaluation results of the first geothermal exploration area.

[0075] In some embodiments, the optimized geothermal parameter prediction model and geothermal resource evaluation model can be tested based on the geological data, geothermal parameters and geothermal resource evaluation labels of the multiple first observation points to obtain geothermal parameters output by the geothermal parameter prediction model and geothermal resource evaluation results output by the geothermal resource evaluation model; abnormal first observation points can be eliminated from the multiple first observation points based on the output geothermal parameters; abnormal first observation points can continue to be eliminated from the remaining first observation points based on the output geothermal resource evaluation results; the optimized geothermal parameter prediction model and geothermal resource evaluation model can be optimized again based on the geological data, geothermal parameters and geothermal resource evaluation labels of the remaining multiple first observation points.

[0076] By removing abnormal first observation points from multiple first observation points twice, the influence of abnormal geological data and geothermal parameters on model training is effectively avoided. These abnormal geological data and geothermal parameters can make the model unable to accurately fit the real data distribution, affecting the effect of the actual application of the model. Therefore, removing abnormal first observation points can make the model fit the real data distribution more accurately, and then obtain more refined geothermal resource evaluation results in the first geothermal exploration area.

[0077] In some embodiments, the process of testing the optimized geothermal parameter prediction model according to the geological data and geothermal parameters of the remaining plurality of first observation points includes:

[0078] The geological data of the plurality of first observation points may be input as input features into the optimized geothermal parameter prediction model, wherein the geological data at least include: Curie point depth, Moho surface depth, aeromagnetic anomaly data, Bouguer gravity anomaly data, altitude data, longitude and latitude, and distances from other faults;

[0079] The geothermal parameter prediction model can predict and output the geothermal parameters of each first observation point. The predicted geothermal parameters at least include: geothermal gradient, geothermal heat flow and 150°C burial depth;

[0080] The performance index of the geothermal parameter prediction model can be calculated based on the predicted geothermal parameters and the measured geothermal parameters.

[0081] According to the geothermal parameters and geothermal resource evaluation labels of the remaining multiple first observation points, the process of testing the optimized geothermal resource evaluation model includes:

[0082] The geothermal parameters of the plurality of first observation points may be input as input features into the optimized geothermal resource evaluation model, wherein the geothermal parameter data at least include: geothermal gradient, geothermal heat flow and 150°C burial depth;

[0083] The geothermal resource evaluation model can predict and output the geothermal resource evaluation results of each first observation point;

[0084] The performance index of the geothermal resource evaluation model can be calculated based on the predicted geothermal resource evaluation results and the measured geothermal resource evaluation labels.

[0085] In some embodiments, according to the output geothermal parameters, the process of removing abnormal first observation points from the plurality of first observation points includes:

[0086] The geothermal parameters predicted by the geothermal parameter prediction model can be compared with the measured geothermal parameters;

[0087] If the difference between the predicted geothermal parameter and the measured geothermal parameter value is greater than or equal to the set threshold, it can be considered that the first observation point corresponding to the geological data input into the geothermal parameter prediction model is abnormal;

[0088] Abnormal first observation points may be eliminated from the plurality of first observation points.

[0089] The process of removing abnormal first observation points from the remaining multiple first observation points again according to the output geothermal resource evaluation results includes:

[0090] The geothermal resource evaluation results predicted by the geothermal resource evaluation model can be compared with the measured geothermal resource evaluation labels;

[0091] When the difference between the predicted geothermal resource evaluation result and the measured geothermal resource evaluation label value is greater than or equal to the set threshold, it can be considered that the first drilling point corresponding to the geothermal parameter input into the geothermal resource evaluation model is abnormal;

[0092] Abnormal first observation points may continue to be eliminated from the remaining plurality of first observation points.

