Mineral resource intelligent prediction system and method based on man-machine fusion

Abstract

The invention relates to the field of intelligent prediction, and particularly discloses a mineral resource intelligent prediction system based on man-machine fusion and a method thereof. By establishing a mapping relation between a plurality of nodes and a plurality of attribute data in the petroleum safety space-time big data, the grade label of the petroleum supply interruption risk is accurately evaluated, so that corresponding response measures such as reserve release, yield increase and traffic control are taken based on the grade label of the risk, and the safety of petroleum is ensured.

Description

Technical Field

[0001] The present application relates to the field of intelligent prediction, and more specifically, to a mineral resource intelligent prediction system and method based on human-machine integration. Background Art

[0002] Oil is the lifeblood of industry and a vital energy source for national development. Its stable supply is crucial for economic development. International developments, geopolitics, energy landscapes, economic policies, disasters, and weather can all impact oil production and supply. To address the impact of one or more major emergencies on national, regional, and global energy supplies (especially oil), scientists have been focusing on the impact and risks posed by various destabilizing factors.

[0003] Currently, although international oil production, consumption, imports, exports, and prices fluctuate over a long period of time, the overall supply and demand of oil remains balanced, achieving dynamic oil security. However, the occurrence of one or more events could disrupt this balance, leading to oil security issues. This impacts the entire oil supply chain, from upstream to downstream. In other words, each link in the oil supply chain, including production, transportation, processing, and consumption, could be affected by multiple factors due to a single or multiple events, disrupting the entire supply chain and potentially disrupting regional, national, and even global oil security balances. However, traditional oil security strategic analysis methods suffer from limitations such as delayed data updates, difficulty processing complex data, and low efficiency, making them difficult to adapt to the rapid analysis requirements of the big data era.

[0004] Therefore, an optimized intelligent prediction system for mineral resources is desired. Summary of the Invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiments of the present application provide a mineral resource intelligent prediction system and method based on human-machine integration, which uses a neural network model based on deep learning to establish a mapping relationship between multiple nodes and multiple attribute data in the spatiotemporal big data of oil security to accurately assess the level label of the oil supply interruption risk, thereby taking corresponding response measures such as releasing reserves, increasing production, and traffic control based on the risk level label to ensure oil safety.

[0006] According to one aspect of the present application, a method for intelligent prediction of mineral resources based on human-machine integration is provided, which includes: Acquiring multiple nodes related to oil security and attribute data of the multiple nodes, wherein the multiple nodes include oil-producing countries, importing countries, pipelines, export ports, import ports, straits, military bases, refineries, and gas stations, and the attribute data includes reserves, production, consumption, export volume, import volume, price, GDP, population, urbanization rate, passenger volume, freight volume, and vehicle ownership; Arranging the attribute data of each node into an input vector and then passing it through a one-dimensional convolutional neural network model to obtain a multi-dimensional data association feature vector of multiple nodes; Constructing a spatial topological matrix of the plurality of nodes, wherein the eigenvalue of each diagonal position in the spatial topological matrix is the distance between the corresponding two nodes; Passing the spatial topology matrix through a convolutional neural network model as a feature extractor to obtain an inter-node spatial topology feature matrix; Arranging the plurality of node multidimensional data association feature vectors into a global node multidimensional data association feature matrix; The global node multidimensional data association feature matrix and the inter-node spatial topology feature matrix are passed through a graph neural network model to obtain a spatial topology global node multidimensional data association feature matrix; The spatial topology global node multidimensional data association feature matrix is passed through a classifier to obtain a classification result, and the classification result is used to represent an oil supply interruption risk level label.

[0007] In the above-mentioned intelligent prediction method of mineral resources based on human-machine integration, the attribute data of each node is arranged as an input vector and then passed through a one-dimensional convolutional neural network model to obtain multiple node multidimensional data association feature vectors, including: using each layer of the one-dimensional convolutional neural network model to perform the following on the input data in the forward pass of the layer: convolution processing on the input data to obtain a convolution feature map; pooling based on a feature matrix on the convolution feature map to obtain a pooled feature map; and nonlinear activation on the pooled feature map to obtain an activation feature map; wherein the output of the last layer of the one-dimensional convolutional neural network is the node multidimensional data association feature vector, and the input of the first layer of the one-dimensional convolutional neural network is the attribute data input vector.

[0008] In the above-mentioned intelligent prediction method of mineral resources based on human-machine fusion, the spatial topology matrix is passed through a convolutional neural network model as a feature extractor to obtain an inter-node spatial topology feature matrix, including: using each layer of the convolutional neural network model as a feature extractor to perform the following on the input data in the forward pass of the layer: convolution processing on the input data to obtain a convolution feature map; pooling the convolution feature map along the channel dimension to obtain a pooled feature map; and nonlinear activation on the pooled feature map to obtain an activation feature map; wherein the output of the last layer of the convolutional neural network as a feature extractor is the inter-node spatial topology feature matrix, and the input of the first layer of the convolutional neural network as a feature extractor is the spatial topology matrix.

[0009] In the above-mentioned intelligent prediction method for mineral resources based on human-machine integration, the multiple node multidimensional data association feature vectors are arranged into a global node multidimensional data association feature matrix, including: obtaining an auxiliary feature map by passing the global node multidimensional data association feature matrix through a convolutional neural network as a feature extractor; modeling the auxiliary feature map based on a feature significance correlation reconstruction mechanism to obtain a weighted feature vector; performing point multiplication of the weighted feature vector with each node multidimensional data association feature vector to obtain the multiple corrected node multidimensional data association feature vectors; and arranging the multiple corrected node multidimensional data association feature vectors into a global node multidimensional data association feature matrix.

[0010] In the above-mentioned intelligent prediction method of mineral resources based on human-machine fusion, the global node multidimensional data association feature matrix is passed through a convolutional neural network as a feature extractor to obtain an auxiliary feature map, including: using each layer of the convolutional neural network as a feature extractor to perform the following on the input data in the forward pass of the layer: convolution processing on the input data to obtain a convolution feature map; pooling the convolution feature map based on the local feature matrix to obtain a pooled feature map; and nonlinear activation on the pooled feature map to obtain an activation feature map; wherein the output of the last layer of the convolutional neural network as a feature extractor is the auxiliary feature map, and the input of the first layer of the convolutional neural network as a feature extractor is the global node multidimensional data association feature matrix.

