Method, device and medium for mineral resource prediction based on three-dimensional geological model

Abstract

The application relates to the technical field of mineral resource prediction, in particular to a mineral resource prediction method and device based on a three-dimensional geological model and a medium, the method comprising the following steps: analyzing mineral resource data to obtain a mineral feature group and a time distribution sequence length; predicting historical mineral resource data based on the mineral feature group and the time sequence length to obtain future resource data and a resource location node; determining location change data according to the resource location node and the location node; analyzing current mineral resource data to obtain a current mineral feature group and current resource data; matching the current mineral feature group with the mineral feature group to obtain resource change data and the location change data; and obtaining updated current mineral resource data and the resource location node according to the resource change data and the location change data. The application helps to maintain the accuracy of mineral resource data predicted by the three-dimensional geological model.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of mineral resource prediction, in particular to a mineral resource prediction method based on a three-dimensional geological model, a device and a medium. BACKGROUND

[0002] With the acceleration of industrialization and urbanization and the continuous development of economic society, the situation of lack of mineral resources is increasingly apparent. If geological exploration is not strengthened and the mode of economic development is not changed, the support and carrying capacity of mineral resources for economic development will face great challenges, and will become a key factor restricting the sustainable development of economy.

[0003] In specific prospecting work, as surface mines, shallow mines and easily identifiable mines are increasingly decreasing, the difficulty of prospecting is increasing, and the effect of prospecting is decreasing. How to focus on finding deep concealed mines, new types of mines and new field mines has become a problem of concern to countries around the world.

[0004] At present, the use of three-dimensional models to carry out the search for deep concealed ore bodies (i.e., the second prospecting space) has become the main object of prospecting in many countries and regions, so the role of large-scale (usually greater than 1:50000) metallogenic prediction is more prominent, and it has become an important part of mineral deposit exploration. The resource prediction of concealed ore bodies in a mining area needs to be converted from two-dimensional plane prediction to three-dimensional space prediction, and the three-dimensional technical methods involved, such as acquisition of three-dimensional prospecting information, three-dimensional space visualization technology, and resource evaluation method based on three-dimensional space, are important topics in the current academic front. The research and development and experimental application of it have important theoretical and practical significance.

[0005] However, although the geological model for the scale of ore concentration area can improve the problem of difficult search for mineral resources to some extent, with the passage of time, the surface morphology of the mine and the resources of the concealed ore body are also prone to change. At this time, if the prediction data of the mineral resources of the three-dimensional geological model is not updated in time according to the changes of the resources of the concealed ore body, the accuracy of the mineral resource data predicted by the three-dimensional geological model will be reduced. SUMMARY

[0006] In order to solve at least one of the above technical problems, the present application provides a mineral resource prediction method based on a three-dimensional geological model, a device, an apparatus and a medium.

[0007] In a first aspect, the present application provides a mineral resource prediction method based on a three-dimensional geological model, which adopts the following technical solution:

[0008] obtain historical three-dimensional geological models and historical mineral resource data, the historical three-dimensional geological models are three-dimensional geological models of different geographical locations of minerals created in a historical period of time, and the historical mineral resource data is mineral resource data of different location nodes in the historical three-dimensional geological models in the historical period of time and data variation corresponding to the mineral resource data;

[0009] analyze the historical mineral resource data to obtain a mineral feature group of different location nodes in the historical mineral resource data and a time distribution sequence length of resource data corresponding to the mineral feature group;

[0010] predict the historical mineral resource data based on the mineral feature group and the time distribution sequence length to obtain future resource data of different mineral feature groups in a future preset time and resource location nodes corresponding to the future resource data;

[0011] determine resource variation data based on the future resource data and mineral resource data;

[0012] determine location variation data according to the resource location nodes and different location nodes in the historical mineral resource data;

[0013] obtain a current three-dimensional geological model and current mineral resource data, and analyze the current mineral resource data to obtain a current mineral feature group of different location nodes in the current mineral resource data and current resource data corresponding to the current mineral feature group, the current three-dimensional geological model is a three-dimensional geological model predicted and created according to a currently predicted mined mineral, and the current mineral resource data is predicted resource data of different location nodes in the current three-dimensional geological model;

[0014] match the current mineral feature group with the mineral feature group to obtain the resource variation data and the location variation data corresponding to the current mineral feature group;

[0015] update different location nodes in the current mineral resource data and the current three-dimensional geological model according to the resource variation data and the location variation data respectively to obtain updated current mineral resource data and resource location nodes corresponding to the current mineral resource data.

[0016] In a possible implementation manner, the prediction of the historical mineral resource data based on the mineral feature group and the time distribution sequence length to obtain future resource data of different mineral feature groups in a future preset time and resource location nodes corresponding to the future resource data comprises:

[0017] unsupervised time series data arrangement is performed on the historical mineral resource data based on the time distribution sequence length and the mineral characteristic group, to obtain resource matrix data;

[0018] The resource matrix data is input into a trained feature extraction model for vector feature extraction, to obtain a feature dimension number, and the obtained feature dimension number is combined with the resource matrix data for data processing, to generate future resource matrix data;

[0019] The data contained in the future resource matrix data is processed to obtain resource matrix data, and the obtained resource matrix data is input into a preset algorithm model for data calculation, to obtain future resource data of different mineral characteristic groups within a future preset time and resource location nodes corresponding to the future resource data.

[0020] In a possible implementation, the resource matrix data is input into a trained feature extraction model for vector feature extraction, to obtain a feature dimension number, including:

[0021] Based on the resource matrix data, the mineral change position, the mineral change data and the mineral change time after each change of the mineral resource data in the historical mineral resource data are determined;

[0022] The mineral change position, the mineral change data and the mineral change time are respectively input into the feature extraction model for vector feature extraction, to obtain a position vector feature corresponding to the mineral change position, a data vector feature corresponding to the mineral change data and a time vector feature corresponding to the mineral change time;

[0023] The position vector feature, the data vector feature and the time vector feature are counted to obtain the feature dimension number.

[0024] In a possible implementation, the obtained feature dimension number is combined with the resource matrix data for data processing, to generate future resource matrix data, including:

[0025] The feature dimension number is integrated with the resource matrix data to generate first matrix data;

[0026] The first matrix data is subjected to basic data distribution exploration, to obtain a relative periodicity rule of change of the mineral resource data in the historical mineral resource data, and a time period length is determined based on the relative periodicity rule;

[0027] The first matrix data is subjected to supervised time series data arrangement based on the time period length, to obtain second matrix data;

[0028] The second matrix data is used to predict the change trend of the mineral resources in a future preset time period, and future resource matrix data is generated.

[0029] In a possible implementation, the data processing on the data contained in the future resource matrix data comprises:

[0030] The normal distribution mean and the normal distribution variance of the resource data contained in the future resource matrix data are calculated, and a 3σ range of the future resource matrix data is determined based on the normal distribution mean and the normal distribution variance;

[0031] It is determined whether the resource data is outside the 3σ range, and if the resource data is outside the 3σ range, a first matrix sequence of the future resource matrix data in which the resource data is located is determined, a sequence mean is calculated according to the first matrix sequence, the resource data is replaced by the sequence mean to obtain a second matrix sequence after replacement, and missing value processing is performed on the second matrix sequence.