[0093] In some embodiments, the process of re-optimizing the optimized geothermal parameter prediction model according to the remaining plurality of first observation points includes:

[0094] According to the geological data of the remaining multiple first observation points, the geothermal parameters of the remaining multiple first observation points can be predicted using the optimized geothermal parameter prediction model;

[0095] The deviation between the predicted geothermal parameters and the measured geothermal parameters can be used to adjust the geothermal parameter prediction model again;

[0096] When the performance index of the geothermal parameter prediction model reaches a certain threshold, the adjustment is terminated again, and a geothermal parameter prediction model after optimization can be obtained.

[0097] The process of re-optimizing the optimized geothermal resource evaluation model according to the remaining multiple first observation points includes:

[0098] According to the geothermal parameters of the remaining multiple first observation points, the geothermal resource evaluation model obtained by training can be used to predict the geothermal resource evaluation labels of the remaining multiple first observation points;

[0099] The geothermal resource evaluation model can be adjusted again by using the deviation between the predicted geothermal resource evaluation result and the measured geothermal resource evaluation label;

[0100] When the performance index of the geothermal resource evaluation model reaches a certain threshold, the further adjustment is ended, and a geothermal resource evaluation model after optimization can be obtained.

[0101] In some embodiments, the process of obtaining the geothermal resource evaluation result of the first geothermal exploration area according to the geological data of the plurality of second observation points using the geothermal parameter prediction model and the geothermal resource evaluation model that have been optimized again includes:

[0102] The geological data of the plurality of second observation points can be input as input features into the geothermal parameter prediction model after optimization again, wherein the geothermal influencing factor data at least include: Curie point depth, Moho surface depth, aeromagnetic anomaly data, Bouguer gravity anomaly data, altitude data, longitude and latitude, and distance from other faults;

[0103] The geothermal parameter prediction model can predict and output the geothermal parameters of each second observation point. The predicted geothermal parameters at least include: geothermal gradient, geothermal heat flow and 150°C burial depth;

[0104] The geothermal parameters of each first observation point predicted and output by the geothermal parameter prediction model can be input as input features into the geothermal resource evaluation model after optimization again;

[0105] The geothermal resource evaluation model can predict and output the geothermal resource evaluation results of each first observation point, and the geothermal resource evaluation results of the first geothermal exploration area can be obtained according to the geothermal resource evaluation results of each first observation point. For example, the geothermal resource evaluation results of the first geothermal exploration area can be obtained by taking the average of the geothermal resource evaluation results of each observation point, or by taking the median of the geothermal resource evaluation results of each observation point, weighted average or other methods to obtain the geothermal resource evaluation results of the first geothermal exploration area.

[0106] The following introduces a specific scenario example of the embodiment of this specification.

[0107] 1. Figure 2Area A within the range of 127°E-129°E, 41°N-43°N is shown as a region lacking drilling data. Area A is taken as the first geothermal exploration area, and an observation point is taken at an interval of 1 km to collect geological data such as Curie point depth, Moho surface depth, aeromagnetic anomaly data, Bouguer gravity anomaly data, altitude data, longitude and latitude, and distance from other faults at each observation point. A small number of geothermal parameters such as geothermal gradient, geothermal heat flow, 150℃ depth, and corresponding geothermal resource evaluation labels are also collected at the observation points, and these data are digitized.

[0108] 2. Based on the geological data of each observation point in area A, area B with a number of drilling points greater than the set threshold and a geological background similarity with area A greater than the set threshold is selected as the second geothermal exploration area, such as Figure 3 shown.

[0109] 3. Collect the geological data such as Curie depth, Moho depth, aeromagnetic anomaly data, Bouguer gravity anomaly data, altitude data, longitude and latitude, and distance from other faults corresponding to each drilling point in area B, geothermal parameters such as geothermal gradient, geothermal heat flow and 150℃ depth measured at each drilling point, and geothermal resource evaluation labels measured at each drilling point, and digitize the above data. According to the geological unit division of area B, such as Figure 4 As shown, the western slope area, southwestern uplift area, and central depression area are divided into training areas; the northern plunge area, northeastern uplift area, and southeastern uplift area are divided into verification areas.

