Mine roadway design model training method, mine roadway design and device

Abstract

The invention discloses a mine roadway design model training method and a mine roadway design and device, and relates to the technical field of deep learning and the technical field of intelligent mines, and the method comprises the steps: obtaining graph structure data for model training of a mine, the graph structure data comprising non-roadway data and roadway data; inputting the non-roadway data into a to-be-trained graph neural network to obtain roadway design data output by the graph neural network, the roadway design data including roadway design node data and roadway design edge data, and the roadway design edge data representing a spatial relationship between each non-roadway node and each roadway design node; and adjusting network parameters of the graph neural network based on errors between the roadway design data and the roadway data. By adopting the scheme, the design efficiency of mine roadway design is improved.

Description

Technical Field

[0001] The present application relates to the fields of deep learning technology and intelligent mining technology, and in particular to a mine tunnel design model training method, mine tunnel design and device. Background Art

[0002] Smart mines are based on the digitalization and informatization of mines. They actively perceive, automatically analyze, and quickly process mine production, occupational health and safety, technical support, and logistics, building smart mines and ultimately achieving the construction of safe, unmanned, efficient, and clean mines.

[0003] The smart mining system includes all aspects of the mine. According to the usual classification method, it can be divided into three aspects: smart production system, smart occupational health and safety system, and smart technical support and logistics guarantee system construction.

[0004] More specifically, with the continuous advancement of smart mine construction, the way mine information is presented is gradually shifting from the traditional two-dimensional plane to three-dimensional space. A 3D mine scene can be built for the mine to show the entire mine and achieve mine transparency.

[0005] Digital twin technology comprehensively utilizes sensors, Internet of Things, virtual reality, artificial intelligence and other technologies to describe and model the characteristics, behaviors, operating processes and performance of physical objects in the real world.

[0006] Applying digital twin technology to the mining industry to realize digital twins of mines can enable a simulated display of underground mine scenes, making it easier for operators to know the underground mine scene conditions in real time, so that mine management affairs can be carried out more safely and efficiently.

[0007] In mine digital twin technology, it is necessary to perform three-dimensional modeling of the underground mine scene, generate a mine model, and display the mine model in the application.

[0008] In the application of the above-mentioned smart mining technologies, it is necessary to design CAD drawings for the underground mine and then build a three-dimensional model of the mine. It can be seen that underground mine design is the prerequisite for the application of various smart mining technologies.

[0009] In underground mine design, the design of mine tunnels is the main design content. At present, mine tunnel design is still done manually based on experience, which has low design efficiency. Summary of the Invention

[0010] The embodiments of the present application provide a mine tunnel design model training method, mine tunnel design and device, which are used to solve the problem of low design efficiency of mine tunnel design in the prior art.

[0011] The present invention provides a method for training a mine tunnel design model, including:

[0012] Obtaining graph structure data of the mine for model training, the graph structure data comprising: non-roadway data and roadway data, the non-roadway data comprising non-roadway node data and non-roadway edge data, the roadway data comprising roadway node data and roadway edge data, wherein the non-roadway edge data represents the spatial relationship between each non-roadway node, and the roadway edge data represents the spatial relationship between each non-roadway node and each roadway node;

[0013] Inputting the non-laneway data into a graph neural network to be trained to obtain laneway design data output by the graph neural network, wherein the laneway design data includes laneway design node data and laneway design edge data, wherein the laneway design edge data represents the spatial relationship between each non-laneway node and each laneway design node;

[0014] Based on the error between the lane design data and the lane data, the network parameters of the graph neural network are adjusted.

[0015] Furthermore, the graph structure data for model training is generated using the following steps:

[0016] Acquire CAD drawing data of a mine, wherein the CAD drawing data includes graphic information of the mine;

[0017] Converting the CAD drawing data into mine data in a GIS exchange format, wherein the mine data in the GIS exchange format includes non-roadway mine data and roadway mine data;

[0018] Based on the mine data in the GIS exchange format, graph structure data for model training is generated.