[0011] In the above-mentioned intelligent prediction method of mineral resources based on human-machine fusion, the auxiliary feature map is modeled based on the feature significance correlation reconstruction mechanism to obtain a weighted feature vector, including: calculating the hyperbolic sine function value and the hyperbolic cosine function value of each eigenvalue of the auxiliary feature map to obtain a first feature map and a second feature map; the first feature map and the second feature map are respectively and Activation is performed to obtain a first saliency-related feature map and a second saliency-related feature map; after dot-adding the first saliency-related feature map and the second saliency-related feature map, a global pooling operation is performed on each feature matrix to obtain the weighted feature vector.

[0012] In the above-mentioned intelligent prediction method of mineral resources based on human-machine fusion, the global node multidimensional data association feature matrix and the corrected inter-node spatial topology feature matrix are passed through a graph neural network model to obtain a spatial topology global node multidimensional data association feature matrix, including: using the graph neural network model with learnable neural network parameters to perform graph structure encoding on the global node multidimensional data association feature matrix and the inter-node spatial topology feature matrix to obtain the spatial topology global node multidimensional data association feature matrix containing irregular node spatial topology features and multiple attribute association features of each node.

[0013] In the above-mentioned intelligent prediction method of mineral resources based on human-machine integration, the spatial topology global node multidimensional data association feature matrix is passed through a classifier to obtain a classification result, and the classification result is used to represent the oil supply interruption risk level label, including: expanding the spatial topology global node multidimensional data association feature matrix into a classification feature vector based on a row vector or a column vector; using multiple fully connected layers of the classifier to fully connect encode the classification feature vector to obtain an encoded classification feature vector; and passing the encoded classification feature vector through the Softmax classification function of the classifier to obtain the classification result.

[0014] According to another aspect of the present application, a mineral resource intelligent prediction system based on human-machine integration is provided, which includes: an attribute data acquisition module, configured to acquire multiple nodes related to oil security and attribute data of the multiple nodes, wherein the multiple nodes include oil-producing countries, importing countries, pipelines, export ports, import ports, straits, military bases, refineries, and gas stations, and the attribute data includes reserves, production, consumption, exports, imports, prices, GDP, population, urbanization rate, passenger volume, freight volume, and vehicle ownership; A one-dimensional convolution module, configured to arrange the attribute data of each node into an input vector and then pass it through a one-dimensional convolutional neural network model to obtain a multi-dimensional data association feature vector for multiple nodes; A spatial topology construction module, configured to construct a spatial topology matrix of the plurality of nodes, wherein the eigenvalue of each diagonal position in the spatial topology matrix is the distance between the corresponding two nodes; A spatial topology feature extraction module, configured to pass the spatial topology matrix through a convolutional neural network model as a feature extractor to obtain an inter-node spatial topology feature matrix; an arrangement module, configured to arrange the plurality of node multidimensional data association feature vectors into a global node multidimensional data association feature matrix; A graph feature extraction module is used to obtain a spatial topology global node multidimensional data association feature matrix by passing the global node multidimensional data association feature matrix and the inter-node spatial topology feature matrix through a graph neural network model; The classification result generating module is used to pass the spatial topology global node multidimensional data association feature matrix through a classifier to obtain a classification result, and the classification result is used to represent the oil supply interruption risk level label.

[0015] According to another aspect of the present application, an electronic device is provided, comprising: a processor; and a memory, wherein computer program instructions are stored in the memory, and when the computer program instructions are executed by the processor, the processor executes the intelligent prediction method for mineral resources based on human-machine integration as described above.

[0016] According to another aspect of the present application, a computer-readable medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the processor executes the above-mentioned intelligent prediction method for mineral resources based on human-computer integration.

[0017] Compared with the existing technology, the present application provides an intelligent mineral resource prediction system and method based on human-machine integration. It adopts a neural network model based on deep learning and establishes a mapping relationship between multiple nodes and multiple attribute data in the spatiotemporal big data of oil safety to accurately evaluate the level label of the oil supply interruption risk, so as to take corresponding response measures such as releasing reserves, increasing production, and traffic control based on the risk level label to ensure the safety of oil. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0019] Figure 1 Flowchart of a method for intelligent prediction of mineral resources based on human-machine fusion according to an embodiment of the present application; Figure 2 Schematic diagram of the architecture of a method for intelligent prediction of mineral resources based on human-machine integration according to an embodiment of the present application; Figure 3A flowchart of convolutional neural network coding in a method for intelligent prediction of mineral resources based on human-machine fusion according to an embodiment of the present application; Figure 4 Flowchart of vector arrangement in the method for intelligent prediction of mineral resources based on human-machine fusion according to an embodiment of the present application; Figure 5 This is a flowchart of classification in the intelligent prediction method of mineral resources based on human-machine integration according to an embodiment of the present application; Figure 6 is a block diagram of a mineral resource intelligent prediction system based on human-machine integration according to an embodiment of the present application; Figure 7 is a block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0020] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described herein.

[0021] As mentioned in the previous background technology, although international oil production, consumption, imports, exports, and prices are in a state of constant fluctuation, the overall supply and demand of oil remains balanced, achieving dynamic oil security. When one or more events occur, this balance will be disrupted, leading to oil security issues. This will affect the entire upstream, midstream, and downstream of the oil industry chain. In other words, each link in oil production, transportation, processing, and consumption may be affected by multiple factors due to a single or multiple events, thus disrupting the regional, national, and even global oil security balance. However, traditional oil security strategic analysis methods have shortcomings such as untimely data updates, difficulty in processing complex data, and low efficiency, making them difficult to adapt to the requirements of rapid analysis in the big data era. Therefore, an optimized intelligent prediction system for mineral resources is desired.

