Index system and method model for deep geothermal resource potential evaluation and dessert area optimization

Abstract

The invention belongs to the technical field of deep geothermal resource evaluation, and particularly relates to an index system and method model for deep geothermal resource potential evaluation and dessert area optimization, and the method comprises the following steps: 1, constructing an evaluation index system according to deep geothermal geologic features, and 2, constructing a deep geothermal resource potential evaluation method based on a MaxEnt model. According to the method, existing information can be fully utilized by utilizing the MaxEnt model on the premise of not introducing additional assumptions, all optimal probability distribution models conforming to priori knowledge are obtained, the method is widely applied to analysis and inference of statistical data, and the method can be widely applied to scenes with sparse sampling points and complex correlation among indexes. The model can obtain a reliable evaluation result and good applicability, comprises the links of evaluation index system construction, model training and inspection, result verification and the like, and can obtain a good evaluation result for the problems of sparse sample data, complex geothermal geological conditions and the like in deep geothermal resource potential evaluation.

Description

Technical Field

[0001] The present invention belongs to the technical field of deep geothermal resource evaluation, and in particular relates to an index system and method model for deep geothermal resource potential evaluation and sweet spot optimization. Background Art

[0002] As an important renewable clean energy, geothermal energy has the advantages of large reserves, good stability, wide distribution, and no restrictions on seasons / day and night. Deep geothermal resources, that is, the heat energy stored in underground rock and fluids buried at a depth of 3-10km, are an important direction of global energy development. It is estimated that the total amount of deep geothermal resources in the world is 30 times the total energy of coal, oil and natural gas, and has great development potential. The development and utilization of deep geothermal resources is of great significance to improving energy supply, promoting economic development and environmental protection, and is also conducive to accelerating the realization of the "dual carbon goals";

[0003] However, the deep burial depth and wide distribution of deep geothermal resources have resulted in a lack of exploration data and a low degree of exploration, which makes resource potential evaluation and sweet spot optimization particularly necessary. At present, comprehensive evaluation based on regional geological surveys and geophysical technology is the main method, among which the analytic hierarchy process (AHP) is a widely used evaluation method. However, the analytic hierarchy process relies on expert experience for index screening and weight assignment, and has subjective and adaptability problems, which may lead to deviations in evaluation results and reduced reliability;

[0004] Data-driven quantitative evaluation methods, such as the weight of evidence method (WofE), have been applied in shallow geothermal resource evaluation, but are limited in deep geothermal resource evaluation by the sparseness and complexity of sampling data. Machine learning methods, especially those based on data-driven methods, can mine potential laws in large amounts of scattered data, avoid subjective influences, and overcome the shortcomings of expert-driven models;

[0005] In the existing technology, Pang Zhonghe et al. proposed an evaluation index system closely related to the occurrence of deep geothermal energy, and used the hierarchical analysis method to optimize the deep geothermal energy exploration and development areas. et al. and Sang et al. used the weight of evidence method to identify geothermal areas in Turkey and eastern Liaoning Province, respectively. Moghaddam et al. studied the spatial relationship between geothermal occurrence and geological and geochemical characteristics. Although these methods have achieved certain results, they still have limitations in the evaluation of deep geothermal resources;

[0006] Therefore, the present invention proposes a method model based on machine learning for deep geothermal resource potential evaluation and sweet spot optimization, aiming to overcome the shortcomings of the existing technology and provide an objective and reliable evaluation method. Summary of the invention

[0007] The purpose of the present invention is to provide an index system and method model for deep geothermal resource potential evaluation and sweet spot optimization, which can achieve better evaluation results for the problems of sparse sample data and complex geothermal geological conditions in deep geothermal resource potential evaluation.

[0008] The technical solution adopted by the present invention is as follows:

[0009] An index system and method model for deep geothermal resource potential evaluation and sweet spot selection, comprising the following steps:

[0010] Step 1: Construct an evaluation index system based on the deep geothermal geological characteristics;

[0011] Step 2: Construct a deep geothermal resource potential evaluation method based on the MaxEnt model.

