A data-driven prospecting prediction method, device, medium and equipment

Through the combination of deep learning and Gaussian mixed model, mineral prediction under complex geological conditions is achieved, the misjudgment problems caused by single background value and artificial kernel function selection in the existing technology are solved, and the accuracy of mineral exploration prediction is improved and automation level.

CN118643408BActive Publication Date: 2024-11-05INST OF GEOPHYSICAL & GEOCHEMICAL EXPLORATION CHINESE ACAD OF GEOLOGICAL SCI
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411102726.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-13
Publication Date
2024-11-05
Estimated Expiration
2044-08-13

AI Technical Summary

Technical Problem

The existing mineral exploration prediction methods have the differences caused by the existence of a single background value under complex geological conditions, resulting in misjudgment and the selection of artificial nuclear functions, making it difficult to achieve accurate mineral exploration.

Method used

Deep learning method is adopted to obtain geological, geochemical and remote sensing data for grid processing, combine it with Gaussian hybrid model to identify spatial structural anomalies, train the prediction model for automated mineral exploration and prediction, and reduce human intervention.

Benefits of technology

More accurate and complete mineral prediction results are achieved, the misjudgment rate is reduced, and the automation level of mineral exploration prediction is improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118643408B_ABST
    Figure CN118643408B_ABST
Patent Text Reader

Abstract

The present application relates to the field of prospecting prediction technology, and is a data-driven prospecting prediction method, device, medium and equipment. The acquired geological data is processed to extract characteristic data related to prospecting, and the characteristic data is passed through a trained prediction model to obtain a prediction result, wherein the prediction result is used to characterize the probability of the existence of a target mine in the geology. The method provided in the embodiment of the present application can realize automated prospecting prediction, and its prediction results are more accurate and complete than those of the prior art.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of prospecting and prediction, and is a data-driven prospecting and prediction method, device, medium and equipment. Background Art

[0002] Mineral resources are an indispensable material basis for human survival and social and economic development. Exploration geochemistry is an important branch of applied geochemistry. It has the characteristics of systematically measuring the spatial variation of elements. Mineral resources are composed of elements. Exploration geochemistry plays an irreplaceable role in mineral exploration practice. Identification and extraction of geochemical anomalies are the main goals of geochemical exploration. By separating the geochemical background and geochemical anomalies, information related to mineral exploration in geochemical data can be mined. However, as the shallow mines on the surface are exhausted, the difficulty of mineral exploration is increasing, and the situation of mineral exploration is becoming more severe. Summary of the invention

[0003] In order to solve the above technical problems, the embodiments of the present application provide a data-driven prospecting prediction method, device, medium and equipment, which can realize automated prospecting prediction through deep learning methods.

[0004] In order to achieve the above purpose, the technical solution adopted in the embodiment of the present application is as follows:

[0005] In a first aspect, a data-driven prospecting prediction method is provided, the method comprising: acquiring geological data, elemental geochemical data, geophysical data, and remote sensing data in a target area, and gridding the elemental geochemical data, geophysical data, and remote sensing data to obtain a geophysical grid map, a geochemical grid map, and a remote sensing grid map; the geological data is geological map structure data formed by processing a geological map and supplementing the attribute information of geological elements, and the geological data is exported based on a MapGIS plain code format; the attribute information includes one or more of geological material composition, structure, and construction; the geophysical grid map, the geochemical grid map, and the remote sensing grid map are integrated into the geological data to obtain a geological grid map, the geological grid map is input into a prediction model that is in a convergent state after training to obtain a prospecting prediction classification map, and the prospecting prediction classification map is projected onto a geological map of the target area to delineate a prospecting prediction area; the prospecting prediction classification map is marked with probabilities corresponding to different mineral types.

[0006] Furthermore, the elemental geochemical data includes geochemical measurement data of abnormal elements in the target area; the geophysical data includes geomagnetic measurement data in the target area; and the remote sensing data includes remote sensing images, on which corresponding address interpretations are configured.

[0007] Furthermore, the elemental geochemical data, geophysical data and remote sensing data are gridded, including: dividing different types of data in the elemental geochemical data, geophysical data and remote sensing data into multiple feature layers, and gridding each of the feature layers in a 0, 1 manner to obtain the geophysical grid map, the geochemical grid map and the remote sensing grid map.