[0032] The second matrix sequence in the future resource matrix data is subjected to sequence normalization processing to obtain resource matrix data.

[0033] In a possible implementation, the method further comprises:

[0034] Actual resource change data and actual position change data detected in a future preset time period are determined for the current mineral feature group.

[0035] It is determined whether the actual resource change data and the actual position change are consistent with the resource change data and the position change data, respectively, and if not, the resource change data and the position change data are subjected to inverse normalization processing based on the actual resource change data and the actual position change data, and the resource change data and the position change data are restored to the actual resource change data and the actual position change data.

[0036] In a possible implementation, if the actual resource change data and the actual position change data are not consistent with the resource change data and the position change data, the resource change data and the position change data are subjected to inverse normalization processing based on the actual resource change data and the actual position change data, and the method further comprises:

[0037] A data root mean square error is determined based on the actual resource change data, the actual position change data, the resource change data, and the position change data.

[0038] According to the data root mean square error, parameters in an epoch training model in the feature extraction model are set, and the epoch training model after setting is iteratively updated in reverse to obtain an updated feature extraction model and a data validation set of each round in the feature extraction model.

[0039] The data validation set is calculated and evaluated to generate a loss value and an evaluation index of the data validation set.

[0040] In a second aspect, the present application provides a mineral resource prediction device based on a three-dimensional geological model, which adopts the following technical solution:

[0041] A mineral resource prediction device based on a three-dimensional geological model comprises:

[0042] A first acquisition module is configured to acquire a historical three-dimensional geological model and historical mineral resource data, the historical three-dimensional geological model being a three-dimensional geological model of minerals at different geographic locations within a historical period of time, and the historical mineral resource data being mineral resource data at different location nodes in the historical three-dimensional geological model within the historical period of time and a data change amount corresponding to the mineral resource data;

[0043] A data analysis module is configured to analyze the historical mineral resource data to obtain a mineral feature group at different location nodes in the historical mineral resource data and a time distribution sequence length of resource data corresponding to the mineral feature group;

[0044] A data prediction module is configured to predict the historical mineral resource data based on the mineral feature group and the time distribution sequence length to obtain future resource data of different mineral feature groups within a future preset time and resource location nodes corresponding to the future resource data;

[0045] A resource determination module is configured to determine resource change data based on the future resource data and the mineral resource data;

[0046] A location determination module is configured to determine location change data according to the resource location nodes and different location nodes in the historical mineral resource data;

[0047] A second acquisition module is configured to acquire a current three-dimensional geological model and current mineral resource data, and analyze the current mineral resource data to obtain a current mineral feature group at different location nodes in the current mineral resource data and current resource data corresponding to the current mineral feature group, the current three-dimensional geological model being a three-dimensional geological model predicted and created according to a currently predicted mineral to be mined, and the current mineral resource data being predicted resource data at different location nodes in the current three-dimensional geological model;

[0048] characteristics matching module, configured to match the current mineral resource characteristics group with the mineral resource characteristics group to obtain the resource change data and the position change data corresponding to the current mineral resource characteristics group;

[0049] a data updating module, configured to update different position nodes in the current mineral resource data and the current three-dimensional geological model according to the resource change data and the position change data respectively to obtain updated current mineral resource data and resource position nodes corresponding to the current mineral resource data.

[0050] In a possible implementation, when the data prediction module predicts the historical mineral resource data based on the mineral resource characteristics group and the time distribution sequence length to obtain future resource data of different mineral resource characteristics groups within a preset future time and resource position nodes corresponding to the future resource data, the data prediction module is specifically configured to:

[0051] perform unsupervised time sequence data arrangement on the historical mineral resource data based on the time distribution sequence length and the mineral resource characteristics group to obtain resource matrix data;

[0052] input the resource matrix data into a trained feature extraction model to perform vector feature extraction to obtain a feature dimension number, and perform data combination processing on the obtained feature dimension number and the resource matrix data to generate future resource matrix data;

[0053] perform data processing on data contained in the future resource matrix data to obtain resource matrix data, and input the obtained resource matrix data into a preset algorithm model to perform data calculation to obtain future resource data of different mineral resource characteristics groups within a preset future time and resource position nodes corresponding to the future resource data.

[0054] In another possible implementation, when the data prediction module inputs the resource matrix data into a trained feature extraction model to perform vector feature extraction to obtain a feature dimension number, the data prediction module is specifically configured to:

[0055] determine, based on the resource matrix data, a mineral change position, mineral change data and mineral change time after each time of data change of the mineral resource in the historical mineral resource data;

[0056] input the mineral change position, the mineral change data and the mineral change time into the feature extraction model respectively to perform vector feature extraction to obtain a position vector feature corresponding to the mineral change position, a data vector feature corresponding to the mineral change data and a time vector feature corresponding to the mineral change time;

[0057] The feature dimension quantity is obtained by counting the position vector feature, the data vector feature and the time vector feature.

[0058] In another possible implementation, when the data prediction module combines the obtained feature dimension quantity with the resource matrix data to generate future resource matrix data, the data prediction module is specifically configured to:

[0059] integrate the feature dimension quantity with the resource matrix data to generate first matrix data;

[0060] probe the first matrix data to obtain a relative periodicity rule of mineral resource change in the historical mineral resource data, and determine a time period length based on the relative periodicity rule;

[0061] perform supervised time series data arrangement on the first matrix data based on the time period length to obtain second matrix data;

[0062] predict a mineral resource change trend in a future preset time period based on the second matrix data to generate future resource matrix data.

[0063] In another possible implementation, when the data prediction module performs data processing on data contained in the future resource matrix data to obtain resource matrix data, the data prediction module is specifically configured to:

[0064] calculate a normal distribution mean and a normal distribution variance of resource data contained in the future resource matrix data, and determine a 3σ range of the future resource matrix data based on the normal distribution mean and the normal distribution variance;

[0065] determine whether the resource data is outside the 3σ range, if the resource data is outside the 3σ range, determine a first matrix sequence of the future resource matrix data in which the resource data is located, calculate a sequence mean value according to the first matrix sequence, replace the resource data with the sequence mean value to obtain second matrix sequence after replacement, and perform missing value processing on the second matrix sequence;

[0066] perform sequence normalization processing on the second matrix sequence in the future resource matrix data to obtain resource matrix data.

[0067] In another possible implementation, the apparatus further includes an actual data determination module and a data normalization module, wherein,

[0068] The actual data determining module is configured to determine actual resource change data and actual position change data detected by the current set of mineral characteristics within a preset future time period;

[0069] The data normalization module is configured to determine whether the actual resource change data and the actual position change data are consistent with the resource change data and the position change data, respectively, and if not, to perform inverse normalization processing on the resource change data and the position change data based on the actual resource change data and the actual position change data, so as to restore the resource change data and the position change data to the actual resource change data and the actual position change data.