[0019] Furthermore, adjusting the network parameters of the graph neural network based on the error between the lane design data and the lane data includes:

[0020] Calculating a mean square error between the lane design data and the lane data as a value of a loss function of the graph neural network;

[0021] When the mean square error is not less than a preset error threshold, the network parameters of the graph neural network are adjusted.

[0022] Furthermore, the non-roadway nodes include: ore body contour lines and chute;

[0023] The lane node is the lane centerline;

[0024] The tunnel design node is the tunnel design centerline.

[0025] The present application also provides a mine tunnel design method, including:

[0026] Acquire non-roadway graph structure data of a mine to be designed, wherein the non-roadway graph structure data includes: non-roadway node data and non-roadway edge data, wherein the non-roadway edge data represents the spatial relationship between each non-roadway node;

[0027] The non-laneway graph structure data is input into a graph neural network to obtain laneway design data output by the graph neural network. The laneway design data serves as laneway graph structure data. The laneway design data includes laneway design node data and laneway design edge data. The laneway design edge data represents the spatial relationship between each non-laneway node and each laneway design node.

[0028] Furthermore, it also includes:

[0029] Converting the lane map structure data into lane data in a GIS exchange format;

[0030] The tunnel data in the GIS exchange format is converted into CAD drawing data of the tunnel.

[0031] Furthermore, the non-lane graph structure data is generated by the following steps:

[0032] Acquire non-tunnel CAD drawing data of a mine, where the non-tunnel CAD drawing data includes graphic information of the non-tunnel of the mine;

[0033] Convert non-laneway CAD drawing data into non-laneway data in GIS exchange format;

[0034] Based on the non-lane data in the GIS exchange format, non-lane graph structure data is generated.

[0035] Furthermore, the non-roadway nodes include: ore body contour lines and chute;

[0036] The tunnel design node is the tunnel design centerline.

[0037] The present application also provides a mine tunnel design model training device, comprising:

[0038] a training data acquisition module, configured to acquire graph structure data of the mine for model training, wherein the graph structure data includes: non-roadway data and roadway data, wherein the non-roadway data includes non-roadway node data and non-roadway edge data, and the roadway data includes roadway node data and roadway edge data, wherein the non-roadway edge data represents the spatial relationship between each non-roadway node, and the roadway edge data represents the spatial relationship between each non-roadway node and each roadway node;

[0039] a training data input module, configured to input the non-laneway data into a graph neural network to be trained, and obtain laneway design data output by the graph neural network, wherein the laneway design data includes laneway design node data and laneway design edge data, wherein the laneway design edge data represents the spatial relationship between each non-laneway node and each laneway design node;

[0040] A model training module is used to adjust the network parameters of the graph neural network based on the error between the lane design data and the lane data.

[0041] Furthermore, the training data acquisition module is further configured to generate the graph structure data for model training by adopting the following steps:

[0042] Acquire CAD drawing data of a mine, wherein the CAD drawing data includes graphic information of the mine; convert the CAD drawing data into mine data in a GIS exchange format, wherein the mine data in the GIS exchange format includes non-roadway mine data and roadway mine data; and generate graph structure data for model training based on the mine data in the GIS exchange format.

[0043] Furthermore, the model training module is specifically used to calculate the mean square error between the tunnel design data and the tunnel data as the value of the loss function of the graph neural network; when the mean square error is not less than a preset error threshold, the network parameters of the graph neural network are adjusted.

[0044] Furthermore, the non-roadway nodes include: ore body contour lines and chute;

[0045] The lane node is the lane centerline;

[0046] The tunnel design node is the tunnel design centerline.