[0008] Furthermore, the prediction model includes an encoder, 4 fully connected layers, 1 Softmax layer and a decoder.

[0009] Furthermore, the encoder includes a convolution layer, a batch normalization layer, a maximum pooling layer and a Relu layer, and the decoder includes a deconvolution layer and a Relu activation function layer.

[0010] Furthermore, the training method of the prediction model enhances the data by means of window transformation to obtain enhanced data samples, identifies spatial structural anomalies of the data samples through a Gaussian mixture model and calculates anomaly scores, and determines score classifications of the anomaly scores based on a preset anomaly score threshold, wherein the score classifications include a first anomaly score and a second anomaly score, wherein the first anomaly score is a score classification higher than the anomaly score threshold, and the second anomaly score is a score classification lower than the anomaly score threshold, and the samples corresponding to the first anomaly score are marked as positive samples, and the samples corresponding to the second anomaly score are marked as negative samples; and the initial prediction model is trained based on the positive samples and the negative samples.

[0011] Furthermore, the Gaussian mixture model is a linear combination of multiple Gaussian models, each of which has a corresponding weight. The data sample is identified with a spatial structure anomaly through a Gaussian mixture model and an anomaly score is calculated, including: determining multiple initial key parameters of the Gaussian mixture model based on maximizing the following log-likelihood function, and iterating the multiple initial key parameters based on expected maximization to obtain multiple target key parameters, determining a probability value corresponding to the data sample based on the multiple target key parameters, and comparing the probability value with a preset probability threshold, determining that the corresponding data sample has a spatial structure anomaly when the probability value is less than the probability threshold, and obtaining a difference between the probability value and the maximum probability value, and the difference is the anomaly score.

[0012] In a second aspect, a data-driven mineral prospecting prediction device is provided, the device comprising: a data processing module for acquiring geological data, elemental geochemical data, geophysical data, and remote sensing data in a target area, and gridding the elemental geochemical data, geophysical data, and remote sensing data to obtain a geophysical grid map, a geochemical grid map, and a remote sensing grid map; a prediction module for fusing the geophysical grid map, the geochemical grid map, and the remote sensing grid map into the geological data to obtain a geological grid map, inputting the geological grid map into a prediction model that is in a convergent state after training to obtain a mineral prospecting prediction classification map, and projecting the mineral prospecting prediction classification map onto a geological map of the study area to delineate a mineral prospecting prediction area.

[0013] In a third aspect, a computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the steps of the data-driven mineral prospecting prediction method as described in any one of the above are implemented.

[0014] In a third aspect, a terminal device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the data-driven mineral prospecting prediction method as described in any one of the above items when executing the computer program.

[0015] In the technical solution provided in the embodiment of the present application, the acquired geological data is processed to extract characteristic data related to mineral exploration, and the characteristic data is passed through a trained prediction model to obtain a prediction result, wherein the prediction result is used to characterize the probability of the existence of the target mine in the geology. The method provided in the embodiment of the present application can realize automated mineral exploration prediction, and its prediction results are more accurate and complete than those of the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0017] The methods, systems and / or programs in the accompanying drawings will be further described according to exemplary embodiments. These exemplary embodiments will be described in detail with reference to the drawings. These exemplary embodiments are non-limiting exemplary embodiments, wherein example numbers represent similar mechanisms in the various views of the accompanying drawings.

[0018] Figure 1 It is a flow chart of the data-driven mineral prospecting prediction method provided in the embodiment of the present application.

[0019] Figure 2 It is a schematic diagram of the structure of a data-driven mineral prospecting prediction device provided in an embodiment of the present application.

[0020] Figure 3 It is a schematic diagram of the terminal device structure provided in an embodiment of the present application. DETAILED DESCRIPTION

[0021] In order to better understand the above technical scheme, the technical scheme of the present application is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical scheme of the present application, rather than limitations on the technical scheme of the present application. In the absence of conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other.

[0022] In the following detailed description, numerous specific details are set forth by way of example in order to provide a comprehensive understanding of the relevant guidance. However, it will be apparent to those skilled in the art that the present application may be practiced without these details. In other cases, well-known methods, procedures, systems, compositions and / or circuits have been described at a relatively high level, without detail, in order to avoid unnecessarily obscuring aspects of the present application.