[0110] 4. Use the DNN model to establish a geothermal parameter prediction model, such as Figure 5 As shown in the figure, the geological data of the training area in area B is used as the input feature to predict the geothermal parameters of the training area in area B. The geothermal parameter prediction model is trained to fit by minimizing the square sum of the errors between the predicted geothermal parameters and the measured geothermal parameters. The DNN model is also used to establish a geothermal resource evaluation model, as shown in Figure 6 As shown in the figure, the geothermal parameters of the training area in area B are used as input features to predict the geothermal resource evaluation labels of the training area in area B, and the geothermal resource evaluation model is trained to fit by minimizing the square sum of the errors between the predicted geothermal resource evaluation results and the measured geothermal resource labels. The geological data and geothermal parameters of the verification area in area B are respectively input into the trained geothermal parameter prediction model and geothermal resource evaluation model, and the geothermal parameters and geothermal resource evaluation results of the verification area are predicted. The performance indicators of the geothermal parameter prediction model and the geothermal resource evaluation model are calculated, and the predicted results are compared with the measured results, abnormal data are eliminated, and the model is optimized until the model performance indicators reach a certain threshold.

[0111] 5. According to a small number of geothermal parameters and geothermal resource evaluation labels in region A, the optimized geothermal parameter prediction model and geothermal resource evaluation model are optimized again. Then, the geothermal parameters and geothermal resource evaluation results in region A are predicted using the optimized geothermal parameter prediction model and geothermal resource evaluation model. Finally, the evaluation results of geothermal parameters and geothermal resources at each observation point in region A are obtained, such as Figure 7 As shown. The predicted terrestrial heat flow at each observation point in region A is between 71.35-87.53 mW / m 2 The average heat flow is about 79.97mW / m 2 ; The geothermal gradient is between 3.16-4.03℃ / 100m, and the average geothermal gradient is about 3.66℃ / 100m; the burial depth at 150℃ is between 3350-4272m, and the average burial depth is about 3689m.

[0112] Based on the above-mentioned geothermal resource evaluation method based on machine learning, this specification also proposes an embodiment of a geothermal resource evaluation device based on machine learning. Figure 8 As shown, the evaluation device may specifically include the following modules:

[0113] The first acquisition module 801 may be used to acquire geological data of a plurality of observation points selected in a first geothermal exploration area;

[0114] A selection module 802 may select a second geothermal exploration area based on geological data of a plurality of observation points in the first geothermal exploration area;

[0115] The second acquisition module 803 may be used to acquire data sets of multiple drilling points in the second geothermal exploration area, wherein the data sets of the multiple drilling points include geological data of the drilling points, geothermal parameters of the drilling points, and geothermal resource evaluation labels of the drilling points;

[0116] A training module 804 may be used to train a geothermal parameter prediction model and a geothermal resource evaluation model based on the data sets of the plurality of drilling points;

[0117] The third acquisition module 805 can be used to obtain the geothermal resource evaluation result of the first geothermal exploration area according to the geological data of the multiple observation points and using the trained geothermal parameter prediction model and geothermal resource evaluation model.

[0118] In some embodiments, the first acquisition module 801 can be specifically used to acquire geological data of multiple observation points selected in the first geothermal exploration area; the geological data at least includes: Curie point depth, Moho surface depth, aeromagnetic anomaly data, Bouguer gravity anomaly data, altitude data, longitude and latitude, and distance from other faults.

[0119] In some embodiments, the selection module 802 can be specifically used to select an area where the number of drilling points is greater than or equal to a set threshold and the similarity between the corresponding geological background and the geological background of the first geothermal exploration area meets the similarity condition as the second geothermal exploration area.

[0120] In some embodiments, the second acquisition module 803 can be specifically used to acquire data sets of multiple drilling points in the second geothermal exploration area, wherein the data sets of the multiple drilling points include geological data of the drilling points, geothermal parameters of the drilling points, and geothermal resource evaluation labels of the drilling points; the geothermal parameters include at least: geothermal gradient, geothermal heat flow, and 150°C burial depth.