[0047] The present application also provides a mine tunnel design device, comprising:

[0048] A non-roadway data acquisition module is used to acquire non-roadway graph structure data of the mine to be designed, wherein the non-roadway graph structure data includes: non-roadway node data and non-roadway edge data, wherein the non-roadway edge data represents the spatial relationship between each non-roadway node;

[0049] A lane design module is used to input the non-lane graph structure data into a graph neural network to obtain lane design data output by the graph neural network. The lane design data serves as lane graph structure data. The lane design data includes lane design node data and lane design edge data. The lane design edge data represents the spatial relationship between each non-lane node and each lane design node.

[0050] Furthermore, the lane design module is also used to convert the lane diagram structure data into lane data in a GIS exchange format; and convert the lane data in a GIS exchange format into CAD drawing data of the lane.

[0051] Furthermore, the non-laneway data acquisition module is further configured to generate the non-laneway graph structure data by adopting the following steps:

[0052] Acquire non-tunnel CAD drawing data of a mine, wherein the non-tunnel CAD drawing data includes graphic information of the non-tunnel of the mine; convert the non-tunnel CAD drawing data into non-tunnel data in a GIS exchange format; and generate non-tunnel graph structure data based on the non-tunnel data in the GIS exchange format.

[0053] Furthermore, the non-roadway nodes include: ore body contour lines and chute;

[0054] The tunnel design node is the tunnel design centerline.

[0055] An embodiment of the present application also provides an electronic device, including a processor and a machine-readable storage medium, wherein the machine-readable storage medium stores machine-executable instructions that can be executed by the processor, and the processor is prompted by the machine-executable instructions to: implement any of the above-mentioned mine tunnel design model training methods, or implement any of the above-mentioned mine tunnel design methods.

[0056] An embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements any of the above-mentioned mine tunnel design model training methods, or implements any of the above-mentioned mine tunnel design methods.

[0057] An embodiment of the present application further provides a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute any of the above-mentioned mine tunnel design model training methods, or to execute any of the above-mentioned mine tunnel design methods.

[0058] The beneficial effects of this application include:

[0059] In the method provided in the embodiment of the present application, graph structure data of a mine for model training is obtained, non-tunnel data in the graph structure data is input into the graph neural network to be trained, tunnel design data output by the graph neural network is obtained, and the network parameters of the graph neural network are adjusted based on the error between the obtained tunnel design data and the tunnel data in the graph structure data, thereby achieving effective training of the graph neural network. Based on the trained graph neural network, when it is necessary to design a mine tunnel, non-tunnel graph structure data of the mine to be designed can be obtained, and this non-tunnel graph structure data can be input into the graph neural network to obtain output tunnel design data, which improves design efficiency compared to manual design in the prior art.

[0060] Other features and advantages of the present application will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present application. The purposes and other advantages of the present application can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] The accompanying drawings are used to provide a further understanding 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 accompanying drawings:

[0062] Figure 1 A flowchart of a mine tunnel design model training method provided in an embodiment of the present application;

[0063] Figure 2 A flowchart for generating graph structure data for model training in the mine tunnel design model training method provided in an embodiment of the present application;

[0064] Figure 3 A flowchart for adjusting network parameters of a graph neural network in the mine tunnel design model training method provided in an embodiment of the present application;

[0065] Figure 4 A flow chart of a mine tunnel design method provided in an embodiment of the present application;

[0066] Figure 5 A schematic diagram of the structure of a mine tunnel design model training device provided in an embodiment of the present application;

[0067] Figure 6 A schematic diagram of the structure of a mine tunnel design device provided in an embodiment of the present application;

[0068] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0069] To provide an implementation plan for improving the design efficiency of mine tunnel design, the present application provides a mine tunnel design model training method, mine tunnel design, and device. The preferred embodiments of the present application are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only intended to illustrate and explain the present application and are not intended to limit the present application. Furthermore, the embodiments and features within the embodiments of the present application may be combined with each other unless there is a conflict.

