Mining quantity prediction method and system based on machine learning

Abstract

The invention discloses a mining quantity prediction method and system based on machine learning, and relates to the technical field of data processing based on the prediction purpose, and the method comprises the steps: obtaining a plurality of data sources corresponding to the mining quantity, and carrying out the preprocessing of the data sources through a distributed calculation framework, and the processing process comprises the steps: removing noise, carrying out the standardization, and filling a missing value; and constructing and training a machine learning model based on the preprocessed data, learning rules from the data through the machine learning model and outputting a mining quantity prediction result, predicting the future mining quantity through the trained machine learning model, and updating the prediction result in real time. The method not only improves the precision and efficiency of mining quantity prediction, but also can adapt to a complex dynamic environment, and has high practicability and application prospects.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing for prediction purposes, and in particular to a mining volume prediction method and system based on machine learning. Background Art

[0002] With the continuous advancement of mineral resource mining technology, the prediction of mining volume has become an important part of mine management and resource optimization. Traditional mining volume prediction methods mainly rely on historical data and empirical formulas. These methods usually ignore the complex spatiotemporal dependencies in the data and cannot dynamically adapt to the real-time changing mining environment. Although there are some mining volume prediction methods based on statistical models and machine learning in the prior art, these methods mostly rely on static model assumptions and have limited mining volume prediction effects in large-scale and complex environments. Existing methods face great challenges in processing large-scale data, spatiotemporal changes, and real-time updates, and it is difficult to meet the needs of high-precision and real-time predictions in the mining production process. Therefore, developing a mining volume prediction method that can effectively integrate spatial and temporal characteristics and adapt to dynamic changes has become a technical problem that needs to be solved urgently in the current mining field. Summary of the invention

[0003] In view of the above existing problems, the present invention is proposed.

[0004] Therefore, the present invention provides a mining volume prediction method based on machine learning, which solves several key technical problems in the existing mining volume prediction technology. First, traditional methods usually deal with spatial and temporal characteristics separately, and fail to effectively combine spatiotemporal characteristics, so as to be unable to fully capture the complex laws of mining volume changes in mining areas. Secondly, most of the existing technologies rely on static models, lack the ability to update in real time and adapt to dynamic changes in data, resulting in the inability to accurately reflect real-time fluctuations in the mining process of mining areas. In addition, traditional methods are inefficient in processing large-scale data and it is difficult to provide fast and accurate predictions. Finally, when facing changing data, existing machine learning models are often prone to overfitting, resulting in poor generalization ability and inability to adapt to new environments or emergencies. The present invention fundamentally solves these technical problems by combining spatiotemporal convolutional networks (ST-CNN) with long short-term memory networks (LSTM), and performing dynamic learning and real-time prediction based on a distributed computing framework.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions, a mining volume prediction method based on machine learning, comprising:

[0006] Obtain multiple data sources corresponding to the mining volume, including historical mining data, geological data, equipment operation data, environmental data and other factors that may affect the mining volume;

[0007] Preprocessing the data source through a distributed computing framework, including removing noise, standardizing, and filling missing values;

[0008] Based on the preprocessed data, a machine learning model is constructed and trained, the rules are learned through the machine learning model and the mining volume prediction results are output, the future mining volume is predicted through the trained machine learning model, and the prediction results are updated in real time.

[0009] As a preferred solution of the mining volume prediction method based on machine learning described in the present invention, the data source includes sensor data, equipment status data, environmental parameter data and historical mining volume data of the mining area; the sensor data of the mining area includes temperature, humidity, air pressure and vibration data in the mine; the equipment status data includes the operating status, fault records and maintenance records of the equipment in the mine; the environmental parameter data includes temperature, precipitation, wind speed and seismic activity data outside the mining area; the historical mining volume data includes the mining volume of the mining area over the years or months, the usage of mining equipment and mining speed data.

[0010] As a preferred solution of the mining volume prediction method based on machine learning described in the present invention, the distributed computing framework adopts Apache Spark to perform data parallel processing, and the distributed computing framework performs distributed storage and parallel computing of data through Apache Spark.

[0011] As a preferred solution of the mining volume prediction method based on machine learning described in the present invention, wherein: the preprocessing includes denoising, standardizing and filling missing values ​​of the data source;

[0012] The denoising step processes the noise in the data source by using a filtering algorithm; the standardization independently standardizes each feature and converts feature data of different dimensions and units into a unified range. The specific method is to normalize according to the minimum and maximum values ​​and compress the data values ​​into the range of [0, 1]; the missing value filling step fills the missing values ​​by using interpolation methods, wherein the interpolation methods include linear interpolation and spline interpolation. Linear interpolation generates an estimated value by connecting the data points before and after the missing value, while spline interpolation uses a polynomial function for smooth fitting.