[0070] In another possible implementation, the apparatus further includes an error determining module, a model updating module, and a data evaluation module, wherein,

[0071] The error determining module is configured to determine a data root mean square error based on the actual resource change data, the actual position change data, the resource change data, and the position change data.

[0072] The model updating module is configured to set parameters in an epoch training model in the feature extraction model according to the data root mean square error, and to perform reverse iterative updating on the set epoch training model to obtain an updated feature extraction model and a data validation set of each round of the feature extraction model.

[0073] The data evaluation module is configured to perform calculation and evaluation on the data validation set to generate a loss value and an evaluation index of the data validation set.

[0074] In a third aspect, the present application provides an electronic device, which adopts the technical scheme as follows:

[0075] at least one processor;

[0076] a memory;

[0077] at least one application program, wherein the at least one application program is stored in the memory and is configured to be executed by the at least one processor, and the at least one application program is configured to perform the mineral resource prediction method based on the three-dimensional geological model as described in any one of the first aspects.

[0078] In a fourth aspect, the present application provides a computer-readable storage medium, which adopts the technical scheme as follows:

[0079] A computer readable storage medium having stored thereon a computer program which, when executed in a computer, causes the computer to perform the method for predicting mineral resources based on a three-dimensional geological model according to any one of the first aspect.

[0080] In summary, the present application includes at least one of the following beneficial technical effects:

[0081] After the three-dimensional geological model is constructed according to the current ore concentration area scale region and mining is not carried out in the ore concentration area scale region for a short time, in order to maintain the prediction accuracy of the three-dimensional geological model of mineral resources, first, the historical three-dimensional geological model and the historical mineral resource data are obtained, wherein the historical three-dimensional geological model is a three-dimensional geological model created for mineral resources at different geographic locations within a historical period of time, and the historical mineral resource data is mineral resource data at different location nodes in the historical three-dimensional geological model within the historical period of time and a data change corresponding to the mineral resource data, then the mineral resource data is analyzed to obtain a mineral feature group at different location nodes in the historical mineral resource data and a time distribution sequence length of the resource data corresponding to the mineral feature group, then the historical mineral resource data is predicted based on the mineral feature group and the time distribution sequence length to obtain future resource data of different mineral feature groups within a future preset time and resource location nodes corresponding to the future resource data, then the resource change data is determined based on the future resource data and the mineral resource data, the location change data is determined according to the resource location nodes and the different location nodes in the historical mineral resource data, then the current three-dimensional geological model and the current mineral resource data are obtained, and the current mineral resource data is analyzed to obtain a current mineral feature group at different location nodes in the current mineral resource data and current resource data corresponding to the current mineral feature group, wherein the current three-dimensional geological model is a three-dimensional geological model created according to the predicted mining of the current mineral resources, and the current mineral resource data is predicted resource data at different location nodes in the current three-dimensional geological model, then the current mineral feature group is matched with the mineral feature group to obtain resource change data corresponding to the current mineral feature group and location change data, and the different location nodes in the current mineral resource data and the current three-dimensional geological model are updated according to the resource change data and the location change data, respectively, to obtain updated current mineral resource data and resource location nodes corresponding to the current mineral resource data, thereby obtaining the relative change rule of the mineral feature group and the time period of the field resource at different location nodes through the analysis of the historical mineral resource data, and then updating the data corresponding to the current field resource data, so that the prediction data of the current three-dimensional geological model can change according to the time period, achieving the effect of maintaining the prediction accuracy of the current three-dimensional geological model. BRIEF DESCRIPTION OF DRAWINGS

[0082] Figure 1A flowchart of a mineral resource prediction method based on a three-dimensional geological model is provided in the embodiments of the present application.

[0083] Figure 2 A structural diagram of a mineral resource prediction device based on a three-dimensional geological model is provided in the embodiments of the present application.

[0084] Figure 3 A structural diagram of an electronic device is provided in the embodiments of the present application. DETAILED DESCRIPTION

[0085] The embodiments of the present application will be described below in detail with reference to the accompanying drawings. Figures 1-3 The present application will be described in further detail.

[0086] The specific embodiments are merely illustrative of the present application, and are not intended to limit the present application. Those skilled in the art can make modifications to the embodiments without creative contribution, and the modifications are within the scope of the present application.

[0087] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative contribution are within the scope of the present application.

[0088] In addition, the term "and / or" in the present application is merely used to describe the association relationship of the associated objects, and can represent the existence of three relationships, for example, A and / or B can represent the existence of A alone, the existence of A and B together, and the existence of B alone. In addition, the character " / " in the present application generally represents an "or" relationship between the associated objects, unless otherwise specified.

[0089] The embodiments of the present application will be described in further detail below with reference to the accompanying drawings.

[0090] The embodiments of the present application provide a mineral resource prediction method based on a three-dimensional geological model, which is executed by an electronic device. The electronic device can be a server or a terminal device. The server can be a physical server, a server cluster composed of multiple physical servers, a distributed system, or a cloud server providing cloud computing services. The terminal device can be a smart phone, a tablet computer, a notebook computer, a desktop computer, etc., but is not limited thereto. The terminal device and the server can be directly or indirectly connected through wired or wireless communication, and the embodiments of the present application do not limit this. Figure 1 As shown in the figure, the method comprises:

[0091] Step S10, obtaining a historical three-dimensional geological model and historical mineral resource data.

[0092] The historical three-dimensional geological model is a three-dimensional geological model of mineral resources at different geographic locations in a historical period of time, and the historical mineral resource data is mineral resource data at different location nodes in the historical three-dimensional geological model in the historical period of time and a data change corresponding to the mineral resource data.

[0093] In the embodiment of the application, the historical period of time can be ten years, twenty years, and the specific value is input by the staff through the terminal device, for example: when the staff inputs twenty years through the terminal device, the electronic device as the execution subject of the application will collect all the mining data occurred in different regional geographic locations in twenty years, and then obtain the three-dimensional geological model constructed for the mining area before mining through the mining data, and the mineral resource data corresponding to different location nodes in the three-dimensional geological model and the data change corresponding to the mineral resource data at different time.

[0094] Specifically, the different location nodes represent the coordinate edge value corresponding to each mineral resource in the three-dimensional geological model, that is, each edge coordinate of the area where the mineral resource is located, and each edge coordinate is connected to form an irregular rectangle, and the irregular rectangle is regarded as a marker point. The specific position of the marker point in the three-dimensional geological model is detected, that is, the position node of the mineral resource is obtained.

[0095] Step S11, analyzing the historical mineral resource data to obtain a mineral feature group of different location nodes in the historical mineral resource data and a time distribution sequence length of resource data corresponding to the mineral feature group.