[0070] The present application embodiment provides a mine tunnel design model training method, such as Figure 1 As shown, including:

[0071] Step 11: Obtain graph structure data of the mine for model training. The graph structure data includes: non-roadway data and roadway data. The non-roadway data includes non-roadway node data and non-roadway edge data. The roadway data includes roadway node data and roadway edge data. The non-roadway edge data represents the spatial relationship between non-roadway nodes, and the roadway edge data represents the spatial relationship between non-roadway nodes and roadway nodes.

[0072] Step 12: Input the non-laneway data into the graph neural network to be trained to obtain the laneway design data output by the graph neural network. The laneway design data includes laneway design node data and laneway design edge data. The laneway design edge data represents the spatial relationship between each non-laneway node and each laneway design node.

[0073] Step 13: Based on the error between the lane design data and the lane data, adjust the network parameters of the graph neural network.

[0074] The above-mentioned mine tunnel design model training method provided in the embodiment of the present application is adopted to obtain the graph structure data of the mine for model training, and the non-tunnel data in the graph structure data is input into the graph neural network to be trained to obtain the tunnel design data output by the graph neural network. Based on the error between the obtained tunnel design data and the tunnel data in the graph structure data, the network parameters of the graph neural network are adjusted to achieve effective training of the graph neural network.

[0075] In one embodiment of the present application, the following steps may be used to generate graph structure data for model training: Figure 2 As shown, including:

[0076] Step 21: Acquire CAD drawing data of the mine, where the CAD drawing data includes graphic information of the mine.

[0077] The CAD drawing data obtained in this step is the underground mine design data for a certain horizontal line in the mine, and is the CAD drawing data of the completed design, which contains graphic information of the mine, such as layers, lines, etc.

[0078] The file format is .dwg, stored in text or binary form, and contains the geometric information and attributes of the graphics, which reflects the design of some scenes in the mine, including the design of the tunnel.

[0079] Step 22: Convert the CAD drawing data into mining data in a GIS exchange format. The mining data in the GIS exchange format includes non-tunnel mining data and tunnel mining data.

[0080] In this step, the mining data in the GIS exchange format obtained by conversion can be Geojson data, and the data types can include point (Point), line (LineString), multiline (MultiLineString), surface (Polygon), etc. The format structure can be expressed as follows:

[0081]

[0082] In the embodiment of the present application, the mine data in the GIS exchange format may include non-roadway mine data and roadway mine data, wherein the non-roadway mine data may include ore body contour line data and chute data. The GIS data structure of the ore body contour line, chute and roadway is shown as follows:

[0083] Ore body outline:

[0084] Ore body unique identification ID;

[0085] Ore body coordinates: a series of coordinate points (in XYZ format) that form the boundary of the ore body. The ore body data type is a closed line (LineString) or a surface (Polygon);

[0086] Ore body related attributes: MineType ore body type (rock, ore, etc.), mining area number, etc.

[0087] Well:

[0088] The unique identifier of the well;

[0089] Well coordinates: a series of coordinate points (in XYZ format) that form the boundary of the well. The well data type is a polygon.

[0090] Passage-related attributes: orePassType, pass type (hoisting shaft, ventilation shaft, cutting shaft, cable shaft, elevator shaft, etc.), mining area number, etc.

[0091] Laneway:

[0092] Lane unique identification id;

[0093] Lane coordinates: a series of coordinate points (in XYZ format) that form the lane centerline. The lane data type is LineString.

[0094] Tunnel related attributes: TunnelType tunnel type (along-vein tunnel, through-vein tunnel, off-vein connecting tunnel, return air shaft connecting tunnel, chute connecting tunnel, etc.), mine area number, etc.

[0095] Step 23: Generate graph structure data for model training based on the mine data in GIS exchange format.

[0096] Graph data consists of a finite non-empty set of nodes and a set of edges between nodes, usually expressed as: G = (V, E), where G represents the graph, V represents the set of nodes in G, and E represents the set of edges between nodes in G.

[0097] In this step, geopandas and networkx (Python library) can be used to read and construct the GIS data of the ore body contour line, chute, and tunnel into a graph data structure.