[0012] The technical effects achieved by the present invention are:

[0013] The present invention establishes an evaluation index system based on the characteristics of deep geothermal resources, and uses the Maximum Entropy (MaxEnt) model to evaluate the potential of deep geothermal resources and optimize the sweet spot. This method makes full use of the MaxEnt model as a typical data-driven and machine learning method, which can make predictions or inferences based on incomplete information. Without introducing additional assumptions, it makes full use of existing information to obtain the optimal probability distribution that conforms to prior knowledge. In scenarios where sampling points are sparse and the correlation between indicators is complex, it can obtain more reliable evaluation results and good applicability.

[0014] It includes building an evaluation index system, model training and testing, and result verification, and can achieve better evaluation results for problems such as sparse sample data and complex geothermal geological conditions in the evaluation of deep geothermal resource potential. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is the technical roadmap of the present invention;

[0016] Figure 2 This is the outline map of the location and structure of the Subei Basin in the present invention;

[0017] Figure 3 It is the index layer in the present invention;

[0018] Figure 4 It is a schematic diagram of the ROC curve of the model accuracy test in the present invention;

[0019] Figure 5 It is a schematic diagram of sensitivity analysis in the present invention;

[0020] Figure 6This is the deep geothermal potential distribution map of the Northern Jiangsu Basin in the present invention. DETAILED DESCRIPTION

[0021] In order to make the purpose and advantages of the present invention more clearly understood, the present invention is specifically described below in conjunction with embodiments. It should be understood that the following text is only used to describe one or several specific embodiments of the present invention, and does not strictly limit the scope of protection of the specific claims of the present invention.

[0022] like Figure 1-6 As shown, an index system and method model for deep geothermal resource potential evaluation and sweet spot area optimization includes the following steps:

[0023] S1: Construct an evaluation index system based on the deep geothermal geological characteristics;

[0024] Based on the formation principle of deep geothermal, geological tectonic action, geophysical response characteristics, etc., the inventors believe that when evaluating deep geothermal resources, an indicator system should be constructed from three aspects:

[0025] S11: Geophysical parameters reflecting the composition of deep heat sources, such as Curie surface, Moho surface, burial depth of low-velocity and high-conductivity layer, etc., reflect the intrinsic structural characteristics of the deep layer and play a decisive role in the composition and distribution of deep heat sources. At the same time, since the P-wave velocity reflects the energy transfer characteristics in the crust-mantle medium and can express the physical characteristics of the crust-mantle structure, it should also be included in the evaluation index system of deep geothermal resources;

[0026] S12: An indicator reflecting the structural geological conditions and magmatic activity of deep strata. As a longitudinal communication channel of the crust, deep and large faults play a vital role in the favorable development location of deep heat sources. The regional tectonic unconformity formed by the Indosinian surface during the Indosinian movement and the thick Mesozoic and Cenozoic terrestrial clastic sedimentary strata on the upper part have a good shielding and heat preservation effect on the deep heat. The distribution of intermediate-acidic magmatic rocks reflects both the regional magmatic activity and the spatial characteristics of radioactive heat sources, and can be used as an indicative parameter of deep heat sources.

[0027] S13: Although the indicative indicators of deep geothermal energy, such as geothermal gradient and geothermal heat flow value, are affected by sampling depth and measurement conditions, they still have a corresponding indicative effect on deep heat sources; the distribution of hot spring points also has a certain indicative effect; in addition, the spatial location of development projects such as deep geothermal development wells is an important sign of the favorable location of deep geothermal resources.

[0028] S2: Construct a deep geothermal resource potential evaluation method based on the MaxEnt model.

[0029] The MaxEnt model was proposed by statistician ET Jaynes (1957) based on information entropy. In information theory, entropy represents the degree of uncertainty of a random variable. Given a random variable X, its information entropy is:

[0030]

[0031] Among them, H(X) represents the information entropy of the random variable (X), and P(x) represents the probability that the random variable (X) takes the value (X);

[0032] Given a training dataset Define the characteristic function:

[0033]

[0034] The random variable X and the target Y are binary functions that are 1 when they both satisfy a certain condition and 0 otherwise. The empirical distribution of the joint distribution P(x, y) for:

[0035]

[0036] Where N is the number of samples in the training data set, which is used to calculate the empirical distribution;

[0037] From the above formula, we can get the empirical distribution of the marginal distribution P(x): for:

[0038]

[0039] At this point, the mathematical expectation of the characteristic function in the training data with respect to the empirical distribution is:

[0040]

[0041] The mathematical expectation of the joint distribution of features obtained by formula (5) is:

[0042]

[0043] Furthermore, according to the constraint get:

[0044]

[0045] If there are n characteristic functions, there are n constraints;

[0046] Furthermore, C is used to represent the set that satisfies the constraints:

[0047]

[0048] Therefore, the MaxEnt model is found in the model set C that satisfies the constraints and maximizes the entropy of the conditional distribution P(Y|X). In formula (8), the entropy of the conditional distribution P(Y|X) is expressed as:

[0049]

[0050] Therefore, the MexEnt model is expressed as:

[0051]

[0052] The MaxEnt model is a typical machine learning model. The receiver operating characteristic curve (ROC) and the area under the curve (AUC) can be used to evaluate the model accuracy. The closer the AUC value is to 1, the better the performance of the model. AUC of 0.5 indicates a close random distribution.

[0053] On the basis of MaxEnt model, geothermal wells are used as the indicator target of deep geothermal. By analyzing the characteristics of deep crust structure, heat generation environment, heat storage structure, heat conduction structure and heat release structure, nine indicators including the burial depth of asthenosphere, Moho surface and Curie surface, the top surface burial depth of low-velocity and high-conductivity layer and P-wave velocity distribution, distance from deep faults, distribution of intermediate-acid magmatic rocks in Mesozoic and Cenozoic eras, burial depth of Indosinia surface and measured heat flow value of drilling holes are included in the evaluation system. The MaxEnt model is used to quantitatively calculate and test the model of the potential distribution of deep geothermal resources in the study area, and the probability distribution results of deep geothermal resources in the study area are obtained. The location of hot spring points is used as a test indicator to verify the reliability of the evaluation results, and finally the distribution of deep geothermal resource potential in the evaluation area and the ranking of influencing factors are obtained. Technical route such as Figure 1 shown.

[0054] Based on the attached Figure 1 The technical route in this invention not only fully considers the factors such as deep geothermal geological conditions, geophysical characteristics, thermal display characteristics, but also takes into account the main characteristics of sparse deep geothermal data. The evaluation results can not only directly obtain the spatial distribution characteristics of deep geothermal resources and delineate the sweet spot, but also clearly point out the main controlling factors affecting the distribution characteristics of deep geothermal resources in the study area;

[0055] The present application is explained below using specific embodiments:

[0056] (1) Geological background of the study area;

[0057] The Subei Basin is located in the central part of Jiangsu Province, a relatively economically developed province in eastern China. It is a Mesozoic-Cenozoic superimposed basin in eastern China. More than half a century of oil and gas exploration has revealed the deep structural geological conditions of the basin and accumulated a large amount of geological data. Geothermal resource exploration in recent years has shown that the deep geothermal resources in the basin have great potential. However, the Subei Basin is extensive, and the spatial distribution law of deep geothermal resources, the dominant areas and the factors affecting the distribution of deep geothermal reservoirs are unclear. Therefore, taking the Subei Basin as an example, using the existing structural geology and geothermal geology theoretical research results in the region, with the help of relatively rich geothermal geological exploration data and geophysical interpretation data, the favorable areas of deep geothermal distribution are studied to reveal the spatial distribution law of deep geothermal resources in the study area and the controlling factors of reservoirs, so as to promote the development and utilization of geothermal energy.

[0058] The Subei Basin is located to the east of the Tanlu Fault Zone. It is a Mesozoic-Cenozoic continental sedimentary basin developed on the Archean, Proterozoic Sinian and Paleozoic basements, with an area of ​​about 32,000 km. 2 The landform is mainly plains, with an average altitude of less than 20 meters. The Subei Basin is located at the junction of the North China Plate and the Yangtze Plate. In the Mesozoic and Cenozoic Era, it was affected by the upthrust of the asthenosphere caused by the subduction of the Pacific Plate to the Eurasian Plate, and became the largest high-heat flow geothermal anomaly area in East China. Due to the compression of the plates, the Subei Basin has undergone many tectonic movements and has a strong structural segmentation, forming a tectonic feature of "two depressions and one uplift", with the Yanfu Depression in the north, the Dongtai Depression in the south, and the Jianhu Uplift in the middle. Figure 2 Shows the structure of the Subei Basin.