[0096] Specifically, the historical mineral resource data refers to the mineral resource data corresponding to the location nodes at different coordinates in the three-dimensional geological model and the data change corresponding to the mineral resource data in the historical period of time, wherein the mineral resource data includes a mineral feature group and a mineral determination time, and the mineral feature group includes mineral distribution characteristics, mineral genesis characteristics, mineral form characteristics, mineral content characteristics, and mineral location characteristics. The mineral determination time represents the time when the historical three-dimensional geological model is discovered and constructed. With the passage of time, the characteristic data corresponding to the mineral feature group will also change, and through the historical mineral resource data, the change of the characteristic data at different times after the mineral determination time can be more obviously perceived. The data amount of the change is the data change mentioned in the application.

[0097] In step S12, the historical mineral resource data is predicted based on the mineral feature group and the time distribution sequence length, to obtain future resource data of different mineral feature groups within a preset time in the future and resource location nodes corresponding to the future resource data.

[0098] Specifically, the historical mineral resource data is unsupervised time series data processing based on the time distribution sequence length and the mineral feature group, to obtain resource matrix data. The resource matrix data is input into the trained feature extraction model for vector feature extraction, to obtain the feature dimension number. The obtained feature dimension number is combined with the resource matrix data for data processing, to generate future resource matrix data. The data in the future resource matrix data is processed, to obtain resource matrix data. The obtained resource matrix data is input into a preset algorithm model for data calculation, to obtain future resource data of different mineral feature groups within a preset time in the future and resource location nodes corresponding to the future resource data.

[0099] Specifically, the time series is a set of random variables sorted by time. It is usually the result of observing a certain potential process according to a given sampling rate at equal intervals. The time series data essentially reflects the trend of one or more random variables changing over time, and the core of the time series prediction method is to mine this rule from the data and make estimates of future data.

[0100] In the embodiments of the present application, the time distribution sequence length represents the length of the change of the mineral resource data over time.

[0101] According to the time distribution sequence length and the mineral feature group, the historical mineral resource data is unsupervised time series data processing, to obtain the following resource matrix data:

[0102] .

[0103] Wherein, m is the mineral feature group, and n is the time distribution sequence length.

[0104] Specifically, before the resource matrix data is input into the feature extraction model for vector feature extraction, the matrix data sample needs to be obtained in advance. The matrix data sample includes the resource matrix data formed by all historical mineral resource data and the vector features in the resource matrix data. Then, the feature extraction model is created, and the feature extraction model is trained based on the matrix data sample, to obtain the trained feature extraction model.

[0105] Specifically, the feature extraction model is a pre-trained neural network model. Neural networks (Neural Networks, NN) are widely interconnected by a large number of simple processing units (called neurons).

[0106] The complex network system formed thereby reflects many basic characteristics of brain function and is a highly complex nonlinear dynamic learning system. The neural network has large-scale parallelism, distributed storage and processing, self-organization, self-adaptation and self-learning ability, and is particularly suitable for processing information processing problems that need to consider many factors and conditions at the same time, and are imprecise and fuzzy. The development of neural networks is related to neuroscience, mathematical science, cognitive science, computer science, artificial intelligence, information science, control theory, robotics, microelectronics, psychology, optical computing, molecular biology, etc.

[0107] Specifically, the resource matrix data is input into a feature extraction model for vector feature extraction, and the extracted vector features are counted to obtain a feature dimension number, wherein the vector features include position vector features, data vector features, and time vector features in historical mineral resource data, and then the feature dimension number is combined with the resource matrix data to obtain future resource matrix data.

[0108] For the embodiments of the present application, a bidirectional LSTM model is used as an example to illustrate the preset algorithm model, including but not limited to the bidirectional LSTM model.

[0109] Specifically, the preset algorithm model is constructed, and the model main body uses a bidirectional LSTM as a trend prediction model. The LSTM mainly has a forgetting gate, an input gate, and an output gate;

[0110] Forgetting gate:

[0111] Input gate:

[0112] After filtering the information through the forgetting gate and the input gate, the historical memory and the memory content of the current stage are combined, and the generated value is:

[0113]

[0114] Output gate:

[0115] After the above-described LSTM, a layer of LSTM network layer is accessed in reverse, and through this process, a BI-LSTM layer can be obtained, which is trained by a plurality of groups of mineral feature groups; an event space feature joint learning layer is added, an association vector matrix with a size of M*V*K is initialized, the output vector of the last layer of the LSTM is taken, the transpose is multiplied by the association vector parameter matrix, and finally a regression loss function is connected to complete the construction of the preset algorithm model.

[0116] Step S13, determining resource change data based on the future resource data and the mineral resource data.

[0117] ​Specifically, after learning the future resource data, the future resource data and the mineral resource data are differentially calculated according to the corresponding mineral characteristics to obtain characteristic change data corresponding to each mineral characteristic, and then the characteristic change data is integrated to obtain resource change data.

[0118] Step S14, determining position change data according to the resource position node and the different position nodes in the historical mineral resource data.

[0119] Specifically, after learning the resource position node, the resource position node and the corresponding position node in the historical mineral resource data are calculated based on the historical three-dimensional geological model to obtain three change vectors in X, Y and Z directions, and the three change vectors are integrated to obtain the position change data.

[0120] Step S15, obtaining a current three-dimensional geological model and current mineral resource data, and analyzing the current mineral resource data to obtain a current mineral characteristic group of different position nodes in the current mineral resource data and current resource data corresponding to the current mineral characteristic group.

[0121] The current three-dimensional geological model is a three-dimensional geological model created according to the prediction of the currently predicted mineral to be mined, and the current mineral resource data is predicted resource data located at different position nodes in the current three-dimensional geological model.

[0122] Step S16, matching the current mineral characteristic group with the mineral characteristic group to obtain resource change data and position change data corresponding to the current mineral characteristic group.

[0123] For the embodiments of the present application, when the current mineral characteristic group is matched with the mineral characteristic group, the proportion data change weight of different characteristics in the mineral characteristic group is analyzed, and then the proportion weight value of different characteristics in the different mineral characteristic groups is obtained, and then the current mineral characteristic group and the characteristics in the different mineral characteristic groups are one-to-one corresponding matching to obtain the matching characteristics of the different mineral characteristic groups and the current mineral characteristic group, the matching degree of the different mineral characteristic groups and the current mineral characteristic group is calculated according to the proportion weight value and the matching characteristics, and then the multiple matching degrees are compared to obtain the mineral characteristic group corresponding to the current mineral characteristic group, and the resource change data and the position change data are determined based on the mineral characteristic group.

[0124] Specifically, the proportion weight value of different features is closely related to the actual data of each mineral feature group, for example: there are two mineral features A and B, when the resource data and location change, the mineral distribution feature of A mineral feature group changes by 52%, the shape feature changes by 2%, and the content feature changes by 12%. Therefore, the proportion weight of the mineral distribution feature in the A mineral feature group is 78.78%, the proportion weight of the content feature is 3%, and the proportion weight of the content feature is 18.18%. However, when the resource data and location change, the mineral distribution feature of B mineral feature group changes by 32%, the shape feature changes by 12%, and the content feature changes by 22%. Therefore, the proportion weight of the mineral distribution feature in the B mineral feature group is 48.48%, the proportion weight of the content feature is 18.18%, and the proportion weight of the content feature is 33.33%.