[0098] Among them, node data and edge data specifically include:

[0099] Node data:

[0100] Includes non-roadway node data and roadway node data. Correspondingly, non-roadway nodes include ore body contour lines and chutes, and roadway nodes are roadway centerlines.

[0101] Type: Identifies whether the node is a tunnel centerline, a chute, or an ore body outline;

[0102] Geometry coordinates: the coordinates of a node, which can be a series of coordinates representing a node;

[0103] TunnelType: tunnel along the vein, tunnel through the vein, tunnel outside the vein, tunnel with return air, tunnel with chute, etc.

[0104] orePassType: chute type: hoist shaft, ventilation shaft, cutting shaft, cable shaft, elevator shaft, etc.

[0105] MineType: rock, ore; the ore type can indicate the type of ore body contour line.

[0106] Edge data:

[0107] Edge data includes non-lane edge data and lane edge data. Non-lane edge data represents the spatial relationship between non-lane nodes, while lane edge data represents the spatial relationship between non-lane nodes and lane nodes.

[0108] Specifically, edge data may include: the spatial relationship between the ore body contour line, the chute and the tunnel centerline (a complete tunnel includes the tunnel centerline, the left waistline of the tunnel, and the right waistline of the tunnel. The tunnel waistline is derived from the tunnel centerline based on the tunnel width Width, with a left and right offset of Width / 2). The relationships include intersection, inclusion, etc. (for example, the ore body contour line includes the tunnel, the ore body contour line includes the chute, the ore body contour line intersects with the tunnel, the chute intersects with the tunnel, and other spatial relationships).

[0109] In the embodiment of the present application, the tunnel node is the tunnel centerline. Correspondingly, the tunnel design node in the tunnel design data output by the graph neural network is the tunnel design centerline, which can be understood as the middle of the tunnel designed based on the input non-tunnel data.

[0110] In one embodiment of the present application, for the above step 13, based on the error between the lane design data and the lane data, the network parameters of the graph neural network are adjusted, such as Figure 3 Specifically, it may include:

[0111] Step 31: Calculate the mean square error between the lane design data and the lane data as the value of the loss function of the graph neural network.

[0112] In the embodiment of the present application, the mean square error between the lane design data and the lane data is used as the loss function of the graph neural network, which can be calculated using the following formula:

[0113] MSE = 1 / n*∑(y_pred-y_true) 2 ;

[0114] Where n is the number of times training has been completed, which can also be understood as the number of samples. One set of samples is used for one training. y_pred is the predicted lane design data, y_true is the lane data used for training, and Z(y_pred-y_true) 2 is the sum of the training results of completed training.

[0115] Step 32: When the mean square error is not less than the preset error threshold, adjust the network parameters of the graph neural network and start the next training.

[0116] In an embodiment of the present application, the preset error threshold MSEValue can be flexibly set based on the needs of actual applications. The smaller the value, the more times of training need to be completed, and the better the performance of the resulting graph neural network. For example, it can be 0.1.

[0117] In an embodiment of the present application, the graph neural network can adopt various known and feasible graph neural networks. For example, a multi-layer graph convolution layer can be used as the graph neural network. The more layers there are, the better the performance of the trained graph neural network, but the more computing resources are consumed. The specific settings can be flexibly made based on the actual application. For example, a two-layer graph convolution layer can be used.

[0118] In an embodiment of the present application, after completing the above-mentioned model training, self-test sample data can also be used to predict the graph data structure of the lane centerline based on the above-mentioned trained model. When the deviation between the position of a lane centerline in the predicted result and the position of the lane centerline in the corresponding self-test sample is within a preset range, it can be determined that the prediction result of this lane centerline is correct. When the prediction result exceeds the preset range, the number of training times for the same level of data can be increased and the model can be retrained.