[0121] In some embodiments, the above-mentioned training module 804 can be specifically used to train a geothermal parameter prediction model based on the geological data and geothermal parameters in the data set of the multiple drilling points; the geothermal parameter prediction model is a multi-input multi-output model; the geothermal resource evaluation model is trained based on the geothermal parameters and geothermal resource evaluation labels in the data set of the multiple drilling points; the geothermal resource evaluation model is a multi-input single-output model.

[0122] In some embodiments, the above-mentioned training module 804 can also be specifically used to divide the second geothermal exploration area into a training area and a verification area, the training area includes multiple first drilling points, and the verification area includes multiple second drilling points; according to the data sets of the multiple first drilling points, the geothermal parameter prediction model and the geothermal resource evaluation model are trained; according to the data sets of the multiple second drilling points, the trained geothermal parameter prediction model and the geothermal resource evaluation model are tested to obtain the geothermal parameters output by the geothermal parameter prediction model and the geothermal resource evaluation results output by the geothermal resource evaluation model; according to the output geothermal parameters and the output geothermal resource evaluation results, the abnormal second drilling points are eliminated from the multiple second drilling points in the verification area; according to the data sets of the remaining second drilling points in the verification area, the trained geothermal parameter prediction model and the geothermal resource evaluation model are optimized.

[0123] In some embodiments, the third acquisition module 805 can be specifically used to input the geological data of the multiple observation points into a geothermal parameter prediction model to obtain the geothermal parameters of the multiple observation points; input the geothermal parameters of the multiple observation points into a geothermal resource evaluation model to obtain the geothermal resource evaluation results of the first geothermal exploration area.

[0124] In some embodiments, the third acquisition module 805 can also be used to divide the multiple observation points into multiple first observation points and multiple second observation points, the first observation points corresponding to geological data, geothermal parameters and geothermal resource evaluation labels, and the second observation points only corresponding to geological data; the first observation points corresponding to abnormal geothermal parameters and abnormal geothermal resource evaluation labels are eliminated from the multiple first observation points; according to the geological data, geothermal parameters and geothermal resource evaluation labels of the remaining multiple first observation points, the optimized geothermal parameter prediction model and geothermal resource evaluation model are optimized again; according to the geological data of the multiple second observation points, the geothermal parameter prediction model and geothermal resource evaluation model that have been optimized again are used to obtain the geothermal resource evaluation results of the first geothermal exploration area.

[0125] It should be noted that the units, devices or modules described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. For the convenience of description, the above devices are described separately by functions divided into various modules. Of course, when implementing this specification, the functions of each module can be implemented in the same or more software and / or hardware, or the modules that implement the same function can be implemented by a combination of multiple sub-modules or sub-units. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0126] As can be seen from the above, a geothermal resource evaluation device based on machine learning provided in the embodiment of this specification utilizes a machine learning model to learn the ability to predict geothermal parameters and evaluate geothermal resources based on geological data, thereby improving the accuracy of geothermal resource exploration in the absence of drilling data, which is of great significance for the delineation and division of favorable areas for geothermal resource exploration.

[0127] The embodiments of the present specification also provide a computer device for a geothermal resource evaluation method based on machine learning, comprising a processor and a memory for storing instructions executable by the processor, wherein the processor can perform the following steps according to the instructions when implemented: obtaining geological data of multiple observation points selected in a first geothermal exploration area; selecting a second geothermal exploration area based on the geological data of multiple observation points in the first geothermal exploration area; obtaining data sets of multiple drilling points in the second geothermal exploration area, wherein the data sets of the multiple drilling points include the geological data of the drilling points, the geothermal parameters of the drilling points, and the geothermal resource evaluation labels of the drilling points; training a geothermal parameter prediction model and a geothermal resource evaluation model based on the data sets of the multiple drilling points; and obtaining geothermal resource evaluation results of the first geothermal exploration area using the trained geothermal parameter prediction model and the geothermal resource evaluation model based on the geological data of the multiple observation points.