[0125] Step S17, according to the resource change data and the location change data, the current mineral resource data and the different location nodes in the current three-dimensional geological model are updated respectively, and the updated current mineral resource data and the resource location nodes corresponding to the current mineral resource data are obtained.

[0126] Based on the above embodiment, after the three-dimensional geological model is constructed according to the current ore concentration area scale region and mining is not carried out in the ore concentration area scale region for a short time, in order to maintain the prediction accuracy of the mineral resources of the three-dimensional geological model, first, the historical three-dimensional geological model and the historical mineral resource data are obtained, wherein the historical three-dimensional geological model is a three-dimensional geological model of mineral resources at different geographical positions in a historical period of time, and the historical mineral resource data is mineral resource data at different position nodes in the historical three-dimensional geological model in the historical period of time and a data change corresponding to the mineral resource data, then the mineral resource data is analyzed to obtain a mineral feature group of the different position nodes in the historical mineral resource data and a time distribution sequence length of the resource data corresponding to the mineral feature group, then the historical mineral resource data is predicted based on the mineral feature group and the time distribution sequence length to obtain future resource data of different mineral feature groups in a future preset time and a resource position node corresponding to the future resource data, then the resource change data is determined based on the future resource data and the mineral resource data, the position change data is determined according to the resource position node and the different position nodes in the historical mineral resource data, then the current three-dimensional geological model and the current mineral resource data are obtained, and the current mineral resource data is analyzed to obtain a current mineral feature group of different position nodes in the current mineral resource data and current resource data corresponding to the current mineral feature group, wherein the current three-dimensional geological model is a three-dimensional geological model predicted and created according to the current mining mineral resources, and the current mineral resource data is predicted resource data at different position nodes in the current three-dimensional geological model, then the current mineral feature group is matched with the mineral feature group to obtain resource change data corresponding to the current mineral feature group and position change data, and the different position nodes in the current mineral resource data and the current three-dimensional geological model are updated according to the resource change data and the position change data respectively to obtain updated current mineral resource data and resource position nodes corresponding to the current mineral resource data, so that the relative change rule of the mineral feature group of the mine resource at different position nodes and the time period is obtained through the analysis of the historical mineral resource data, and the data corresponding to the current mine resource data is updated accordingly, so that the prediction data of the current three-dimensional geological model can be changed according to the time period, and the effect of maintaining the prediction accuracy of the current three-dimensional geological model is achieved.

[0127] In a possible implementation of the embodiment of the application, the resource matrix data is input into the trained feature extraction model to perform vector feature extraction to obtain the number of feature dimensions, specifically including:

[0128] The mineral change position, the mineral change data and the mineral change time after each mineral resource change in the historical mineral resource data is determined based on the resource matrix data, and the mineral change position, the mineral change data and the mineral change time are respectively input into a feature extraction model for vector feature extraction to obtain a position feature vector corresponding to the mineral change position, a data feature vector corresponding to the mineral change data and a time feature vector corresponding to the mineral change time. The position feature vector, the data feature vector and the time feature vector are subjected to quantity statistics to obtain the feature dimension quantity.

[0129] Specifically, the total number of feature vectors of a matrix is calculated as follows: the number = n - the rank of the feature matrix, and the number = n - r (into E-A). Where n is the order, and not every matrix can be diagonalized. If the eigenvalues of a matrix are all different, it can be diagonalized. The projection of the feature vector on the basis vector (i.e., the coordinates), assuming that the vector space is h-dimensional. Thus, it can be directly represented by a coordinate vector. Using the basis vector, linear transformation can also be represented by a simple matrix multiplication.

[0130] In one possible implementation of the embodiment of the present application, the obtained feature dimension quantity is combined with the resource matrix data for data processing to generate future resource matrix data, specifically including:

[0131] The feature dimension quantity is integrated with the resource matrix data to generate first matrix data, the first matrix data is subjected to basic data distribution exploration to obtain the relative periodicity law of the mineral resource change in the historical mineral resource data, and the time period length is determined based on the relative periodicity law, the first matrix data is subjected to supervised time series data arrangement based on the time period length to obtain second matrix data, the future mineral resource change trend in a preset time period is predicted based on the second matrix data, and future resource matrix data is generated.

[0132] Specifically, the feature dimension quantity is integrated as a dimension with the resource matrix data, and the pytorch technology is used for illustration in the embodiment of the present application, including but not limited to one possible implementation of the pytorch technology. The feature dimension quantity is added to the resource matrix data in the form of a dimension by the instruction "out. unsqueeze (-1)" in pytorch to realize dimension integration.

[0133] Specifically, PyTorch is a Torch-based Python open-source machine learning library for applications such as natural language processing. It is mainly developed by the artificial intelligence team of Facebook, which not only enables powerful GPU acceleration, but also supports dynamic neural networks, which is not supported by many mainstream frameworks such as TensorFlow. PyTorch provides two high-level functions: 1. Tensor calculation with powerful GPU acceleration (such as Numpy); 2. Deep neural networks containing automatic differentiation systems In addition to Facebook, institutions such as Twitter, GMU, and Salesforce have adopted PyTorch.

[0134] After integrating the number of feature dimensions with the resource matrix data, the following first matrix data is obtained:

[0135]

[0136] Wherein, v represents the number of feature dimensions.

[0137] Specifically, the first matrix data is imported into an Excel table, and a Python integrated jupyter environment is configured. The basic data distribution of the n time series in the first matrix data is probed, and the main purpose is to find the relative periodicity of the corresponding sequence. Then, the length of the time period is determined according to the relative periodicity.

[0138] Specifically, t is used to replace the length of the time periodicity, and n in the first matrix data is replaced by t to obtain new second matrix data:

[0139] .

[0140] Specifically, it is assumed that the preset future preset time period is k, that is, the moving step is k steps for prediction, and the future resource matrix data is obtained:

[0141] .

[0142] In one possible implementation of the embodiment of the application, the data contained in the future resource matrix data is processed to obtain resource matrix data, specifically including:

[0143] The normal distribution mean and the normal distribution variance of the resource data contained in the future resource matrix data are calculated, and a 3σ range of the future resource matrix data is determined based on the normal distribution mean and the normal distribution variance, whether the resource data is outside the 3σ range is judged, if the resource data is outside the 3σ range, a first matrix sequence of the future resource matrix data where the resource data is located is determined, a sequence mean is calculated according to the first matrix sequence, and the resource data is replaced by the sequence mean to obtain a second matrix sequence after replacement, and the second matrix sequence is processed for missing values, and the second matrix sequence in the future resource matrix data is processed for sequence normalization to obtain the resource matrix data.