[0023] Flowcharts are used in the present application to illustrate the execution process performed by the system according to the embodiment of the present application. It should be clearly understood that the execution process of the flowchart may not be performed in order. On the contrary, these execution processes may be performed in reverse order or simultaneously. In addition, at least one other execution process may be added to the flowchart. One or more execution processes may be deleted from the flowchart.

[0024] Before further describing the embodiments of the present invention in detail, the nouns and terms involved in the embodiments of the present invention are described. The nouns and terms involved in the embodiments of the present invention are subject to the following interpretations.

[0025] (1) In response to, it is used to indicate the conditions or states on which the executed operation depends. When the dependent conditions or states are met, one or more operations executed may be in real time or with a set delay. Unless otherwise specified, there is no restriction on the order in which the multiple operations executed are executed.

[0026] (2) Based on: used to indicate the conditions or states on which the executed operations depend. When the dependent conditions or states are met, one or more operations executed may be in real time or with a set delay. Unless otherwise specified, there is no restriction on the order in which the multiple operations executed are executed.

[0027] Geochemical exploration has been one of the most commonly used and effective methods of geological and mineral exploration since its birth in the 1930s, and has played a key role in mineral exploration. In order to identify geochemical anomalies, domestic and foreign scholars have done a lot of exploration. In the existing technology, there are mainly two technical routes for mineral exploration prediction methods: single element identification and multi-element identification.

[0028] Among them, the single-element geochemical anomaly identification method has some defects. First, a single background value is used in a large area, and second, the data must meet certain assumptions. Some scholars believe that when the geological conditions are more complex, a unified background value for the entire area will lead to areas with high element background values ​​being identified as anomalies. The core of the multi-element geochemical statistical method is to use the kernel function to map the original data to a high-dimensional feature space, and then separate the background and anomalies in the high-dimensional feature space. However, the current selection of kernel functions still relies on human selection, and different kernel functions will lead to large differences. This type of method has a large degree of artificiality.

[0029] Therefore, in response to the above background technical problems, an embodiment of the present application provides a mineral prospecting prediction method based on multi-source data of a neural network. This method uses geological characteristics to perform comprehensive multi-information mineral prospecting prediction. It is necessary to vectorize the geological information, and then use a deep learning algorithm to integrate the geological information with geochemical and geophysical data for mineral prospecting prediction.

[0030] See also Figure 1 , the method provided in the embodiment of the present application comprises the following steps:

[0031] Step S110. Acquire geological data, elemental geochemical data, geophysical data, and remote sensing data in the target area, and grid the elemental geochemical data, geophysical data, and remote sensing data to obtain a geophysical grid map, a geochemical grid map, and a remote sensing grid map.

[0032] In an embodiment of the present application, the geological data is geological map structure data formed by processing the geological map and supplementing the attribute information of the geological elements, and the geological data is exported based on the MapGIS plain code format. Among them, the attribute information for the geological elements includes one or more of the geological material composition, structure, and structure. The element geochemical data includes the geochemical measurement data of the abnormal elements in the target area, and the geophysical data includes the geomagnetic measurement data in the target area, which includes the polarization processing results, the upward extension processing result data, and the directional derivative processing result data. The remote sensing data includes remote sensing images, and the corresponding address interpretation is configured on the remote sensing images.

[0033] The method provided in the embodiment of the present application is based on deep learning, and its prospecting prediction is carried out on the basis of gridding of different types of data. Different gridding methods should be used for different types of data. Gridding of geological maps refers to the process of segmenting existing geological maps according to the same grid. The amount of data formed in the process of gridding of geological maps is very huge, and often thousands or even tens of thousands of dimensions of grid data can be formed. In order to improve computing efficiency, data screening and principal component analysis are needed to process and compress the above data.