[0128] In order to complete the above instructions more accurately, refer to Fig. 9 As shown, the embodiment of this specification also provides another specific electronic device, wherein the electronic device includes a network communication port 901, a processor 902 and a memory 903, and the above structures are connected through internal cables so that each structure can perform specific data interaction.

[0129] Among them, the network communication port 901 can be specifically used to obtain geological data of multiple observation points selected in the first geothermal exploration area; obtain data sets of multiple drilling points in the second geothermal exploration area, and the data sets of the multiple drilling points include geological data of the drilling points, geothermal parameters of the drilling points, and geothermal resource evaluation labels of the drilling points.

[0130] The processor 902 can be specifically used to select a second geothermal exploration area based on the geological data of multiple observation points in the first geothermal exploration area; train a geothermal parameter prediction model and a geothermal resource evaluation model based on the data sets of the multiple drilling points; and obtain a geothermal resource evaluation result of the first geothermal exploration area using the trained geothermal parameter prediction model and geothermal resource evaluation model based on the geological data of the multiple observation points.

[0131] The memory 903 may be specifically used to store corresponding instruction programs.

[0132] In this embodiment, the network communication port 901 can be a virtual port that is bound to different communication protocols so that different data can be sent or received. For example, the network communication port can be a port responsible for web data communication, a port responsible for FTP data communication, or a port responsible for email data communication. In addition, the network communication port can also be a physical communication interface or communication chip. For example, it can be a wireless mobile network communication chip, such as GSM, CDMA, etc.; it can also be a Wifi chip; it can also be a Bluetooth chip.

[0133] In this embodiment, the processor 902 may be implemented in any appropriate manner. For example, the processor may take the form of a microprocessor or processor and a computer-readable medium storing a computer-readable program code (such as software or firmware) executable by the (micro)processor, a logic gate, a switch, an application specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller, etc. This specification does not limit this.

[0134] In this embodiment, the memory 903 may include multiple levels. In a digital system, anything that can store binary data can be a memory; in an integrated circuit, a circuit with a storage function that has no physical form is also called a memory, such as RAM, FIFO, etc.; in a system, a storage device with a physical form is also called a memory, such as a memory stick, TF card, etc.

[0135] Corresponds to Figure 1 In the method, an embodiment of the present invention further provides a computer-readable storage medium, on which computer instructions are stored, and when the computer instructions are executed by a processor, the above steps are executed.

[0136] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0137] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0138] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0139] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0140] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A geothermal resource evaluation method based on machine learning, characterized in that: include: Acquiring geological data for a plurality of observation points selected within a first geothermal exploration area; selecting a second geothermal exploration area based on geological data of a plurality of observation points in the first geothermal exploration area; Acquire a data set of multiple drilling points in the second geothermal exploration area, wherein the data set of the multiple drilling points includes geological data of the drilling points, geothermal parameters of the drilling points, and geothermal resource evaluation labels of the drilling points; Training a geothermal parameter prediction model and a geothermal resource evaluation model based on the data sets of the plurality of drilling points; According to the geological data of the multiple observation points, the geothermal parameter prediction model and the geothermal resource evaluation model obtained by training are used to obtain the geothermal resource evaluation results of the first geothermal exploration area.

2. The method according to claim 1, characterized in that: The geological data at least include: Curie surface depth, Moho surface depth, aeromagnetic anomaly data, Bouguer gravity anomaly data, altitude data, longitude and latitude, and distance from other faults; The geothermal parameters include at least: geothermal gradient, geothermal flow and 150°C burial depth; The geothermal parameter prediction model is a multi-input multi-output model; The geothermal resource evaluation model is a multi-input single-output model.