[0144] Specifically, the 3σ range is based on the equal precision repeated measurement of the normal distribution, and the interference or noise of the singular data is difficult to meet the normal distribution. If the absolute value of the residual error of a measurement value in a group of measurement data is greater than 3σ, the measurement value is a bad value and should be removed. Generally, the error equal to ±3σ is regarded as the limit error. For the random error of the normal distribution, the probability of falling outside ±3σ is only 0.27%, which is very small in a limited number of measurements, so there is a 3σ criterion. The 3σ criterion is the most commonly used and the simplest gross error discrimination criterion, which is generally applied to the case where the number of measurements is sufficient (n≥30) or when n>10 is roughly discriminated.

[0145] Specifically, the missing value refers to the data clustering, grouping, deletion or truncation caused by the lack of information in the matrix sequence. The processing of the missing value is generally divided into deleting the case with missing value and imputing the missing value. The embodiment of the application deletes the case with missing value to process the matrix sequence. The simple deletion method and the weight method are mainly used to delete the case with missing value. The simple deletion method is the most original method for processing the missing value. It deletes the case with missing value. If the data missing problem can be solved by simply deleting a small part of samples, this method is the most effective. When the type of missing value is non-complete random missing, the deviation can be reduced by weighting the complete data. After marking the incomplete data cases, different weights are assigned to the complete data cases. The weight of the case can be obtained by logistic or probit regression.

[0146] Specifically, there are two forms of normalization method, one is to change the number to a decimal number between 0 and 1, and the other is to change the dimensional expression to a dimensionless expression. The normalization method is mainly proposed for the convenience of data processing. The data is mapped to the range of 0-1 for processing, which is more convenient and fast.

[0147] The specific processing normalization method is: .

[0148] In a possible implementation of the method, the method further includes:

[0149] The actual resource change data and the actual position change data of the current mineral feature group are determined within a preset time period in the future, and it is determined whether the actual resource change data and the actual position change data are consistent with the resource change data and the position change data, respectively. If not, the resource change data and the position change data are inversely normalized based on the actual resource change data and the actual position change data, and the resource change data and the position change data are restored to the actual resource change data and the actual position change data.

[0150] In a possible implementation of the method, if the resource change data and the position change data are not consistent, the resource change data and the position change data are inversely normalized based on the actual resource change data and the actual position change data, and the method further includes:

[0151] The root mean square error of the data is determined based on the actual resource change data, the actual position change data, the resource change data, and the position change data. The parameters in the epoch training model in the feature extraction model are set according to the root mean square error of the data, and the epoch training model after the setting is iteratively updated in reverse to obtain an updated feature extraction model and a data validation set of each round in the feature extraction model. The data validation set is calculated and evaluated to generate a loss value and an evaluation index of the data validation set.

[0152] Specifically, when a complete data set passes through a neural network once and returns once, the process is called an epoch training model. An epoch refers to the process of feeding all data into the network to complete a forward calculation and a backward propagation. Since an epoch is often too large, a computer cannot handle it, and we will divide it into several smaller batches. When training, it is not enough to iterate the training of all data once, and it needs to be repeated several times to fit and converge. In actual training, we divide all data into several batches, and each time we feed in a part of the data. Gradient descent is an iterative process itself. According to the root mean square error, the parameters in the epoch training model are set to obtain a validation set of each round in the feature extraction model, i.e., a validation set.

[0153] Specifically, the calculation and evaluation of the validation set include the following steps:

[0154] The validation set and the training set of the feature extraction model are calculated by percentage to obtain a loss value;

[0155] The loss value is compared with a standard loss value table to obtain an evaluation index.

[0156] For example, the current loss value is 50%, and the standard loss value table corresponding to 50% is a second-level indicator.

[0157] The following describes a mineral resource prediction device based on a three-dimensional geological model provided by an embodiment of the present application. The mineral resource prediction device based on a three-dimensional geological model described below can correspond to the mineral resource prediction method based on a three-dimensional geological model described above. Please refer to Figure 2 , Figure 2 FIG. 20 is a structural schematic diagram of a mineral resource prediction device 20 based on a three-dimensional geological model provided by an embodiment of the present application, which comprises:

[0158] The first acquisition module 21 is configured to acquire historical three-dimensional geological models and historical mineral resource data. The historical three-dimensional geological models are three-dimensional geological models of different geographical locations of mineral resources in a historical period of time. The historical mineral resource data is mineral resource data of different location nodes in the historical three-dimensional geological models in the historical period of time and data variation corresponding to the mineral resource data.

[0159] The data analysis module 22 is configured to analyze the mineral resource data to obtain mineral feature groups of different location nodes in the historical mineral resource data and time distribution sequence lengths of resource data corresponding to the mineral feature groups.

[0160] The data prediction module 23 is configured to predict the historical mineral resource data based on the mineral feature groups and the time distribution sequence lengths to obtain future resource data of different mineral feature groups in a future preset time and resource location nodes corresponding to the future resource data.

[0161] The resource determination module 24 is configured to determine resource variation data based on the future resource data and the mineral resource data.

[0162] The location determination module 25 is configured to determine location variation data according to the resource location nodes and different location nodes in the historical mineral resource data.

[0163] The second acquisition module 26 is configured to acquire a current three-dimensional geological model and current mineral resource data, and analyze the current mineral resource data to obtain current mineral feature groups of different location nodes in the current mineral resource data and current resource data corresponding to the current mineral feature groups. The current three-dimensional geological model is a three-dimensional geological model predicted and created according to a currently predicted mineral to be mined. The current mineral resource data is predicted resource data of different location nodes in the current three-dimensional geological model.

[0164] The feature matching module 27 is configured to match the current mineral feature groups with the mineral feature groups to obtain resource variation data and location variation data corresponding to the current mineral feature groups.

[0165] The data updating module 28 is configured to update different position nodes in the current mineral resource data and the current three-dimensional geological model according to the resource change data and the position change data respectively, to obtain updated current mineral resource data and resource position nodes corresponding to the current mineral resource data.

[0166] In one possible implementation of the embodiments of the present application, when the data prediction module 23 predicts the historical mineral resource data based on the mineral feature groups and the time distribution sequence length, to obtain future resource data of different mineral feature groups within a preset future time and resource position nodes corresponding to the future resource data, the data prediction module 23 is specifically configured to:

[0167] perform unsupervised time sequence data arrangement on the historical mineral resource data based on the time distribution sequence length and the mineral feature groups, to obtain resource matrix data;

[0168] input the resource matrix data into the trained feature extraction model to perform vector feature extraction, to obtain a feature dimension number, and perform data combination processing on the obtained feature dimension number and the resource matrix data, to generate future resource matrix data;

[0169] perform data processing on data contained in the future resource matrix data, to obtain resource matrix data, and input the obtained resource matrix data into a preset algorithm model to perform data calculation, to obtain future resource data of different mineral feature groups within a preset future time and resource position nodes corresponding to the future resource data.

[0170] In another possible implementation of the embodiments of the present application, when the data prediction module 23 inputs the resource matrix data into the trained feature extraction model to perform vector feature extraction, to obtain a feature dimension number, the data prediction module 23 is specifically configured to:

[0171] determine a mineral change position, mineral change data and mineral change time after each mineral resource data change in the historical mineral resource data based on the resource matrix data;

[0172] input the mineral change position, the mineral change data and the mineral change time into the feature extraction model respectively to perform vector feature extraction, to obtain a position feature vector corresponding to the mineral change position, a data feature vector corresponding to the mineral change data and a time feature vector corresponding to the mineral change time;

[0173] perform quantity statistics on the position feature vector, the data feature vector and the time feature vector, to obtain a feature dimension number.