[0034] Among them, the gridding processing of element geochemical data, geophysical data and remote sensing data is carried out by dividing different types of data in the element geochemical data, geophysical data and remote sensing data into multiple feature layers, and gridding each feature layer in the form of 0 and 1 to obtain the geophysical grid map, the geochemical grid map and the remote sensing grid map. If there are k feature layers, the characteristics of the same grid unit can be represented by a one-dimensional vector, such as [0,1,0,0,0,1,1,1,1,1…]. A grid feature consisting of m rows and n columns can be represented by a three-dimensional array of m×n×k. In this way, the main information considered to be related to mineralization or prospecting can be used, but all the information contained in the geological map cannot be fully utilized. The binary representation method uses a series of layers to represent different geological attributes. For data compression and processing, geological elements can be classified according to their characteristics, and the characteristics of geological elements can be superimposed, thereby reducing data compression. For example, if there are 50 faults in a prediction area, 20 of which are NE-directed faults, 10 are EW-directed faults, and 20 are SN-directed faults, the fault grid layers in the same direction can be superimposed to form three grid layers corresponding to the fault structures. After this processing, the original 50 grid layers can be compressed into three grid layers, greatly reducing the amount of data.

[0035] Through the above processing, the features required for subsequent prediction models are extracted to obtain the geophysical grid maps, the geochemical grid maps, and the remote sensing grid maps corresponding to the above different types.

[0036] Step S120. Integrate the geophysical grid map, the geochemical grid map, and the remote sensing grid map into the geological data to obtain a geological grid map, input the geological grid map into a prediction model that is in a converged state after training to obtain a mineral prospecting prediction classification map, and project the mineral prospecting prediction classification map onto the geological map of the study area to delineate the mineral prospecting prediction area.

[0037] In the embodiment of the present application, the probabilities corresponding to different mineral types are marked on the mineral prospecting prediction classification map.

[0038] In step S110, the geophysical grid map, the geochemical grid map, and the remote sensing grid map corresponding to different types of data are obtained, and the geophysical grid map, the geochemical grid map, the remote sensing grid map and the geological map are fused to obtain the final geological grid map, and the geological grid map is input into the prediction model in a converged state after training.

[0039] Among them, the basic structure of the prediction model includes an encoder, 4 fully connected layers, 1 Softmax layer and a decoder.

[0040] Specifically, the encoder in the embodiment of the present application includes a convolution layer, a batch normalization layer, a maximum pooling layer and a Relu layer, and the decoder includes a deconvolution layer and a Relu activation function layer. The convolution layer mainly extracts input data, and the convolution operation is implemented by the following formula: ,in represents the row and column coordinates in the input data, w represents the convolution kernel in the convolution operation, and b is the bias. After the convolution operation, the output features are activated by executing the RELU function, retaining the positive values ​​in the calculation results and setting the negative values ​​to 0, thereby avoiding overfitting in the training process, and x is the feature map.

[0041] The pooling layer is used to reduce the dimension of the feature map. Generally, the maximum pooling is used to retain the typical features in the feature map, while reducing the data dimension and the amount of calculation. The pooling operation is mainly expressed by the following formula: ,in represents the jth region in the feature map x, and i represents the feature value in the jth region.

[0042] The BN layer is usually placed after the convolution layer and before the activation function to readjust the data distribution of each batch and normalize the activation function.

[0043] The decoder consists of a deconvolution layer and a Relu activation function. The encoder maps the input features from top to bottom into the potential feature space, while the decoder maps the feature space back to the input space, hoping to reconstruct the original input data. Deconvolution is also called transposed convolution, which is the inverse process of convolution. The size of the input data for the decoder is , containing 15 geochemical element data in total, and each element data is used as a data channel. The network autoencoder used in the embodiment of the present application is composed of 4 convolutional layers, 4 BN layers and 3 maximum pooling layers, the decoder is composed of 4 deconvolutional layers, and the decoder and the encoder are connected using 1 convolutional layer.

[0044] After multiple convolution and pooling operations, the feature dimension extracted by the network model is reduced. By setting a fully connected layer, the feature matrix is ​​expanded into a one-dimensional vector, and the feature matrix is ​​mapped to the classification sample space to obtain the probability value of the input data in each classification. The index value with the maximum probability is the current classification. The fully connected operation is represented by the following formula: , where m, n, d represent the width, height and layer value of the input feature map respectively, w represents the weight value, b represents the bias, and i represents the i-th value of the one-dimensional vector output by the fully connected layer.