[0022] Specifically, in the technical solution of the present application, with the support of technologies such as big data, cloud computing, and the Internet of Things, the spatiotemporal big data of oil safety is used as a data resource. When one or more events suddenly occur, key abnormal information is quickly extracted and identified, and the short-term or medium- and long-term impacts of one or more links in the entire industrial chain are analyzed. According to the warning level, response measures such as releasing reserves, increasing production, and traffic control are taken, thereby realizing a new "event-computation-response-response measure" working mode for oil safety.

[0023] Accordingly, considering the actual use of spatiotemporal oil security big data for risk anomaly warnings, and the subsequent implementation of countermeasures such as releasing reserves, increasing production, and controlling traffic according to the warning level, the most important thing is to analyze the spatiotemporal oil security big data to output an oil supply interruption risk level label. Specifically, the spatiotemporal oil security big data includes spatial data and attribute data. The spatial data is various types of oil-related data with spatial geometric features (points, lines, and surfaces). In other words, the spatial data is node data, which specifically includes oil-producing countries, importing countries, pipelines, export ports, import ports, straits, military bases, refineries, gas stations, etc. The attribute data is the attribute data of each node in the spatial data, which mainly includes global oil reserves, production (measured in units such as millions of tons and thousands of barrels per day), consumption, exports, imports, prices (year, month, day), GDP, population, urbanization rate, passenger volume, freight volume, and vehicle ownership. Some of this data is manually acquired, while others can be obtained through machine crawlers.

[0024] Based on this, considering that the multiple node data in the oil security spatiotemporal big data have spatial line and surface correlations, and the attribute data of the multiple nodes also have attribute correlations, it is not possible to conduct oil risk warnings based on changes in a single node or attribute data, which brings difficulties to the intelligent risk prediction of oil and mineral resources. In this process, the difficulty lies in how to establish a mapping relationship between the multiple nodes and multiple attribute data to accurately assess the level label of the oil supply interruption risk, so as to take corresponding response measures such as releasing reserves, increasing production, and traffic control based on the risk level label to ensure oil safety.

[0025] In recent years, deep learning and neural networks have been widely used in fields such as computer vision, natural language processing, and text signal processing. Furthermore, deep learning and neural networks have demonstrated capabilities approaching or even surpassing those of humans in areas such as image classification, object detection, semantic segmentation, and text translation.

[0026] The development of deep learning and neural networks has provided new solutions and solutions for mining the complex mapping relationships between multiple nodes and multiple attribute data. Those skilled in the art will appreciate that deep learning-based deep neural network models can be trained using appropriate strategies, such as a gradient descent backpropagation algorithm, to adjust the parameters of the deep neural network model so that it can simulate complex nonlinear relationships between objects. This is clearly suitable for simulating and establishing the complex mapping relationships between multiple nodes and multiple attribute data.

[0027] Figure 1This is a flowchart of a method for intelligent prediction of mineral resources based on human-machine integration according to an embodiment of the present application. Figure 2 FIG. 1 is a schematic diagram of the architecture of the intelligent prediction method for mineral resources based on human-machine fusion according to an embodiment of the present application. Figure 1 and Figure 2 As shown, according to the embodiment of the present application, the intelligent prediction method of mineral resources based on human-machine fusion includes the steps of: S110, obtaining multiple nodes related to oil security and attribute data of the multiple nodes, wherein the multiple nodes include oil-producing countries, importing countries, pipelines, export ports, import ports, straits, military bases, refineries and gas stations, and the attribute data include reserves, production, Consumption, export, import, price, GDP, population, urbanization rate, passenger volume, freight volume and car ownership; S120, arrange the attribute data of each node into an input vector and then pass it through a one-dimensional convolutional neural network model to obtain a plurality of node multidimensional data association feature vectors; S130, construct a spatial topology matrix of the plurality of nodes, wherein the eigenvalue of each position on the diagonal position in the spatial topology matrix is the distance between the corresponding two nodes; S140, pass the spatial topology matrix through a convolutional neural network model as a feature extractor to obtain an inter-node spatial topology feature matrix; S150, arrange the plurality of node multidimensional data association feature vectors into a global node multidimensional data association feature matrix; S160, pass the global node multidimensional data association feature matrix and the inter-node spatial topology feature matrix through a graph neural network model to obtain a spatial topology global node multidimensional data association feature matrix; S170, pass the spatial topology global node multidimensional data association feature matrix through a classifier to obtain a classification result, and the classification result is used to represent the oil supply interruption risk level label.

[0028] Specifically, in step S110, multiple nodes related to oil security and attribute data for these nodes are obtained. The multiple nodes include oil-producing countries, importing countries, pipelines, export ports, import ports, straits, military bases, refineries, and gas stations. The attribute data includes reserves, production, consumption, exports, imports, prices, GDP, population, urbanization rate, passenger volume, freight volume, and vehicle ownership. It should be understood that when using spatiotemporal oil security big data to generate risk anomaly warnings and implement response measures such as releasing reserves, increasing production, and controlling traffic based on the warning level, the most important step is to analyze the spatiotemporal oil security big data to output an oil supply interruption risk level label. In particular, the spatiotemporal big data of oil security includes spatial data and attribute data, wherein the spatial data is various types of oil-related data with spatial geometric features (points, lines, and surfaces), that is, the spatial data is node data, which specifically mainly includes oil-producing countries, importing countries, pipelines, export ports, import ports, straits, military bases, refineries, gas stations, etc. The attribute data is the attribute data of each node in the spatial data, which mainly includes the reserves, production (measured in units such as millions of tons and thousands of barrels per day), consumption, export volume, import volume, price (year, month, day), GDP, population, urbanization rate, passenger volume, freight volume, car ownership, etc. of oil in various countries around the world. In a specific example of the present application, among the multiple nodes related to oil security and the attribute data of the multiple nodes, the multiple nodes can be manually acquired and set, while the attribute data of the multiple nodes can be acquired by machine crawlers, so as to realize collaborative and interactive mineral resource prediction of human-machine integration.