3. The method according to claim 1, characterized in that: The selected second geothermal exploration area comprises: An area where the number of drilling points is greater than or equal to a set threshold and the similarity between the corresponding geological background and the geological background of the first geothermal exploration area meets the similarity condition is selected as the second geothermal exploration area.

4. The method according to claim 1, characterized in that: The training of geothermal parameter prediction model and geothermal resource evaluation model includes: training a geothermal parameter prediction model based on the geological data and geothermal parameters in the data set of the plurality of drilling points; A geothermal resource evaluation model is trained based on the geothermal parameters and geothermal resource evaluation labels in the data sets of the multiple drilling points.

5. The method according to claim 4, characterized in that: The training of geothermal parameter prediction model and geothermal resource evaluation model includes: dividing the second geothermal exploration area into a training area and a verification area, wherein the training area includes a plurality of first drilling points and the verification area includes a plurality of second drilling points; Training a geothermal parameter prediction model and a geothermal resource evaluation model based on the data sets of the plurality of first drilling points; Testing the trained geothermal parameter prediction model and geothermal resource evaluation model according to the data sets of the plurality of second drilling points to obtain geothermal parameters output by the geothermal parameter prediction model and geothermal resource evaluation results output by the geothermal resource evaluation model; Eliminating abnormal second drilling points from a plurality of second drilling points in the verification area according to the output geothermal parameters and the output geothermal resource evaluation results; The trained geothermal parameter prediction model and geothermal resource evaluation model are optimized according to the data set of the remaining second drilling points in the verification area.

6. The method according to claim 1, characterized in that: The obtaining of geothermal resource evaluation results of the first geothermal exploration area includes: Inputting the geological data of the plurality of observation points into a geothermal parameter prediction model to obtain geothermal parameters of the plurality of observation points; The geothermal parameters of the plurality of observation points are input into a geothermal resource evaluation model to obtain a geothermal resource evaluation result of the first geothermal exploration area.

7. The method according to claim 6, characterized in that: The obtaining of geothermal resource evaluation results of the first geothermal exploration area includes: The plurality of observation points are divided into a plurality of first observation points and a plurality of second observation points, wherein the first observation points correspond to geological data, geothermal parameters and geothermal resource evaluation labels, and the second observation points correspond to only geological data; Eliminating the first observation points corresponding to abnormal geothermal parameters and abnormal geothermal resource evaluation labels from the plurality of first observation points; According to the geological data, geothermal parameters and geothermal resource evaluation labels of the remaining multiple first observation points, the optimized geothermal parameter prediction model and geothermal resource evaluation model are further optimized; According to the geological data of the plurality of second observation points, the geothermal parameter prediction model and the geothermal resource evaluation model that have been optimized again are used to obtain the geothermal resource evaluation results of the first geothermal exploration area.

8. A geothermal resource evaluation device based on machine learning, characterized in that: include: A first acquisition module, used to acquire geological data of a plurality of observation points selected in a first geothermal exploration area; A selection module, used to select a second geothermal exploration area based on geological data of a plurality of observation points in the first geothermal exploration area; A second acquisition module is used to acquire data sets of multiple drilling points in the second geothermal exploration area, wherein the data sets of the multiple drilling points include geological data of the drilling points, geothermal parameters of the drilling points, and geothermal resource evaluation labels of the drilling points; A training module, used for training a geothermal parameter prediction model and a geothermal resource evaluation model according to the data sets of the plurality of drilling points; The third acquisition module is used to obtain the geothermal resource evaluation results of the first geothermal exploration area based on the geological data of the multiple observation points and using the trained geothermal parameter prediction model and geothermal resource evaluation model.

9. A geothermal resource evaluation device based on machine learning, characterized in that: The method comprises a processor and a memory for storing processor-executable instructions, wherein the processor implements the steps of the method according to any one of claims 1 to 7 when executing the instructions.

10. A computer-readable storage medium, characterized in that: Computer instructions are stored thereon, and when the instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images