[0045] In an embodiment of the present application, a training method for a prediction model uses a window transformation method to enhance the data to obtain enhanced data samples, and uses a Gaussian mixture model to identify spatial structural anomalies of the data samples and calculate anomaly scores, and marks high anomaly score samples as positive samples and low anomaly score samples as negative samples; the initial prediction model is trained based on the positive samples and the negative samples.

[0046] Among them, window shift is used for window transformation, which is equivalent to observing the geochemical characteristics of the deposit from different positions. If the deposit is located in a grid cell, it is offset in 8 directions such as N, NE, E, SE, S, SW, W, and NW according to the window size, and a training data is generated for each offset of 1 cell. You can set different offset sizes, such as 1, 2...n, etc. If a window includes 10×10 grid cells, 100 learning data can be generated through window shift. In this way, multiple learning data for training are obtained to construct the corresponding data set.

[0047] The data set is classified again through the Gaussian mixture model, where the purpose of data classification is to screen each sample data in the data set into positive samples and negative samples. The screening logic is to use the Gaussian mixture model to identify the spatial structural anomaly of each sample data and calculate the anomaly score, marking the samples with high anomaly scores as positive samples and the samples with low anomaly scores as negative samples.

[0048] Among them, the Gaussian mixture model is a linear combination of multiple Gaussian models, based on the following formula: , where L is the number of Gaussian models, The weight of each Gaussian model satisfies and , Indicates that the sample belongs to The probability of a Gaussian model.

[0049] The Gaussian mixture model has three unknown parameters, namely , and , the above parameters are the initial key parameters, which can be determined by maximizing the following log-likelihood function: . Because the above formula contains the logarithm of and, it is very difficult to use the maximum derivative, so the expected maximization algorithm is generally used for iterative solution to obtain the final target key parameters. And update the Gaussian mixture model, specifically, use an n*L matrix to represent the unobserved data of L components, Indicates the relationship between the i-th sample and the The contribution rate of a Gaussian distribution, Meet the following conditions ,and Finally, the target key parameters are obtained according to the calculation , , The prior probability can be determined That is The weight of the Gaussian model : .

[0050] Among them, based on the calculated target key parameters and the above-mentioned Gaussian mixture model, the probability value corresponding to each sample data can be calculated, wherein the abnormal probability calculation of the sample data for the Gaussian mixture model can be performed using an existing processing algorithm, and its processing logic is based on the mean error and covariance calculated by the Gaussian mixture model, which will not be repeated in the embodiments of the present application. And compare the probability value with a preset probability threshold, and when the probability value is less than the probability threshold, it is determined that the corresponding data sample has a spatial structure abnormality. Specifically, based on the target key parameters , , The probability value of each sample can be determined, and the size of the probability value can reflect the abnormality of the spatial structure corresponding to the sample data.

[0051] The larger the probability value, the more the sample obeys the multivariate Gaussian distribution, and the smaller the probability value, the more likely the sample is abnormal. In order to more clearly represent the abnormality, define To express The degree of abnormality is the corresponding abnormality score: , The larger the value of is, the more likely the sample is abnormal. is the abnormality score, is the maximum probability value among multiple probability values ​​in multiple sample data. is the probability value of the sample data.

[0052] The abnormal score obtained is used to screen positive samples and negative samples, wherein the abnormal score is compared with a preset abnormal score threshold to determine the score classification of the abnormal score. In an embodiment of the present application, the score classification includes a first abnormal score and a second abnormal score, wherein the first abnormal score is a score classification higher than the abnormal score threshold, and the second abnormal score is a score classification lower than the abnormal score threshold, and the sample corresponding to the first abnormal score is marked as a positive sample, and the sample corresponding to the second abnormal score is marked as a negative sample.

[0053] In the embodiment of the present application, the positive samples and negative samples required for training are obtained through the above-mentioned processing, and the model training is converged based on the process of network training, and the final prediction model is obtained, and the classification result obtained by the model is obtained through the trained prediction model to delineate the prospecting prediction area of ​​the target area. When delineating, the classification map obtained by the prediction is redrawn and colored, and the transparency is adjusted and superimposed on the geological map of the target area to complete the projection of the prediction results on the geological map. The connected or adjacent prediction areas are circled together with dotted circles according to the spatial position, and the numbering is convenient for the next step of reasoning and analysis. When evaluating, the ontology model of the target area is first constructed, and then knowledge reasoning is performed to dig out the implicit knowledge therein, and the evaluation work is completed after summarization. The prediction results are projected onto the geological map of the target area according to the above steps, and the position of the delineated prediction area can be intuitively seen.