[0029] Specifically, in step S120, the attribute data of each node is arranged into an input vector and then passed through a one-dimensional convolutional neural network model to obtain a plurality of node multidimensional data association feature vectors. In order to mine the implicit association feature information between the multiple attribute data of each node, the multiple attribute data of each node is arranged into an input vector to integrate the data distribution information between the multiple attribute data of each node. Then, the vector is passed through a one-dimensional convolutional neural network model for feature mining to extract high-dimensional implicit association feature distribution information between the multiple attribute data of each node, thereby obtaining a plurality of node multidimensional data association feature vectors.

[0030] That is, after the attribute data of each node is arranged as an input vector, the one-dimensional convolutional neural network model is used to mine the implicit correlation feature information between the multiple attribute data of each node. Specifically, first, the multiple attribute data of each node are arranged as an input vector to integrate the data distribution information between the multiple attribute data of each node, and then the one-dimensional convolutional neural network model is used to extract the high-dimensional implicit correlation feature distribution information between the multiple attribute data of each node, thereby obtaining multiple node multi-dimensional data correlation feature vectors. In a specific example of the present application, each layer of the one-dimensional convolutional neural network model is used to perform the following on the input data in the forward pass of the layer: convolution processing on the input data to obtain a convolution feature map; pooling based on the feature matrix on the convolution feature map to obtain a pooled feature map; and nonlinear activation on the pooled feature map to obtain an activation feature map; wherein the output of the last layer of the one-dimensional convolutional neural network is the node multi-dimensional data correlation feature vector, and the input of the first layer of the one-dimensional convolutional neural network is the attribute data input vector. Taking into account the relationship between each node.

[0031] Specifically, in step S130, a spatial topological matrix of the plurality of nodes is constructed, wherein the eigenvalue of each position on the diagonal position in the spatial topological matrix is the distance between the corresponding two nodes. Considering that the petroleum safety spatiotemporal big data includes spatial data and attribute data, wherein the spatial data is various types of petroleum-related data with spatial geometric features (points, lines, and surfaces), that is, the spatial data is node data, and the attribute data is multiple attribute data of each node in the spatial data. Therefore, there is a spatial topological association between the nodes, and the multiple attribute data of each node have a mutual data association relationship. Therefore, in the technical solution of the present application, a spatial topological matrix of the plurality of nodes is constructed based on the topological structure information between the plurality of nodes, wherein the eigenvalue of each position on the diagonal position in the spatial topological matrix is the distance between the corresponding two nodes.

[0032] Specifically, in step S140, the spatial topology matrix is passed through a convolutional neural network model as a feature extractor to obtain an inter-node spatial topology feature matrix. That is, a convolutional neural network model as a feature extractor that has excellent performance in extracting implicit correlation features is used to perform feature mining on the spatial topology matrix to extract spatial topology feature information about each node in the spatial topology matrix, thereby obtaining an inter-node spatial topology feature matrix. In a specific example, the convolutional neural network as a feature extractor includes multiple neural network layers cascaded together, wherein each neural network layer includes a convolution layer, a pooling layer, and an activation layer. In the encoding process of the convolutional neural network as a feature extractor, each layer of the convolutional neural network as a feature extractor performs convolution processing based on a convolution kernel on the input data using the convolution layer, performs pooling processing on the convolution feature map output by the convolution layer using the pooling layer, and performs activation processing on the pooled feature map output by the pooling layer using the activation layer during the forward pass of the layer.

[0033] Figure 3 The flowchart of the convolutional neural network coding in the mineral resource intelligent prediction method based on human-machine fusion according to the embodiment of the present application is as follows. Figure 3 As shown, in the encoding process of the convolutional neural network, it includes: using each layer of the convolutional neural network model as a feature extractor to perform the following on the input data in the forward pass of the layer: S210, convolution processing on the input data to obtain a convolution feature map; S220, pooling the convolution feature map along the channel dimension to obtain a pooled feature map; and, S230, performing nonlinear activation on the pooled feature map to obtain an activation feature map; wherein, the output of the last layer of the convolutional neural network as a feature extractor is the spatial topological feature matrix between nodes, and the input of the first layer of the convolutional neural network as a feature extractor is the spatial topological matrix.

[0034] Specifically, in step S150, the plurality of node multidimensional data association feature vectors are arranged into a global node multidimensional data association feature matrix. In the technical solution of the present application, the spatial topology global node multidimensional data association feature matrix is composed of the association feature representations of the node multidimensional data association feature vectors corresponding to each node under the spatial topology. Therefore, if the node data association feature vectors corresponding to each node can be weighted based on the association degree of the node multidimensional data association feature vectors of the plurality of nodes under the sample dimension, the feature representation of the node multidimensional data association feature vector of each node can be optimized, thereby improving the accuracy of the classification results of the spatial topology global node multidimensional data association feature matrix.

[0035] Based on this, the global node multidimensional data association feature matrix is firstly passed through the convolutional neural network as a feature extractor to obtain an auxiliary feature graph containing high-dimensional association representation. , used to express the high-dimensional correlation between samples of the multidimensional data association feature vector of the multiple nodes, wherein the number of channels of the convolutional neural network as a feature extractor is the same as the length of the multidimensional data association feature vector, and then the auxiliary feature map Perform feature-saliency-related reconstruction mechanism modeling to obtain weighted feature vectors The specific process includes: Calculate the auxiliary feature map The hyperbolic sine function value and the hyperbolic cosine function value of each eigenvalue of obtain the first characteristic map and the second feature map : ; ; in, and Represents the auxiliary feature map For each eigenvalue of , find the hyperbolic sine and cosine functions.

[0036] The first feature map And the second feature map Respectively and Activate to obtain the first significant correlation feature map and the second significant correlation feature map : ; ; Then the first significant correlation feature map And the second significant correlation feature map After point addition, a global pooling operation is performed on each feature matrix to obtain the weighted feature vector : ; in and are different preset weight matrices, Represents the global pooling operation of each feature matrix of the point plus feature map, Indicates point addition, Represents matrix multiplication.