[0174] In another possible implementation manner of the embodiment of the present application, the data prediction module 23 is specifically used for:

[0175] integrating the feature dimension number with the resource matrix data to generate first matrix data;

[0176] performing basic data distribution exploration on the first matrix data, obtaining a relative periodicity law of mineral resource change in historical mineral resource data, and determining a time period length based on the relative periodicity law;

[0177] performing supervised time series data arrangement on the first matrix data based on the time period length to obtain second matrix data;

[0178] predicting a mineral resource change trend in a future preset time period based on the second matrix data to generate future resource matrix data.

[0179] In another possible implementation manner of the embodiment of the present application, the data prediction module 23 is specifically used for:

[0180] calculating a normal distribution mean and a normal distribution variance of resource data contained in the future resource matrix data, and determining a 3σ range of the future resource matrix data based on the normal distribution mean and the normal distribution variance;

[0181] judging whether the resource data is outside the 3σ range, if the resource data is outside the 3σ range, determining a first matrix sequence of the future resource matrix data in which the resource data is located, calculating a sequence mean value according to the first matrix sequence, replacing the resource data with the sequence mean value to obtain a second matrix sequence after replacement, and performing missing value processing on the second matrix sequence;

[0182] performing sequence normalization processing on the second matrix sequence in the future resource matrix data to obtain the resource matrix data.

[0183] In another possible implementation manner of the embodiment of the present application, the device 20 further includes an actual data determination module and a data normalization module, wherein,

[0184] the actual data determination module is configured to determine actual resource change data and actual position change data detected by a current mineral feature group in a future preset time period;

[0185] The data normalization module is configured to determine whether the actual resource change data and the actual position change data are consistent with the resource change data and the position change data respectively, and if not, to perform inverse normalization processing on the resource change data and the position change data based on the actual resource change data and the actual position change data, so as to restore the resource change data and the position change data to the actual resource change data and the actual position change data.

[0186] In another possible implementation of the embodiment of the application, the apparatus 20 further includes an error determination module, a model updating module, and a data evaluation module, wherein,

[0187] The error determination module is configured to determine a root mean square error of data based on the actual resource change data, the actual position change data, the resource change data, and the position change data.

[0188] The model updating module is configured to set parameters in an epoch training model in the feature extraction model according to the root mean square error of data, and to perform reverse iterative updating on the set epoch training model to obtain an updated feature extraction model and a data validation set of each round in the feature extraction model.

[0189] The data evaluation module is configured to perform calculation and evaluation on the data validation set to generate a loss value and an evaluation index of the data validation set.

[0190] An electronic device provided in the embodiment of the application is described below. The electronic device described below can be referred to in correspondence with the method of predicting mineral resources based on a three-dimensional geological model described above.

[0191] The embodiment of the application provides an electronic device, as shown in Figure 3 Figure 3 FIG. 1 is a structural schematic diagram of an electronic device provided in the embodiment of the application, Figure 3 The electronic device 300 shown in FIG. 1 includes a processor 301 and a memory 303. The processor 301 and the memory 303 are connected, for example, through a bus 302. Optionally, the electronic device 300 can further include a transceiver 304. It should be noted that the transceiver 304 is not limited to one in actual application, and the structure of the electronic device 300 does not constitute a limitation on the embodiment of the application.

[0192] ​The processor 301 can be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic device, transistor logic device, hardware component, or any combination thereof. The processor 301 can implement or execute various exemplary logical blocks, modules and circuits described in combination with the disclosure of the embodiments of the present application. The processor 301 can also be a combination of computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, and the like.

[0193] The bus 302 can include a path for transmitting information between the above-mentioned components. The bus 302 can be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, or the like. The bus 302 can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, Figure 3 Only one thick line is used in the figure, but it does not mean that there is only one bus or only one type of bus.

[0194] The memory 303 can be a ROM (Read Only Memory) or other type of static storage device that can store static information and instructions, a RAM (Random Access Memory) or other type of dynamic storage device that can store information and instructions, an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer, but is not limited thereto.

[0195] The memory 303 is used to store application program codes for implementing the embodiments of the present application, and is controlled by the processor 301 to perform. The processor 301 is used to execute the application program codes stored in the memory 303 to realize the contents shown in the foregoing method embodiments.

[0196] The electronic device includes, but is not limited to, a mobile terminal such as a mobile phone, a notebook computer, a digital broadcast receiver, a PDA (Personal Digital Assistant), a PAD (Tablet Personal Computer), a PMP (Portable Multimedia Player), a car terminal (for example, a car navigation terminal), and the like, and a stationary terminal such as a digital TV, a desktop computer, and the like. Figure 3 The illustrated electronic device is merely an example and should not impose any limitation on the function and use range of the embodiments of the present application.

[0197] A computer readable storage medium according to an embodiment of the present application is described below, and the computer readable storage medium described below can be referred to in correspondence with the method described above.

[0198] The embodiment of the present application provides a computer readable storage medium, and the computer readable storage medium stores a computer program. The computer program is executed by a processor to realize the steps of the mineral resource prediction method based on a three-dimensional geological model.

[0199] Since the embodiments of the computer readable storage medium part correspond to the embodiments of the method part, the embodiments of the computer readable storage medium part are described with reference to the description of the embodiments of the method part.

[0200] It should be understood that, although each step in the flowchart of the accompanying drawings is displayed in sequence according to the indication of the arrow, these steps are not necessarily executed in sequence according to the indication of the arrow. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and they can be executed in other sequences. Moreover, at least part of the steps in the flowchart of the accompanying drawings can include multiple sub-steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence is not necessarily sequential, but can be executed in rotation or alternation with at least part of other steps or sub-steps or stages of other steps.

[0201] The above is only some embodiments of the present application, and it should be pointed out that, for those skilled in the art, without departing from the principles of the present application, some improvements and refinements can be made, and these improvements and refinements should be regarded as the protection scope of the present application.