[0059] Data show that NJB is rich in deep geothermal resources and is one of the richest regions in eastern China, with great development potential. However, due to the thick Cenozoic sedimentary strata in the Subei Basin, weak surface thermal structural traces such as thermal alteration mineral precipitation, volcanic structures, and earthquakes, and few surface thermal display features such as hot springs, it is difficult to directly judge the distribution of deep geothermal resources through surface thermal display features, resulting in great difficulty and uncertainty in the evaluation of deep geothermal potential areas in the Subei Basin.

[0060] (2) indicator layer;

[0061] 1) Depth of asthenosphere;

[0062] The asthenosphere is a plastic rheological layer composed of partially molten rock material located below the lithosphere. The temperature of the asthenosphere layer is 500-600℃ higher than that of the lithosphere layer at the same depth. Therefore, when the high-temperature asthenosphere rises to a certain height, its upward heat conduction can cause the earth's heat flow above it to form a high calorific value anomaly zone. Therefore, the shallower parts of the asthenosphere are important deep heat source areas. According to the research results of Chen Husheng et al. (1999) and (Xing Jishan et al. 2019) on the thickness of the lithosphere in the Lower Yangtze and adjacent areas, the depth distribution of the asthenosphere in the Subei Basin ( Figure 3 (a)).

[0063] 2) Moho depth;

[0064] The Moho is the interface between the crust and the upper mantle. The shallow depth of the Moho indicates that the deep heat source is closer to the surface, which is conducive to the upward conduction of deep heat. In the Cenozoic, the eastern part of China was subjected to tension, which led to large-scale tectonic thinning of the lithosphere and formed a shallow Moho (Liu Demin et al. 2021). The Subei Basin is located in the uplift area of ​​the Moho and is generally shallow. The Moho contour lines compiled from the gravity data of Chen Husheng et al. (1999) are as follows Figure 3 (b) as shown.

[0065] 3) depth of burial;

[0066] The Curie point is the temperature interface where ferromagnetic minerals in rocks become paramagnetic due to temperature rise to the Curie point. It can reflect the thermal state of the deep crust and the distribution characteristics of the underground temperature field. The shallow depth of the Curie point indicates that the high heat area is close to the surface, which is conducive to the conduction of deep heat flow to the surface and the formation of heat energy enrichment areas. It is an important indicator of geothermal prospecting areas.

[0067] Based on the reinterpretation of the Curie depth results of Jiangsu Province based on the correction and fusion of magnetic survey data from different periods and scales, this paper obtains the distribution map of the Curie depth in the northern Jiangsu Basin. The overall depth is shallow in the west and deep in the east. The Curie depth on the west side of the basin near the Tanlu Fault is the shallowest. Figure 3 (c).

[0068] 4) Low speed and high conductor burial depth;

[0069] Tectonic activities deep in the earth's crust can cause changes in rock temperature and pressure, resulting in melting, rheology and other phenomena, changing the original conductivity of the rock. The low-velocity and high-conductivity layer obtained through magnetotelluric detection may be a partial melt (magma chamber), which is also an important deep heat source (Pang Zhonghe et al. 2020), and can gather the heat energy transmitted from the asthenosphere diapir (Liu Demin et al. 2022).

[0070] 5) Distance from deep faults;

[0071] As an important structural feature in the earth's crust, deep and large faults are structural surfaces where crustal materials are broken and moved under stress. They usually have excellent permeability and connectivity, and are channels for the upward migration of high-temperature fluids deep in the crust. Especially at the intersection of deep and large faults, they are often called the center of deep material and energy exchange, which is conducive to the formation of high-temperature geothermal resources. Studies have shown that the distribution and scale of fault zones have a significant impact on the formation and distribution of geothermal resources. Generally, the closer the distance to the fault, the higher the temperature of deep geothermal resources. Based on the distribution of deep and large fault structures in the study area, a thematic map of the distance from deep and large faults is obtained, such as Figure 3 (f) as shown.

[0072] 6) Distribution of intermediate-acidic igneous rocks;

[0073] The heat energy brought by magmatic activity can directly affect the formation and distribution of geothermal resources. The existence and distribution of magmatic rocks also have an important impact on geothermal distribution. Late-formed, highly radioactive magmatic rocks (mainly intermediate-acidic rocks) are favorable magmatic heat sources. At the same time, magmatic rocks have high thermal conductivity and heat capacity, and can store and transfer a large amount of heat energy.