[0037] Here, the feature saliency related reconstruction mechanism first performs a filtering operation on the feature map by using a hyperbolic sine function and a hyperbolic cosine function to model the auxiliary feature map while taking into account both class smoothness and class resolution. The local correlation of the global class propagation, and then through similar but different and The activation function performs different sparsity activations on the local correlation propagation representation in a neural network-like architecture to achieve reset and calibration of the feature distribution state of the non-aligned feature graph. Then, the sparse correlation of feature significance relative to the overall feature is accumulated through differentiated feature aggregation methods with different weight matrices, thereby explicitly modeling the significance weight sequence through significance prior reconstruction. Finally, the propagation hierarchical weight coefficient used to express the feature significance association is determined based on the channel weight parameter.

[0038] Then, the weighted feature vector By performing a dot product with the node multidimensional data associated feature vector corresponding to each node, the feature representation of the node multidimensional data associated feature vector for each node can be optimized, thereby improving the accuracy of the classification results of the spatial topology global node multidimensional data associated feature matrix. In this way, based on the actual spatiotemporal big data of oil security, it is possible to accurately assess the level label of the oil supply interruption risk through human-computer fusion, and then take corresponding countermeasures such as releasing reserves, increasing production, and controlling traffic based on the risk level label to ensure oil security.

[0039] Figure 4 FIG. 1 is a flow chart of vector arrangement in the method for intelligent prediction of mineral resources based on human-machine fusion according to an embodiment of the present application. Figure 4As shown, in the process of vector arrangement, it includes: S310, obtaining an auxiliary feature map by passing the global node multidimensional data association feature matrix through a convolutional neural network as a feature extractor; S320, modeling the auxiliary feature map based on a feature significance correlation reconstruction mechanism to obtain a weighted feature vector; S330, performing point multiplication of the weighted feature vector with each node multidimensional data association feature vector to obtain the multiple corrected node multidimensional data association feature vectors; and, S340, arranging the multiple corrected node multidimensional data association feature vectors into a global node multidimensional data association feature matrix. The global node multidimensional data association feature matrix is passed through a convolutional neural network as a feature extractor to obtain an auxiliary feature map, including: using each layer of the convolutional neural network as a feature extractor to perform the following on the input data in the forward pass of the layer: convolution processing on the input data to obtain a convolution feature map; pooling the convolution feature map based on a local feature matrix to obtain a pooled feature map; and nonlinear activation on the pooled feature map to obtain an activation feature map; wherein the output of the last layer of the convolutional neural network as a feature extractor is the auxiliary feature map, and the input of the first layer of the convolutional neural network as a feature extractor is the global node multidimensional data association feature matrix.

[0040] Specifically, in step S160, the global node multidimensional data association feature matrix and the inter-node spatial topology feature matrix are passed through a graph neural network model to obtain a spatial topology global node multidimensional data association feature matrix. Considering that each node contains the attribute data of the node, in the technical solution of the present application, each node multidimensional data association feature vector in the multiple node multidimensional data association feature vectors is used as the feature representation of the node, and the inter-node spatial topology feature matrix is used as the feature representation of the edge between the nodes. The global node multidimensional data association feature matrix obtained by two-dimensional arrangement of the multiple node multidimensional data association feature vectors and the inter-node spatial topology feature matrix are passed through a graph neural network to obtain a spatial topology global node multidimensional data association feature matrix. In a specific example of the present application, the graph neural network model is used to perform graph structure encoding on the global node multidimensional data association feature matrix and the inter-node spatial topology feature matrix with learnable neural network parameters to obtain the spatial topology global node multidimensional data association feature matrix containing irregular node spatial topology features and multiple attribute association features of each node.

[0041] Specifically, in step S170, the spatial topology global node multidimensional data association feature matrix is passed through a classifier to obtain a classification result, and the classification result is used to represent the oil supply interruption risk level label. That is, the spatial topology global node multidimensional data association feature matrix is used as a classification feature matrix and passed through a classifier to obtain a classification result. Specifically, the classifier processes the spatial topology global node multidimensional data association feature matrix using the following formula to obtain a classification result, wherein the formula is: ; in It represents the projection of the spatial topology global node multidimensional data association feature matrix into a vector, to is the weight matrix of each fully connected layer, to Represents the bias vector of each fully connected layer. Specifically, the classifier includes multiple fully connected layers and a Softmax layer cascaded with the last fully connected layer of the multiple fully connected layers. In the classification process of the classifier, the spatial topology global node multidimensional data association feature matrix is first projected into a vector. For example, in a specific example, the spatial topology global node multidimensional data association feature matrix is expanded along the row vector or column vector to form a classification feature vector. Then, the classification feature vector is fully connected multiple times using the multiple fully connected layers of the classifier to obtain an encoded classification feature vector. Furthermore, the encoded classification feature vector is input into the Softmax layer of the classifier, that is, the encoded classification feature vector is classified using the Softmax classification function to obtain a classification label. In the technical solution of the present application, the label of the classifier is an oil supply interruption risk level label, wherein the classifier determines to which classification label the classification feature matrix belongs using a soft maximum function. In this way, the level label of the oil supply interruption risk can be accurately assessed, and corresponding response measures such as releasing reserves, increasing production, and traffic control can be taken based on the risk level label to ensure oil safety.

[0042] Figure 5 This is a flow chart of the classification of the intelligent prediction method of mineral resources based on human-machine integration according to the embodiment of the present application. Figure 5 As shown, in the classification process, it includes: S410, expanding the spatial topology global node multidimensional data association feature matrix into a classification feature vector based on a row vector or a column vector; S420, using multiple fully connected layers of the classifier to fully connect encode the classification feature vector to obtain an encoded classification feature vector; and, S430, passing the encoded classification feature vector through the Softmax classification function of the classifier to obtain the classification result.

[0043] In summary, according to the embodiment of the present application, the intelligent prediction method of mineral resources based on human-machine integration is explained. It adopts a neural network model based on deep learning and establishes a mapping relationship between multiple nodes and multiple attribute data in the spatiotemporal big data of oil safety to accurately evaluate the level label of the oil supply interruption risk, so as to take corresponding response measures such as releasing reserves, increasing production, and traffic control based on the risk level label to ensure the safety of oil.