Claims

1. A mineral resource prediction method based on a three-dimensional geological model, characterized in that, include: Acquire historical three-dimensional geological models and historical mineral resource data. The historical three-dimensional geological model is a three-dimensional geological model created for minerals at different geographical locations within a historical period. The historical mineral resource data consists of mineral resource data located at different nodes in the historical three-dimensional geological model within the historical period and the corresponding data changes. The historical mineral resource data is analyzed to obtain the mineral feature groups at different locations in the historical mineral resource data and the time distribution sequence length of the resource data corresponding to the mineral feature groups; Based on the mineral feature groups and the length of the time distribution sequence, the historical mineral resource data is predicted to obtain the future resource data of different mineral feature groups within a preset time period and the resource location nodes corresponding to the future resource data. Resource change data is determined based on the aforementioned future resource data and mineral resource data; The location change data is determined based on the resource location nodes and different location nodes in the historical mineral resource data; The current three-dimensional geological model and current mineral resource data are acquired, and the current mineral resource data is analyzed to obtain the current mineral feature groups of different location nodes in the current mineral resource data and the current resource data corresponding to the current mineral feature groups. The current three-dimensional geological model is a three-dimensional geological model created based on the current pre-mining minerals, and the current mineral resource data is the predicted resource data located at different location nodes in the current three-dimensional geological model. The current mineral feature group is matched with the mineral feature group to obtain the resource change data and the location change data corresponding to the current mineral feature group; The current mineral resource data and different location nodes in the current three-dimensional geological model are updated based on the resource change data and the location change data, respectively, to obtain the updated current mineral resource data and the resource location nodes corresponding to the current mineral resource data.

2. The mineral resource prediction method based on a three-dimensional geological model according to claim 1, characterized in that, The method of predicting historical mineral resource data based on the mineral feature groups and the length of the time distribution sequence to obtain future resource data for different mineral feature groups within a preset time period and the resource location nodes corresponding to the future resource data includes: Based on the time distribution sequence length and the mineral feature group, the historical mineral resource data is processed in an unsupervised time series data manner to obtain resource matrix data; The resource matrix data is input into a trained feature extraction model to extract vector features, thereby obtaining the number of feature dimensions. The obtained number of feature dimensions is then combined with the resource matrix data to generate future resource matrix data. The data contained in the future resource matrix data is processed to obtain resource matrix data, and the obtained resource matrix data is input into a preset algorithm model for data extrapolation to obtain future resource data of different mineral characteristic groups within a preset time period and resource location nodes corresponding to the future resource data.

3. The mineral resource prediction method based on a three-dimensional geological model according to claim 2, characterized in that, The resource matrix data is input into a trained feature extraction model for vector feature extraction to obtain the number of feature dimensions, including: Based on the resource matrix data, determine the location, data, and time of each mineral resource change in the historical mineral resource data. The location of the mineral change, the data of the mineral change, and the time of the mineral change are respectively input into the feature extraction model for vector feature extraction, to obtain the location vector feature corresponding to the location of the mineral change, the data vector feature corresponding to the data of the mineral change, and the time vector feature corresponding to the time of the mineral change; The number of feature dimensions is obtained by statistically analyzing the location vector features, data vector features, and time vector features.

4. The mineral resource prediction method based on a three-dimensional geological model according to claim 2, characterized in that, The obtained number of feature dimensions is combined with the resource matrix data to generate future resource matrix data, including: The number of feature dimensions is integrated with the resource matrix data to generate the first matrix data; The basic data distribution of the first matrix data is explored to obtain the relative periodicity of mineral resource changes in the historical mineral resource data, and the time period length is determined based on the relative periodicity. Based on the time period length, the first matrix data is processed into supervised time series data to obtain the second matrix data; Based on the data in the second matrix, the trend of mineral resource changes within a preset time period is predicted, and future resource matrix data is generated.

5. A mineral resource prediction method based on a three-dimensional geological model according to claim 2, characterized in that, The process of processing the data contained in the future resource matrix data to obtain the resource matrix data includes: Calculate the mean and variance of the normal distribution of the resource data contained in the future resource matrix data, and determine the 3σ range of the future resource matrix data based on the mean and variance of the normal distribution; Determine whether the resource data is outside the 3σ range. If the resource data is outside the 3σ range, determine the first matrix sequence of the future resource matrix data in which the resource data is located. Calculate the sequence average value based on the first matrix sequence and replace the resource data with the sequence average value to obtain the replaced second matrix sequence. Then, process the missing values ​​in the second matrix sequence. The second matrix sequence in the future resource matrix data is normalized to obtain the resource matrix data.

6. The mineral resource prediction method based on a three-dimensional geological model according to claim 2, characterized in that, The method further includes: Determine the actual resource change data and actual location change data of the current mineral feature group within a future preset time period; It is determined whether the actual resource change data and the actual location change data are consistent with the resource change data and location change data. If they are inconsistent, the resource change data and the location change data are reverse normalized based on the actual resource change data and the actual location change data to restore the resource change data and the location change data to the actual resource change data and the actual location change data.

7. The mineral resource prediction method based on a three-dimensional geological model according to claim 6, characterized in that, If there is a discrepancy, then the resource change data and the location change data are subjected to inverse normalization processing based on the actual resource change data and the actual location change data, and then the process further includes: The root mean square error of the data is determined based on the actual resource change data, the actual location change data, the resource change data, and the location change data. The parameters in the epoch training model of the feature extraction model are set according to the root mean square error of the data, and the epoch training model is updated in reverse iteration to obtain the updated feature extraction model and the data validation set of each round in the feature extraction model. The data validation set is calculated and evaluated to generate the loss value and evaluation index for the data validation set.

8. A mineral resource prediction device based on a three-dimensional geological model, characterized in that, include: The first geological acquisition module is used to acquire historical three-dimensional geological models and historical mineral resource data. The historical three-dimensional model is a three-dimensional geological model created for minerals at different geographical locations within a historical period. The historical mineral resource data consists of mineral resource data located at different nodes in the historical three-dimensional geological model within the historical period and the corresponding data changes. The data analysis module is used to analyze the historical mineral resource data to obtain the mineral feature groups of different nodes in the historical mineral resource data and the time distribution sequence length of the resource data corresponding to the mineral feature groups. The data prediction module is used to predict the historical mineral resource data based on the mineral feature group and the time distribution sequence length, so as to obtain the future resource data of different mineral feature groups within a preset time period and the resource location nodes corresponding to the future resource data. The resource determination module is used to determine resource change data based on the future resource data and mineral resource data. The location determination module is used to determine location change data based on the resource location nodes and different location nodes in the historical mineral resource data. The second acquisition module is used to acquire the current three-dimensional geological model and the current mineral resource data, and to analyze the current mineral resource data to obtain the current mineral feature groups of different location nodes in the current mineral resource data and the current resource data corresponding to the current mineral feature groups. The current three-dimensional geological model is a three-dimensional geological model created based on the current pre-mining minerals, and the current mineral resource data is the predicted resource data located at different location nodes in the current three-dimensional geological model. The feature matching module is used to match the current mineral feature group with the mineral feature group to obtain the resource change data and the location change data corresponding to the current mineral feature group; The data update module updates the current mineral resource data and different location nodes in the current three-dimensional geological model according to the resource change data and the location change data, respectively, to obtain the updated current mineral resource data and the resource location nodes corresponding to the current mineral resource data.

9. An electronic device, characterized in that, include: At least one processor; Memory; At least one application, wherein the at least one application is stored in memory and configured to be executed by at least one processor, said at least one application being configured to: perform a mineral resource prediction method based on a three-dimensional geological model as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer program is stored and can be loaded by a processor and executed as any one of the mineral resource prediction methods based on a three-dimensional geological model as described in claims 1 to 7.

Images