[0074] There was relatively little magmatic activity in the early Mesozoic in the Subei Basin, and only in the middle and late Yanshan period did a few intermediate-acidic rock bodies develop; the Cenozoic basalt magma was relatively active, and experienced intense activity in three periods of the Paleocene, Neogene and Quaternary, with a wide distribution and a large range. There is no large-scale granite intrusion in the Subei Basin, which is mostly distributed in small blocks, and the heat generation rate of radioactive elements is low, belonging to low-radioactive rocks (Liu Demin et al. 2020). The distribution of intermediate-acidic magmatic rocks in the Subei Basin is as follows: Figure 3 (g) as shown.

[0075] 7) P-wave velocity;

[0076] As an important parameter for the propagation of seismic waves in the strata, the P-wave velocity can reflect the physical properties of the strata, such as lithology, density, and elasticity. Strata with different lithologies and structures have different responses to the propagation velocity of P-waves. By measuring and analyzing the changes in P-wave velocity, the distribution of geothermal resources and potential heat sources can be indirectly inferred. The distribution of P-wave velocity at a depth of 15 to 20 km in the Subei Basin is shown in the figure. Figure 3 (e).

[0077] 8) Earth heat flow value;

[0078] Terrestrial heat flow is a parameter that directly reflects the thermal state of the earth's crust. As a direct indicator of the abundance of geothermal resources, it reflects the intensity of heat loss through the surface per unit area and per unit time. The change in terrestrial heat flow is the result of the combined influence of multiple factors such as geological structure, magma activity, and stratum burial depth. The distribution and potential of geothermal resources can be relatively accurately evaluated through the terrestrial heat flow values ​​measured by drilling. Figure 3 (h) shown.

[0079] 9) Indosinian surface burial depth;

[0080] The Indosinian surface refers to the regional tectonic unconformity surface formed in the Lower Yangtze region where the Subei Basin is located during the Indosinian movement (Liu Hehua et al. 2019). Above the Indosinian surface is a thick layer of Mesozoic and Cenozoic terrestrial clastic sedimentary strata with relatively low thermal conductivity. It has a good shielding and insulation effect on the heat conducted upward from the deep, and can be regarded as the cap layer of deep geothermal. The greater its burial depth, the thicker the cap layer, and the easier it is to store geothermal resources. The burial depth of the Indosinian surface in the Subei Basin drawn based on petroleum exploration data is shown in the figure. Figure 3 (i) as shown.

[0081] (3) Calculation results;

[0082] 1) Result inspection;

[0083] The spatial locations of the 34 geothermal wells mentioned above are used as indicator variables for existing deep geothermal resource points. The 9 layer data are used as conditional indicators for calculating the correlation measurement between deep geothermal and indicator variables. After 10 rounds of iterative calculations (in each round of calculation, 10 of the 34 existing geothermal wells are randomly selected as test data, and the rest are used as training data), Figure 4 The ROC curve and AUC test results are shown. The AUC value is 0.893, indicating that the selected indicator system has a high-precision prediction result for the target variable, and the probability distribution calculation result is highly reliable, which means that the MaxEnt model can obtain reliable prediction results for the potential distribution of deep geothermal resources.

[0084] 2) Indicator sensitivity analysis;

[0085] The Jackknife Test was used to compare the AUCs of the nine conditional indicators. Figure 5 (a)) It can be seen that the distance to a deep fault (X6) has obvious characteristics that are easy to be compared with other indicators. The AUC without value of the distance to a deep fault is the lowest among all indicators ( Figure 5(a)), is 0.7685, and its AUC with only value is the largest, which is 0.6578, indicating that the distance from the deep fault has the most obvious impact on the AUC when it is included and excluded from the conditional indicator system, that is, the spatial correlation between the distance from the deep fault and the location of the geothermal well is the strongest. On the other hand, the asthenosphere burial depth (X1) shows completely opposite characteristics, with its AUC without being the highest globally, reaching 0.8017, and the AUC with only being the lowest globally ( Figure 5 (a)), is 0.5014, indicating that the correlation between the depth of the asthenosphere and the spatial location of the geothermal well is weak.

[0086] At the same time, from the indicator gain of training data ( Figure 5 (b)), the Training gain with only of the distance from deep faults (X6) is the highest globally (0.2639), and the Training gain without is the lowest globally (0.5907), indicating that the distance from deep faults is of the highest importance in the conditional index system. The Training gain with only of the asthenosphere (X1) is 0.0252, the lowest globally, and the Training gain without is 0.7606, the highest globally, indicating that when X6 and X1 are considered as single factors, their sharing of AUC is at two extremes, respectively, and they are the most sensitive and least sensitive elements of the deep geothermal spatial location.