[0044] Example systems: Figure 6 FIG is a block diagram of a mineral resource intelligent prediction system based on human-machine fusion according to an embodiment of the present application. Figure 6 As shown, the intelligent mineral resource prediction system 300 based on human-machine integration according to the embodiment of the present application includes: an attribute data acquisition module 310; a one-dimensional convolution module 320; a spatial topology construction module 330; a spatial topology feature extraction module 340; an arrangement module 350; a graph feature extraction module 360; and a classification result generation module 370.

[0045] Among them, the attribute data acquisition module 310 is used to obtain multiple nodes related to oil security and attribute data of the multiple nodes, wherein the multiple nodes include oil-producing countries, importing countries, pipelines, export ports, import ports, straits, military bases, refineries and gas stations, and the attribute data include reserves, production, consumption, export volume, import volume, price, GDP, population, urbanization rate, passenger volume, freight volume and car ownership; the one-dimensional convolution module 320 is used to arrange the attribute data of each node into an input vector and then pass it through a one-dimensional convolutional neural network model to obtain a plurality of node multidimensional data associated feature vectors; the spatial topology construction module 330 is used to construct a spatial topology matrix of the multiple nodes, wherein the eigenvalue of each position on the diagonal position in the spatial topology matrix is the distance between the corresponding two nodes; the spatial topology feature extraction module 340 is used to convolve the spatial topology matrix as a feature extractor A neural network model is used to obtain a spatial topology feature matrix between nodes; the arrangement module 350 is used to arrange the multiple node multidimensional data association feature vectors into a global node multidimensional data association feature matrix; the graph feature extraction module 360 is used to pass the global node multidimensional data association feature matrix and the spatial topology feature matrix between nodes through a graph neural network model to obtain a spatial topology global node multidimensional data association feature matrix; the classification result generation module 370 is used to pass the spatial topology global node multidimensional data association feature matrix through a classifier to obtain a classification result, and the classification result is used to represent the oil supply interruption risk level label.

[0046] In one example, in the above-mentioned intelligent prediction system 300 for mineral resources based on human-machine fusion, the one-dimensional convolution module 320 is used to: use each layer of the one-dimensional convolutional neural network model to perform the following on the input data in the forward pass of the layer: convolution processing on the input data to obtain a convolution feature map; pooling based on the feature matrix on the convolution feature map to obtain a pooled feature map; and, nonlinear activation on the pooled feature map to obtain an activation feature map; wherein the output of the last layer of the one-dimensional convolutional neural network is the node multidimensional data association feature vector, and the input of the first layer of the one-dimensional convolutional neural network is the attribute data input vector.

[0047] In one example, in the above-mentioned intelligent prediction system 300 for mineral resources based on human-machine fusion, the spatial topology feature extraction module 340 is used to: use the layers of the convolutional neural network model as a feature extractor to perform the following on the input data in the forward pass of the layer: convolution processing on the input data to obtain a convolution feature map; pooling the convolution feature map along the channel dimension to obtain a pooled feature map; and, nonlinear activation on the pooled feature map to obtain an activation feature map; wherein the output of the last layer of the convolutional neural network as a feature extractor is the spatial topology feature matrix between nodes, and the input of the first layer of the convolutional neural network as a feature extractor is the spatial topology matrix.

[0048] In one example, in the above-mentioned intelligent prediction system 300 for mineral resources based on human-machine fusion, the arrangement module 350 is used to: obtain an auxiliary feature map by using a convolutional neural network as a feature extractor for the global node multidimensional data association feature matrix; model the auxiliary feature map based on a feature significance correlation reconstruction mechanism to obtain a weighted feature vector; perform point multiplication of the weighted feature vector with each node multidimensional data association feature vector to obtain the multiple corrected node multidimensional data association feature vectors; and arrange the multiple corrected node multidimensional data association feature vectors into a global node multidimensional data association feature matrix. wherein, the global node multidimensional data association feature matrix is passed through a convolutional neural network as a feature extractor to obtain an auxiliary feature map, including: using each layer of the convolutional neural network as a feature extractor to perform the following on the input data in the forward pass of the layer: convolution processing on the input data to obtain a convolution feature map; pooling the convolution feature map based on the local feature matrix to obtain a pooling feature map; and nonlinear activation on the pooling feature map to obtain an activation feature map; wherein the output of the last layer of the convolutional neural network as a feature extractor is the auxiliary feature map, and the input of the first layer of the convolutional neural network as a feature extractor is the global node multidimensional data association feature matrix; more specifically, modeling the auxiliary feature map based on a feature significance correlation reconstruction mechanism to obtain a weighted feature vector, including: calculating the hyperbolic sine function value and the hyperbolic cosine function value of each eigenvalue of the auxiliary feature map to obtain a first feature map and a second feature map; the first feature map and the second feature map are respectively and Activation is performed to obtain a first saliency-related feature map and a second saliency-related feature map; after dot-adding the first saliency-related feature map and the second saliency-related feature map, a global pooling operation is performed on each feature matrix to obtain the weighted feature vector.

[0049] In one example, in the above-mentioned intelligent prediction system 300 for mineral resources based on human-machine fusion, the graph feature extraction module 360 is used to: use the graph neural network model to perform graph structure encoding on the global node multidimensional data association feature matrix and the inter-node spatial topology feature matrix with learnable neural network parameters to obtain the spatial topology global node multidimensional data association feature matrix containing irregular node spatial topology features and multiple attribute association features of each node.

[0050] In one example, in the above-mentioned intelligent mineral resource prediction system 300 based on human-machine fusion, the classification result generation module 370 is used to: expand the spatial topology global node multidimensional data association feature matrix into a classification feature vector based on a row vector or a column vector; use the multiple fully connected layers of the classifier to fully connect encode the classification feature vector to obtain an encoded classification feature vector; and pass the encoded classification feature vector through the Softmax classification function of the classifier to obtain the classification result.