[0013] As a preferred solution of the mining volume prediction method based on machine learning described in the present invention, the construction and training of the machine learning model includes incorporating the time dimension into the convolution calculation and capturing the local characteristics of space and time, as follows: ;

[0014] in, Indicates that at time step and spatial location The data characteristics on is the spatial and temporal convolution kernel, Represents the convolution operation; is the length of the time dimension, is the number of spatial dimensions; b is the bias term, which indicates the offset of the convolution result; It is the output feature after the spatiotemporal convolution operation;

[0015] In order to further enhance the model's capabilities, the prediction results of the previous time step are introduced As the input of LSTM, historical prediction information is used to adjust the modeling at the current moment, as follows: ;

[0016] in, is the hidden state at the current time step; is the hidden state at the previous time step; is the input feature of the current time step; is the prediction result of the previous time step; The cell state at the current time step; is the output gate of the current time step, represents element-wise product, is the activation function;

[0017] Extracted from ST-CNN Extracted with LSTM Fusion, introducing adaptive weighting mechanism and spatiotemporal correlation matrix , which enables spatial and temporal features to be dynamically weighted according to their importance, and captures the interaction between spatial and temporal features through the Hadamard product: ;

[0018] in, and is a matrix used to weight spatial and temporal features; is the spatiotemporal correlation matrix, which is used to capture the dependencies between spatial and temporal features; represents the Hadamard product;

[0019] The result of spatiotemporal feature fusion is input into the output layer to finally generate the predicted value of mining volume. In order to prevent overfitting and improve the generalization ability of the model, an L2 regularization term is added to the output layer: ;

[0020] in, is the predicted mining volume; is the fused spatiotemporal feature, is the weight matrix of the output layer, is the bias term, is the L2 regularization term, is the L2 norm of the weight matrix, is the regularization coefficient.

[0021] As a preferred embodiment of the mining volume prediction method based on machine learning described in the present invention, the prediction results include combining the prediction results with the geographic information system GIS to generate a three-dimensional visualization image including the mining volume prediction and risk area, and pushing the three-dimensional visualization image to the decision support system.

[0022] As a preferred solution of the mining volume prediction method based on machine learning described in the present invention, wherein: the decision support system includes a visualization interface for displaying the mining volume prediction results of different regions and time periods, wherein the visualization interface of the decision support system is used to display the changing trends of the mining volume prediction results in different regions and time periods, and users can select regions and adjust time according to needs, and adjust the input parameters of the machine learning model.

[0023] As a preferred solution of the mining volume prediction system based on machine learning described in the present invention, it includes: a data acquisition module, an algorithm module and a decision support system; the data acquisition module is used to collect data in real time; the algorithm module is used to carry the algorithms required for calculation and the geographic information system GIS; the decision support system is used to output and feedback the prediction results.

[0024] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of a mining volume prediction method based on machine learning when executing the computer program.

[0025] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a mining volume prediction method based on machine learning.

[0026] Beneficial effects of the present invention: The present invention provides a mining volume prediction method based on spatiotemporal convolutional network and long short-term memory network, which has significant beneficial effects. First, by introducing spatiotemporal convolutional network (ST-CNN) and long short-term memory network (LSTM), the present invention can extract complex features in mining area data from two dimensions of space and time, comprehensively capture the spatiotemporal laws of mining volume changes, and improve the accuracy of prediction. Secondly, the present invention adopts an online learning mechanism, which can adapt to the dynamic changes of data in real time, and ensure timely adjustment and prediction of mining volume in the production environment. Furthermore, combined with a distributed computing framework, the present invention can efficiently process large-scale data sets and update model parameters in real time, thereby effectively solving the problem of low efficiency in large-scale data calculation. Finally, the model introduces regularization technology and spatiotemporal feature fusion mechanism, which enhances generalization ability, avoids overfitting, and ensures that high prediction accuracy can be maintained in different mining area environments. In summary, the present invention not only improves the accuracy and efficiency of mining volume prediction, but also can adapt to complex dynamic environments, and has strong practicality and application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0028] Figure 1 A schematic flow chart of a method for predicting production based on machine learning according to an embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0030] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0031] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0032] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0033] At the same time, in the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper, lower, inner and outer" are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0034] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0035] Example 1, reference Figure 1 , which is the first embodiment of the present invention, provides a mining volume prediction method based on machine learning, comprising:

[0036] S1: Obtain multiple data sources corresponding to the mining volume, including historical mining data, geological data, equipment operation data, environmental data, and other factors that may affect the mining volume.