[0087] 3) Importance of indicators;

[0088] Based on the average value of the calculation results of 10 iterations, the contribution percentages of the 9 indicators to the calculation of deep geothermal resources are shown in Table 1. It can be seen that under the deep geothermal conditions of the Subei Basin, the distance from the deep fault is the most important parameter. Together with the burial depth of the top surface of the low-velocity high-conductor and the P-wave velocity, the three indicators contribute more than 62.5% to the final probability distribution results and the ranking importance exceeds 63%, which together play a decisive role.

[0089]

[0090] 4) Potential and verification of deep geothermal resources in the study area;

[0091] The predicted distribution probability of geothermal energy in the Subei Basin calculated using the MaxEnt model is as follows: Figure 6 The favorable areas are distributed in the central part of Yancheng City in the middle of the basin, the southeastern coastal area, Taizhou City in the south, and the southwest of Huai'an City.

[0092] The locations of 16 known hot springs distributed in the Subei Basin are superimposed on the calculated results (e.g. Figure 6), the results showed that 13 hot springs were located in the medium or above area, accounting for 81.25%, of which 2 were excellent, accounting for 12.5%, 3 were good, accounting for 18.75%, 7 were average, accounting for 43.75%, and 1 was poor, accounting for 0.625%. Only 3 hot springs were in the extreme area of ​​the calculated results.

[0093] In addition, the "Sure 1 Well" is the first hot dry rock exploration verification well in Jiangsu Province. The well was completed at the end of January 2021 with a depth of about 5 kilometers. The carbonate thermal reservoir was successfully discovered, and its location is in the high value area of ​​the calculated results (such as Figure 6 ).

[0094] The above location distribution shows that the calculation results of the deep geothermal resource potential in the Subei Basin in this paper are consistent with the known hot spring locations and hot dry rock drilling locations in the study area, which also shows that the prediction results of this paper are highly reliable.

[0095] The above is only a preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications should also be considered as the protection scope of the present invention. The structures, devices and operating methods not specifically described and explained in the present invention shall be implemented according to the conventional means in the art unless otherwise specified and limited.

Claims

1. An index system and method model for deep geothermal resource potential evaluation and sweet spot optimization, characterized in that: The following steps are involved: S1: Construct an evaluation index system based on the deep geothermal geological characteristics; S2: Construct a deep geothermal resource potential evaluation method based on the MaxEnt model.

2. The index system and method model for deep geothermal resource potential evaluation and sweet spot optimization according to claim 1, characterized in that: The evaluation index system in S1 includes the following index parameters: S11: geophysical parameters reflecting the composition of deep heat sources; S12: An indicator reflecting the structural geological conditions and magmatic activity of deep strata; S13: Indicative indicators of deep geothermal heat.

3. The index system and method model for deep geothermal resource potential evaluation and sweet spot optimization according to claim 1, characterized in that: Given a random variable X in S2, its information entropy is: Given a training dataset Define the characteristic function:

4. The index system and method model for deep geothermal resource potential evaluation and sweet spot optimization according to claim 3, characterized in that: The random variable X and the target Y are binary functions that are 1 when they both satisfy a certain condition and 0 otherwise. The empirical distribution of the joint distribution P(x, y) for: Where N is the number of samples in the training data set, which is used to calculate the empirical distribution; From the above formula, we can get the empirical distribution of the marginal distribution P(x): for: At this point, the mathematical expectation of the characteristic function in the training data with respect to the empirical distribution is:

5. The index system and method model for deep geothermal resource potential evaluation and sweet spot optimization according to claim 4, characterized in that: The mathematical expectation of the joint distribution of features obtained from formula (5) is: Furthermore, according to the constraint get:

6. The index system and method model for deep geothermal resource potential evaluation and sweet spot optimization according to claim 5, characterized in that: If there are n characteristic functions, there are n constraints; Then, we use C to represent the set that satisfies the constraints: Therefore, the MaxEnt model is found in the model set C that satisfies the constraints and maximizes the entropy of the conditional distribution P(Y|X). In formula (8), the entropy of the conditional distribution P(Y|X) is expressed as: Therefore, the model is expressed as:

Images