[0051] In summary, according to the embodiment of the present application, the intelligent prediction system 300 for mineral resources based on human-machine integration is explained. It adopts a neural network model based on deep learning and establishes a mapping relationship between multiple nodes and multiple attribute data in the spatiotemporal big data of oil safety to accurately evaluate the level label of the oil supply interruption risk, so as to take corresponding response measures such as releasing reserves, increasing production, and traffic control based on the risk level label to ensure the safety of oil.

[0052] As described above, the intelligent prediction system for mineral resources based on human-machine fusion according to the embodiment of the present application can be implemented in various terminal devices. In one example, the intelligent prediction system for mineral resources based on human-machine fusion according to the embodiment of the present application 300 can be integrated into the terminal device as a software module and / or hardware module. For example, the intelligent prediction system for mineral resources based on human-machine fusion 300 can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the intelligent prediction system for mineral resources based on human-machine fusion 300 can also be one of the many hardware modules of the terminal device.

[0053] Alternatively, in another example, the intelligent prediction system 300 for mineral resources based on human-machine fusion and the terminal device may also be separate devices, and the intelligent prediction system 300 for mineral resources based on human-machine fusion may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0054] Example electronic devices: Below, reference Figure 7 To describe the electronic device according to the embodiment of the present application.

[0055] Figure 7 The figure shows a block diagram of an electronic device according to an embodiment of the present application.

[0056] like Figure 7 As shown, the electronic device 10 includes one or more processors 11 and a memory 12 .

[0057] The processor 11 may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 10 to perform desired functions.

[0058] The memory 12 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), a hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 11 may execute the program instructions to implement the functions of the intelligent mineral resource prediction method based on human-computer integration described in the various embodiments of the present application as described above and / or other desired functions. The computer-readable storage medium may also store various contents such as a spatial topology global node multidimensional data association feature matrix.

[0059] In one example, the electronic device 10 may further include an input device 13 and an output device 14 , and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0060] The input device 13 may include, for example, a keyboard, a mouse, and the like.

[0061] The output device 14 can output various information to the outside, including classification results, etc. The output device 14 can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, and the like.

[0062] Of course, to simplify, Figure 7 Only some of the components related to the present application in the electronic device 10 are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, the electronic device 10 may further include any other appropriate components according to specific application scenarios.

[0063] Exemplary computer program products and computer-readable storage media: In addition to the above-mentioned methods and devices, an embodiment of the present application may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the steps in the functions of the intelligent prediction method of mineral resources based on human-computer fusion according to various embodiments of the present application described in the above "Exemplary Method" section of this specification.

[0064] The computer program product may be written in any combination of one or more programming languages to implement the program code for performing the operations of the embodiments of the present application, including object-oriented programming languages such as Java, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0065] In addition, an embodiment of the present application may also be a computer-readable storage medium having computer program instructions stored thereon, which, when executed by a processor, enable the processor to execute the steps in the functions of the intelligent prediction method for mineral resources based on human-machine fusion according to various embodiments of the present application described in the above "Exemplary Method" section of this specification.

[0066] The computer-readable storage medium may be any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may include, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0067] The basic principles of the present application have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, and effects mentioned in this application are merely illustrative and not restrictive, and it should not be assumed that these advantages, strengths, and effects are required of each embodiment of this application. In addition, the specific details disclosed above are merely illustrative and facilitating understanding, and are not restrictive. The above details do not limit this application to necessarily being implemented using the above specific details.

[0068] The block diagrams of the devices, devices, equipment, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As will be appreciated by those skilled in the art, these devices, devices, equipment, and systems can be connected, arranged, or configured in any manner. Words such as "include," "comprise," "have," and the like are open-ended words, meaning "including but not limited to," and can be used interchangeably therewith. The words "or" and "and" used herein refer to the words "and / or" and can be used interchangeably therewith, unless the context clearly indicates otherwise. The word "such as" used herein refers to the phrase "such as but not limited to," and can be used interchangeably therewith.

[0069] It should also be noted that in the apparatus, device, and method of the present application, each component or each step can be decomposed and / or recombined, and such decomposition and / or recombination should be regarded as equivalent solutions of the present application.

[0070] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0071] The above description has been provided for the purpose of illustration and description. Furthermore, this description is not intended to limit the embodiments of the present application to the forms disclosed herein. Although a number of example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. A method for intelligent prediction of mineral resources based on human-machine integration, characterized in that: include: Acquiring multiple nodes related to oil security and attribute data of the multiple nodes, wherein the multiple nodes include oil-producing countries, importing countries, pipelines, export ports, import ports, straits, military bases, refineries, and gas stations, and the attribute data includes reserves, production, consumption, export volume, import volume, price, GDP, population, urbanization rate, passenger volume, freight volume, and vehicle ownership; Arranging the attribute data of each node into an input vector and then passing it through a one-dimensional convolutional neural network model to obtain a multi-dimensional data association feature vector of multiple nodes; Constructing a spatial topological matrix of the plurality of nodes, wherein the eigenvalue of each diagonal position in the spatial topological matrix is the distance between the corresponding two nodes; Passing the spatial topology matrix through a convolutional neural network model as a feature extractor to obtain an inter-node spatial topology feature matrix; Arranging the plurality of node multidimensional data association feature vectors into a global node multidimensional data association feature matrix; The global node multidimensional data association feature matrix and the inter-node spatial topology feature matrix are passed through a graph neural network model to obtain a spatial topology global node multidimensional data association feature matrix; The spatial topology global node multidimensional data association feature matrix is passed through a classifier to obtain a classification result, and the classification result is used to represent an oil supply interruption risk level label.

2. The method for intelligent prediction of mineral resources based on human-machine integration according to claim 1 is characterized in that: Arranging the attribute data of each node into an input vector and then passing it through a one-dimensional convolutional neural network model to obtain a plurality of node multi-dimensional data associated feature vectors, including: using each layer of the one-dimensional convolutional neural network model to perform the following on the input data in the forward pass of the layer: Perform convolution on the input data to obtain a convolution feature map; Performing feature matrix-based pooling on the convolution feature map to obtain a pooled feature map; and Performing nonlinear activation on the pooled feature map to obtain an activated feature map; Among them, the output of the last layer of the one-dimensional convolutional neural network is the node multi-dimensional data association feature vector, and the input of the first layer of the one-dimensional convolutional neural network is the attribute data input vector.