[0037] S2: Preprocessing the data source through a distributed computing framework, including removing noise, standardizing, and filling missing values.

[0038] S3: Based on the preprocessed data, a machine learning model is constructed and trained, and the rules are learned from the data through the machine learning model and the mining volume prediction results are output, and the future mining volume is predicted through the trained machine learning model, and the prediction results are updated in real time.

[0039] The data sources include sensor data, equipment status data, environmental parameter data and historical mining volume data of the mining area; among them, the sensor data of the mining area includes temperature, humidity, air pressure and vibration data in the mine; the equipment status data includes the operating status, fault records and maintenance records of the equipment in the mine; the environmental parameter data includes temperature, precipitation, wind speed and seismic activity data outside the mining area; the historical mining volume data includes the mining volume of the mining area over the years or months, the usage of mining equipment and mining speed data.

[0040] The distributed computing framework uses Apache Spark to perform data parallel processing, and the distributed computing framework performs distributed storage and parallel computing of data through Apache Spark.

[0041] The preprocessing includes denoising, standardizing and filling missing values ​​on the data source;

[0042] The denoising step processes the noise in the data source by using a filtering algorithm; the standardization independently standardizes each feature and converts feature data of different dimensions and units into a unified range. The specific method is to normalize according to the minimum and maximum values ​​and compress the data values ​​into the range of [0, 1]; the missing value filling step fills the missing values ​​by using interpolation methods, wherein the interpolation methods include linear interpolation and spline interpolation. Linear interpolation generates an estimated value by connecting the data points before and after the missing value, while spline interpolation uses a polynomial function for smooth fitting.

[0043] The construction and training of the machine learning model includes incorporating the time dimension into the convolution calculation and capturing the local features of space and time, as follows: ;

[0044] in, Indicates that at time step and spatial location The data characteristics on is the spatial and temporal convolution kernel, Represents the convolution operation; is the length of the time dimension, is the number of spatial dimensions; b is the bias term, which indicates the offset of the convolution result; It is the output feature after the spatiotemporal convolution operation;

[0045] In order to further enhance the model's capabilities, the prediction results of the previous time step are introduced As the input of LSTM, historical prediction information is used to adjust the modeling at the current moment, as follows: ;

[0046] in, is the hidden state at the current time step; is the hidden state at the previous time step; is the input feature of the current time step; is the prediction result of the previous time step; is the cell state at the current time step; is the output gate of the current time step, represents element-wise product, is the activation function;

[0047] Extracted from ST-CNN Extracted with LSTM Fusion, introducing adaptive weighting mechanism and spatiotemporal correlation matrix , which enables spatial and temporal features to be dynamically weighted according to their importance, and captures the interaction between spatial and temporal features through the Hadamard product: ;

[0048] in, and is a matrix used to weight spatial and temporal features; is the spatiotemporal correlation matrix, which is used to capture the dependencies between spatial and temporal features; represents the Hadamard product;

[0049] The result of spatiotemporal feature fusion is input into the output layer to finally generate the predicted value of mining volume. In order to prevent overfitting and improve the generalization ability of the model, an L2 regularization term is added to the output layer: ;

[0050] in, is the predicted mining volume; is the fused spatiotemporal feature, is the weight matrix of the output layer, is the bias term, is the L2 regularization term, is the L2 norm of the weight matrix, is the regularization coefficient.

[0051] The prediction results include combining the prediction results with a geographic information system (GIS) to generate a three-dimensional visualization image including mining volume prediction and risk areas, and pushing the three-dimensional visualization image to a decision support system.

[0052] The decision support system includes a visualization interface for displaying mining volume forecast results for different regions and time periods, wherein the visualization interface of the decision support system is used to display the changing trends of mining volume forecast results for different regions and time periods, and users can select regions and adjust time according to needs, and adjust the input parameters of the machine learning model.

[0053] Embodiment 2 is the second embodiment of the present invention, which provides a mining volume prediction system based on machine learning, including: a data acquisition module, an algorithm module and a decision support system; the data acquisition module is used to collect data in real time; the algorithm module is used to carry the algorithms required for calculation and the geographic information system GIS; the decision support system is used to output and feedback the prediction results.