[0119] Based on the mine tunnel design model trained by the above-mentioned mine tunnel design model training method, the embodiment of the present application also provides a mine tunnel design method, such as Figure 4 As shown, including:

[0120] Step 41: Obtain non-roadway graph structure data of the mine to be designed. The non-roadway graph structure data includes: non-roadway node data and non-roadway edge data, wherein the non-roadway edge data represents the spatial relationship between each non-roadway node;

[0121] Step 42: Input the non-laneway graph structure data into the graph neural network to obtain the laneway design data output by the graph neural network. The laneway design data serves as the laneway graph structure data. The laneway design data includes laneway design node data and laneway design edge data. The laneway design edge data represents the spatial relationship between each non-laneway node and each laneway design node.

[0122] By adopting the above-mentioned mine tunnel design method provided in the embodiment of the present application, based on the trained graph neural network, when mine tunnel design is required, the non-tunnel graph structure data of the mine to be designed can be obtained, and the non-tunnel graph structure data can be input into the graph neural network to obtain output tunnel design data, which improves the design efficiency compared with the manual design in the existing technology.

[0123] Moreover, when tunnel design is performed manually, the quality of the design results depends on the experience of the designer. Therefore, there may be a problem that the designed mine tunnel is not reasonable enough. The above-mentioned mine tunnel design method provided in the embodiment of the present application can ensure that the designed mine tunnel meets a certain design rationality. Moreover, the designer can further modify the design results based on the output of the graph neural network to obtain the final mine tunnel design results for practical application, thereby significantly improving the design efficiency while taking into account the design rationality.

[0124] In the embodiment of the present application, the tunnel design data output by the graph neural network is graph structure data. Furthermore, the tunnel design data can be converted to obtain CAD drawing data that can be directly used by designers, which may specifically include:

[0125] Convert the tunnel diagram structure data into tunnel data in GIS exchange format;

[0126] Convert the tunnel data in GIS exchange format into tunnel CAD drawing data.

[0127] Converting graph structure data into GIS exchange format data, and converting GIS exchange format data into CAD drawing data is the same as the above Figure 2 The reverse operation of the data format conversion shown is not described in detail here.

[0128] In one embodiment of the present application, the following steps may be used to generate non-lane graph structure data:

[0129] Acquire non-tunnel CAD drawing data of the mine, where the non-tunnel CAD drawing data includes graphic information of the non-tunnel of the mine;

[0130] Convert non-laneway CAD drawing data into non-laneway data in GIS exchange format;

[0131] Generate non-lane map structure data based on non-lane data in GIS exchange format.

[0132] In the above step of generating non-lane graph structure data, converting CAD drawing data into data in GIS exchange format and generating graph structure data based on the data in GIS exchange format are the same as the above step. Figure 2 The operations similar to the data format conversion shown are not described in detail here.

[0133] In the mine tunnel design method provided in the embodiment of the present application, non-tunnel nodes may include: ore body contour lines and chute;

[0134] The lane design node is the lane design centerline, that is, the designed lane centerline. On this basis, the designer can continue to add other attribute information of the lane, such as lane type, lane width and other attribute information.

[0135] The above-mentioned mine tunnel design model training method and mine tunnel design method provided in the embodiment of the present application can realize the automated design of the tunnel centerline based on the data other than the tunnel in the known CAD drawing data, such as the ore body contour line and the chute, thereby improving the design efficiency of the mine tunnel excavation design plane drawings. Mine designers can draw the tunnel left waistline, tunnel right waistline and other annotation information based on the automatically designed tunnel centerline drawing data, and only need to draw the tunnel left waistline, tunnel right waistline and other annotation information according to the tunnel width and left and right offsets, without having to consider the collaborative design of geological exploration data, engineering measurement data, etc., to a certain extent, improving the efficiency of the mine tunnel excavation design plane drawings.

[0136] In addition, it provides possibilities for the use of AI algorithms in the intelligent design of mine tunnels, accelerating the intelligent upgrade of mines.