3. The method for intelligent prediction of mineral resources based on human-machine integration according to claim 2 is characterized in that: The spatial topology matrix is passed through a convolutional neural network model as a feature extractor to obtain an inter-node spatial topology feature matrix, comprising: using each layer of the convolutional neural network model as a feature extractor to perform the following on the input data in a forward pass of the layer: Perform convolution on the input data to obtain a convolution feature map; Pooling the convolutional feature map along the channel dimension to obtain a pooled feature map; and Performing nonlinear activation on the pooled feature map to obtain an activated feature map; The output of the last layer of the convolutional neural network as a feature extractor is the spatial topology feature matrix between nodes, and the input of the first layer of the convolutional neural network as a feature extractor is the spatial topology matrix.

4. The method for intelligent prediction of mineral resources based on human-machine integration according to claim 3 is characterized in that: Arranging the plurality of node multidimensional data association feature vectors into a global node multidimensional data association feature matrix comprises: The global node multidimensional data correlation feature matrix is passed through a convolutional neural network as a feature extractor to obtain an auxiliary feature graph; Modeling the auxiliary feature map based on a feature saliency-related reconstruction mechanism to obtain a weighted feature vector; Performing point multiplication of the weighted feature vector and each node multidimensional data associated feature vector to obtain the plurality of corrected node multidimensional data associated feature vectors; and The plurality of corrected node multidimensional data association feature vectors are arranged into a global node multidimensional data association feature matrix.

5. The method for intelligent prediction of mineral resources based on human-machine integration according to claim 4 is characterized in that: The global node multidimensional data correlation feature matrix is passed through a convolutional neural network as a feature extractor to obtain an auxiliary feature map, including: using each layer of the convolutional neural network as a feature extractor to perform the following on the input data in the forward pass of the layer: Perform convolution on the input data to obtain a convolution feature map; Performing pooling on the convolutional feature map based on a local feature matrix to obtain a pooled feature map; and Performing nonlinear activation on the pooled feature map to obtain an activated feature map; Among them, the output of the last layer of the convolutional neural network serving as a feature extractor is the auxiliary feature map, and the input of the first layer of the convolutional neural network serving as a feature extractor is the global node multidimensional data association feature matrix.

6. The method for intelligent prediction of mineral resources based on human-machine integration according to claim 5 is characterized in that: Modeling the auxiliary feature map based on a feature saliency-related reconstruction mechanism to obtain a weighted feature vector includes: Calculating a hyperbolic sine function value and a hyperbolic cosine function value of each eigenvalue of the auxiliary feature map to obtain a first feature map and a second feature map; The first feature map and the second feature map are respectively and Performing activation to obtain a first saliency-related feature map and a second saliency-related feature map; After performing point-wise addition on the first saliency-related feature map and the second saliency-related feature map, a global pooling operation is performed on each feature matrix to obtain the weighted feature vector.

7. The method for intelligent prediction of mineral resources based on human-machine integration according to claim 6 is characterized in that: The global node multidimensional data association feature matrix and the corrected inter-node spatial topology feature matrix are passed through a graph neural network model to obtain a spatial topology global node multidimensional data association feature matrix, including: using the graph neural network model to perform graph structure encoding on the global node multidimensional data association feature matrix and the inter-node spatial topology feature matrix with learnable neural network parameters to obtain the spatial topology global node multidimensional data association feature matrix containing irregular node spatial topology features and multiple attribute association features of each node.

8. The method for intelligent prediction of mineral resources based on human-machine integration according to claim 7 is characterized in that: The spatial topology global node multidimensional data association feature matrix is passed through a classifier to obtain a classification result, and the classification result is used to represent the oil supply interruption risk level label, including: Expanding the spatial topology global node multidimensional data association feature matrix into a classification feature vector based on row vectors or column vectors; Performing full-connection encoding on the classification feature vector using multiple fully-connected layers of the classifier to obtain an encoded classification feature vector; and The encoded classification feature vector is passed through the Softmax classification function of the classifier to obtain the classification result.

9. An intelligent mineral resource prediction system based on human-machine integration, characterized in that: include: an attribute data acquisition module, configured to acquire multiple nodes related to oil security and attribute data of the multiple nodes, wherein the multiple nodes include oil-producing countries, importing countries, pipelines, export ports, import ports, straits, military bases, refineries, and gas stations, and the attribute data includes reserves, production, consumption, exports, imports, prices, GDP, population, urbanization rate, passenger volume, freight volume, and vehicle ownership; A one-dimensional convolution module, configured to arrange the attribute data of each node into an input vector and then pass it through a one-dimensional convolutional neural network model to obtain a multi-dimensional data association feature vector for multiple nodes; A spatial topology construction module, configured to construct a spatial topology matrix of the plurality of nodes, wherein the eigenvalue of each diagonal position in the spatial topology matrix is the distance between the corresponding two nodes; A spatial topology feature extraction module, configured to pass the spatial topology matrix through a convolutional neural network model as a feature extractor to obtain an inter-node spatial topology feature matrix; an arrangement module, configured to arrange the plurality of node multidimensional data association feature vectors into a global node multidimensional data association feature matrix; A graph feature extraction module is used to obtain a spatial topology global node multidimensional data association feature matrix by passing the global node multidimensional data association feature matrix and the inter-node spatial topology feature matrix through a graph neural network model; The classification result generating module is used to pass the spatial topology global node multidimensional data association feature matrix through a classifier to obtain a classification result, and the classification result is used to represent the oil supply interruption risk level label.

10. The mineral resource intelligent prediction system based on human-machine integration according to claim 9 is characterized in that: The classification result generation module is used to: Expanding the spatial topology global node multidimensional data association feature matrix into a classification feature vector based on row vectors or column vectors; Performing full-connection encoding on the classification feature vector using multiple fully-connected layers of the classifier to obtain an encoded classification feature vector; as well as The encoded classification feature vector is passed through the Softmax classification function of the classifier to obtain the classification result.

Images