[0054] See also Figure 2 , a prospecting prediction device 200 based on multi-source data is provided, and the device 200 comprises:

[0055] The data processing module 210 is used to obtain geological data, element geochemical data, geophysical data, and remote sensing data in the target area, and perform grid processing on the element geochemical data, geophysical data, and remote sensing data to obtain a geophysical grid map, a geochemical grid map, and a remote sensing grid map;

[0056] The prediction module 220 is used to integrate the geophysical grid map, the geochemical grid map, and the remote sensing grid map into the geological data to obtain a geological grid map, input the geological grid map into the prediction model that is in a convergent state after training to obtain a mineral prospecting prediction classification map, and project the mineral prospecting prediction classification map onto the geological map of the study area to delineate the mineral prospecting prediction area.

[0057] See also Figure 3, the above method can also be integrated into the provided terminal device 300, and the device may have relatively large differences due to different configurations or performances, and may include one or more processors 301 and memory 302, and the memory 302 may store one or more storage applications or data. Among them, the memory 302 can be a temporary storage or a permanent storage. The application stored in the memory 302 may include one or more modules (not shown in the figure), and each module may include a series of computer executable instructions in the terminal device. Furthermore, the processor 301 can be configured to communicate with the memory 302, and the terminal device executes a series of computer executable instructions in the memory 302. The terminal device may also include one or more power supplies 303, one or more wired / wireless network interfaces 304, one or more input / output interfaces 305, one or more keyboards 306, etc.

[0058] In a specific embodiment, the terminal device includes a memory and one or more programs, wherein the one or more programs are stored in the memory, and the one or more programs may include one or more modules, and each module may include a series of computer executable instructions in the terminal device, and the one or more programs are configured to be executed by one or more processors, including computer executable instructions for performing the following:

[0059] Acquire geological data, element geochemical data, geophysical data, and remote sensing data in the target area, and grid the element geochemical data, geophysical data, and remote sensing data to obtain a geophysical grid map, a geochemical grid map, and a remote sensing grid map;

[0060] The geophysical grid map, the geochemical grid map, and the remote sensing grid map are integrated into the geological data to obtain a geological grid map, the geological grid map is input into a prediction model that is in a converged state after training to obtain a mineral prospecting prediction classification map, and the mineral prospecting prediction classification map is projected onto a geological map of the study area to delineate a mineral prospecting prediction area.

[0061] The following is a detailed introduction to the various components of the processor:

[0062] In this embodiment, the processor is an application specific integrated circuit (ASIC), or is configured to implement one or more integrated circuits of the embodiments of the present application, such as one or more microprocessors (digital signal processor, DSP), or one or more field programmable gate arrays (field programmable gate array, FPGA).

[0063] Optionally, the processor may execute various functions by running or executing a software program stored in the memory and calling data stored in the memory, such as executing the above-mentioned Figure 2 The method shown.

[0064] In a specific implementation, as an embodiment, the processor may include one or more microprocessors.

[0065] Among them, the memory is used to store the software program that executes the solution of the present application, and the execution is controlled by the processor. The specific implementation method can refer to the above method embodiment and will not be repeated here.

[0066] Optionally, the memory may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory may be integrated with the processor, or may exist independently and be coupled to the processing unit through the interface circuit of the processor, which is not specifically limited in the embodiments of the present application.

[0067] It should be noted that the structure of the processor shown in this embodiment does not constitute a limitation on the device, and the actual device may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0068] In addition, the technical effects of the processor can refer to the technical effects of the method described in the above method embodiment, which will not be repeated here.

[0069] It should be understood that the processor in the embodiments of the present application may be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0070] It should also be understood that the memory in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0071] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0072] In this application, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0073] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0074] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0075] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0076] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, 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.