[0137] Based on the same inventive concept, according to the mine tunnel design model training method provided in the above embodiment of the present application, another embodiment of the present application also provides a mine tunnel design model training device, the structural diagram of which is shown in FIG. Figure 5 As shown, specifically including:

[0138] A training data acquisition module 51 is configured to acquire graph structure data of the mine for model training, wherein the graph structure data includes non-roadway data and roadway data. The non-roadway data includes non-roadway node data and non-roadway edge data. The roadway data includes roadway node data and roadway edge data. The non-roadway edge data represents the spatial relationship between non-roadway nodes, and the roadway edge data represents the spatial relationship between non-roadway nodes and roadway nodes.

[0139] A training data input module 52 is configured to input the non-laneway data into a graph neural network to be trained, and obtain laneway design data output by the graph neural network, wherein the laneway design data includes laneway design node data and laneway design edge data, wherein the laneway design edge data represents the spatial relationship between each non-laneway node and each laneway design node;

[0140] The model training module 53 is used to adjust the network parameters of the graph neural network based on the error between the lane design data and the lane data.

[0141] Furthermore, the training data acquisition module 51 is further configured to generate the graph structure data for model training by adopting the following steps:

[0142] Acquire CAD drawing data of a mine, wherein the CAD drawing data includes graphic information of the mine; convert the CAD drawing data into mine data in a GIS exchange format, wherein the mine data in the GIS exchange format includes non-roadway mine data and roadway mine data; and generate graph structure data for model training based on the mine data in the GIS exchange format.

[0143] Furthermore, the model training module 53 is specifically used to calculate the mean square error between the tunnel design data and the tunnel data as the value of the loss function of the graph neural network; when the mean square error is not less than a preset error threshold, the network parameters of the graph neural network are adjusted.

[0144] Furthermore, the non-roadway nodes include: ore body contour lines and chute;

[0145] The lane node is the lane centerline;

[0146] The tunnel design node is the tunnel design centerline.

[0147] Based on the same inventive concept, according to the mine tunnel design method provided in the above embodiment of the present application, another embodiment of the present application also provides a mine tunnel design device, the structural diagram of which is shown in FIG. Figure 6 As shown, specifically including:

[0148] The non-roadway data acquisition module 61 is used to acquire non-roadway graph structure data of the mine to be designed, wherein the non-roadway graph structure data includes: non-roadway node data and non-roadway edge data, wherein the non-roadway edge data represents the spatial relationship between each non-roadway node;

[0149] The lane design module 62 is used to input the non-lane graph structure data into the graph neural network to obtain the lane design data output by the graph neural network. The lane design data serves as the lane graph structure data. The lane design data includes lane design node data and lane design edge data. The lane design edge data represents the spatial relationship between each non-lane node and each lane design node.

[0150] Furthermore, the lane design module 62 is further configured to convert the lane diagram structure data into lane data in a GIS exchange format; and convert the lane data in a GIS exchange format into CAD drawing data of the lane.

[0151] Furthermore, the non-lane data acquisition module 61 is further configured to generate the non-lane graph structure data by using the following steps:

[0152] Acquire non-tunnel CAD drawing data of a mine, wherein the non-tunnel CAD drawing data includes graphic information of the non-tunnel of the mine; convert the non-tunnel CAD drawing data into non-tunnel data in a GIS exchange format; and generate non-tunnel graph structure data based on the non-tunnel data in the GIS exchange format.

[0153] Furthermore, the non-roadway nodes include: ore body contour lines and chute;

[0154] The tunnel design node is the tunnel design centerline.

[0155] The functions of the above modules can correspond to Figures 1 to 4 The corresponding processing steps in the shown process will not be repeated here.

[0156] The mine tunnel design model training device and mine tunnel design device provided in the embodiments of the present application can be implemented via a computer program. Those skilled in the art will appreciate that the aforementioned module division is only one of many module division methods, and that other module divisions or no module divisions, as long as the mine tunnel design model training device and mine tunnel design device possess the aforementioned functions, are within the scope of protection of the present application.