[0054] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0055] Example 3

[0056] The third embodiment of the present invention is different from the first two embodiments in that:

[0057] 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 invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0058] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0059] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0060] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0061] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0062] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A mining volume prediction method based on machine learning, characterized in that: include, Obtain multiple data sources corresponding to the amount of mining; Preprocessing the data source through a distributed computing framework, including removing noise, standardizing, and filling missing values; Based on the preprocessed data, a machine learning model is constructed and trained, rules are learned from the machine learning model and a mining volume prediction result is output, future mining volumes are predicted through the trained machine learning model, and the prediction result is updated in real time; The construction and training of the machine learning model includes incorporating the time dimension into the convolution calculation and capturing the local features of space and time, as follows: ； in, Indicates that at time step and spatial location The data characteristics on is the spatial and temporal convolution kernel, Represents the convolution operation; is the length of the time dimension, is the number of spatial dimensions; b is the bias term, which indicates the offset of the convolution result; It is the output feature after the spatiotemporal convolution operation; In order to further enhance the model's capabilities, the prediction results of the previous time step are introduced As the input of LSTM, historical prediction information is used to adjust the modeling at the current moment, as follows: ； in, is the hidden state at the current time step; is the hidden state at the previous time step; is the input feature of the current time step; is the prediction result of the previous time step; is the cell state at the current time step; is the output gate of the current time step, represents element-wise product, is the activation function; Extracted from ST-CNN Extracted with LSTM Fusion, introducing adaptive weighting mechanism and spatiotemporal correlation matrix , which enables spatial and temporal features to be dynamically weighted according to their importance, and captures the interaction between spatial and temporal features through the Hadamard product: ； in, and is a matrix used to weight spatial and temporal features; is the spatiotemporal correlation matrix, which is used to capture the dependencies between spatial and temporal features; represents the Hadamard product; The result of spatiotemporal feature fusion is input into the output layer to finally generate the predicted value of mining volume. In order to prevent overfitting and improve the generalization ability of the model, an L2 regularization term is added to the output layer: ； in, is the predicted mining volume; is the fused spatiotemporal feature, is the weight matrix of the output layer, is the bias term, is the L2 regularization term, is the L2 norm of the weight matrix, is the regularization coefficient.

2. The mining volume prediction method based on machine learning according to claim 1, characterized in that: The data sources include sensor data, equipment status data, environmental parameter data and historical mining volume data of the mining area; among them, the sensor data of the mining area includes temperature, humidity, air pressure and vibration data in the mine; the equipment status data includes the operating status, fault records and maintenance records of the equipment in the mine; the environmental parameter data includes temperature, precipitation, wind speed and seismic activity data outside the mining area; the historical mining volume data includes the mining volume of the mining area over the years or months, the usage of mining equipment and mining speed data.

3. The mining volume prediction method based on machine learning as claimed in claim 2, characterized in that: The distributed computing framework uses Apache Spark to perform data parallel processing, and the distributed computing framework performs distributed storage and parallel computing of data through Apache Spark.

4. The mining volume prediction method based on machine learning as claimed in claim 3, characterized in that: The preprocessing includes denoising, standardizing and filling missing values ​​on the data source; The denoising step processes the noise in the data source by using a filtering algorithm; the standardization independently standardizes each feature and converts feature data of different dimensions and units into a unified range. The specific method is to normalize according to the minimum and maximum values ​​and compress the data values ​​into the range of [0, 1]; the missing value filling step fills the missing values ​​by using interpolation methods, wherein the interpolation methods include linear interpolation and spline interpolation. Linear interpolation generates an estimated value by connecting the data points before and after the missing value, while spline interpolation uses a polynomial function for smooth fitting.

5. The mining volume prediction method based on machine learning according to claim 4, characterized in that: The prediction results include combining the prediction results with a geographic information system (GIS) to generate a three-dimensional visualization image including mining volume prediction and risk areas, and pushing the three-dimensional visualization image to a decision support system.

6. The mining volume prediction method based on machine learning according to claim 5, characterized in that: The decision support system includes a visualization interface for displaying the mining volume forecast results for different regions and time periods, wherein the visualization interface of the decision support system is used to display the changing trends of the mining volume forecast results for different regions and time periods, and the user selects the region and adjusts the time according to needs, and adjusts the input parameters of the machine learning model.

7. A mining yield prediction system based on machine learning, applied to the mining yield prediction method based on machine learning as claimed in any one of claims 1 to 6, characterized in that: Including data acquisition module, algorithm module and decision support system; The data acquisition module is used to collect data in real time; The algorithm module is used to carry the algorithms and geographic information system GIS required for calculation; The decision support system is used to output and provide feedback on the prediction results.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the production volume prediction method based on machine learning described in any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the mining volume prediction method based on machine learning described in any one of claims 1 to 6 are implemented.

Images