[0077] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0078] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0079] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0080] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A data-driven prospecting prediction method, characterized in that: The method comprises: Acquire geological data, element geochemical data, geophysical data, and remote sensing data in the target area, and grid the element geochemical data, geophysical data, and remote sensing data to obtain geophysical grid maps, geochemical grid maps, and remote sensing grid maps; the geological data is geological map structure data formed by processing geological maps and supplementing the attribute information of geological elements in the geological maps, and the geological data is exported based on the MapGIS plain code format; the attribute information includes one or more of geological material composition, structure, and structure; the element geochemical data includes geochemical measurement data of abnormal elements in the target area; the geophysical data includes geomagnetic measurement data in the target area; the remote sensing data includes remote sensing images, and corresponding address interpretations are configured on the remote sensing images; The geophysical grid map, the geochemical grid map, and the remote sensing grid map are integrated into the geological data to obtain a geological grid map, the geological grid map is input into a prediction model that is in a convergent state after training to obtain a mineral prospecting prediction classification map, and the mineral prospecting prediction classification map is projected onto a geological map of a target area to delineate a mineral prospecting prediction area; the mineral prospecting prediction classification map is marked with probabilities corresponding to different mineral types; The prediction model includes an encoder, 4 fully connected layers, 1 Softmax layer and a decoder. The training method of the prediction model enhances the data by window transformation to obtain enhanced data samples. The data samples are identified with a Gaussian mixture model for spatial structural anomalies and anomaly scores are calculated. The score classification of the anomaly scores is determined based on a preset anomaly score threshold. The score classification includes a first anomaly score and a second anomaly score. The first anomaly score is a score classification higher than the anomaly score threshold, and the second anomaly score is a score classification lower than the anomaly score threshold. The samples corresponding to the first anomaly score are marked as positive samples, and the samples corresponding to the second anomaly score are marked as negative samples. The initial prediction model is trained based on the positive samples and the negative samples. The Gaussian mixture model is a linear combination of multiple Gaussian models, each of which has a corresponding weight. The data sample is identified with a spatial structure anomaly through the Gaussian mixture model and an anomaly score is calculated, including: determining multiple initial key parameters of the Gaussian mixture model based on maximizing the following log-likelihood function, iterating the multiple initial key parameters based on expectation maximization to obtain multiple target key parameters, determining a probability value corresponding to the data sample based on the multiple target key parameters, and comparing the probability value with a preset probability threshold. When the probability value is less than the probability threshold, it is determined that the corresponding data sample has a spatial structure anomaly, and obtaining a difference between the probability value and the maximum probability value, where the difference is the anomaly score; The Gaussian mixture model is a linear combination of multiple Gaussian models, which is expressed based on the following formula: , where L is the number of Gaussian models, The weight of each Gaussian model satisfies and , Indicates that the sample belongs to The probability of a Gaussian model; the multiple initial key parameters of the Gaussian mixture model are , and , determined by maximizing the following log-likelihood function: .

2. The data-driven prospecting prediction method according to claim 1, characterized in that: The elemental geochemical data, geophysical data and remote sensing data are gridded, including: dividing different types of data in the elemental geochemical data, geophysical data and remote sensing data into multiple feature layers, and gridding each of the feature layers in a 0, 1 manner to obtain the geophysical grid map, the geochemical grid map and the remote sensing grid map.

3. The data-driven prospecting prediction method according to claim 1, characterized in that: The encoder includes a convolution layer, a batch normalization layer, a maximum pooling layer and a Relu layer, and the decoder includes a deconvolution layer and a Relu activation function layer.

4. A data-driven prospecting prediction device, used to implement the data-driven prospecting prediction method according to any one of claims 1 to 3, characterized in that: The device comprises: A data processing module is used to obtain geological data, elemental geochemical data, geophysical data, and remote sensing data in the target area, and to perform grid processing on the elemental geochemical data, geophysical data, and remote sensing data to obtain a geophysical grid map, a geochemical grid map, and a remote sensing grid map; The prediction module is used to fuse the geophysical grid map, the geochemical grid map, and the remote sensing grid map into the geological data to obtain a geological grid map, input the geological grid map into a prediction model that is in a convergent state after training to obtain a mineral prospecting prediction classification map, and project the mineral prospecting prediction classification map onto a geological map of the target area to delineate a mineral prospecting prediction area.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the data-driven mineral prospecting prediction method as described in any one of claims 1 to 3 is implemented.

6. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the data-driven mineral prospecting prediction method according to any one of claims 1 to 3 is implemented.