[0157] The present application also provides an electronic device, such as Figure 7 As shown, it includes a processor 71 and a machine-readable storage medium 72, and the machine-readable storage medium 72 stores machine-executable instructions that can be executed by the processor 71. The processor 71 is prompted by the machine-executable instructions to: implement any of the above-mentioned mine tunnel design model training methods, or implement any of the above-mentioned mine tunnel design methods.

[0158] An embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements any of the above-mentioned mine tunnel design model training methods, or implements any of the above-mentioned mine tunnel design methods.

[0159] An embodiment of the present application further provides a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute any of the above-mentioned mine tunnel design model training methods, or to execute any of the above-mentioned mine tunnel design methods.

[0160] The machine-readable storage medium in the electronic device may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage. Optionally, the memory may be at least one storage device located away from the processor.

[0161] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0162] Each embodiment in this specification is described in a related manner. Similar portions between the various embodiments can be referenced to each other. Each embodiment focuses on the differences between the other embodiments. In particular, the device, electronic device, computer-readable storage medium, and computer program product embodiments are generally similar to the method embodiments, so their descriptions are relatively simple. For related portions, reference can be made to the descriptions of the method embodiments.

[0163] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or equipment. In the absence of further restrictions, the elements defined by the sentence "comprise a..." do not exclude the presence of other identical elements in the process, method, article or equipment including the elements.

[0164] 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 produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1A device that provides the functions specified in a block or multiple blocks.

[0165] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work 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 The function specified in one or more boxes.

[0166] 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 the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0167] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A mine tunnel design model training method, characterized in that: include: Obtaining graph structure data of the mine for model training, the graph structure data comprising: non-roadway data and roadway data, the non-roadway data comprising non-roadway node data and non-roadway edge data, the roadway data comprising roadway node data and roadway edge data, wherein the non-roadway edge data represents the spatial relationship between each non-roadway node, and the roadway edge data represents the spatial relationship between each non-roadway node and each roadway node; Inputting the non-laneway data into a graph neural network to be trained to obtain laneway design data output by the graph neural network, wherein the laneway design data includes laneway design node data and laneway design edge data, wherein the laneway design edge data represents the spatial relationship between each non-laneway node and each laneway design node; Based on the error between the lane design data and the lane data, the network parameters of the graph neural network are adjusted.

2. The method according to claim 1, wherein The graph structure data for model training is generated using the following steps: Acquire CAD drawing data of a mine, wherein the CAD drawing data includes graphic information of the mine; Converting the CAD drawing data into mine data in a GIS exchange format, wherein the mine data in the GIS exchange format includes non-roadway mine data and roadway mine data; Based on the mine data in the GIS exchange format, graph structure data for model training is generated.

3. The method according to claim 1, wherein The adjusting the network parameters of the graph neural network based on the error between the lane design data and the lane data includes: Calculating a mean square error between the lane design data and the lane data as a value of a loss function of the graph neural network; When the mean square error is not less than a preset error threshold, the network parameters of the graph neural network are adjusted.

4. The method according to any one of claims 1 to 3, wherein The non-roadway nodes include: ore body contour lines and chute; The lane node is the lane centerline; The tunnel design node is the tunnel design centerline.

5. A mine tunnel design method, characterized in that: include: Acquire non-roadway graph structure data of a mine to be designed, wherein the non-roadway graph structure data includes: non-roadway node data and non-roadway edge data, wherein the non-roadway edge data represents the spatial relationship between each non-roadway node; The non-laneway graph structure data is input into a graph neural network to obtain laneway design data output by the graph neural network. The laneway design data serves as laneway graph structure data. The laneway design data includes laneway design node data and laneway design edge data. The laneway design edge data represents the spatial relationship between each non-laneway node and each laneway design node.

6. The method according to claim 5, wherein Also includes: Converting the lane map structure data into lane data in a GIS exchange format; The tunnel data in the GIS exchange format is converted into CAD drawing data of the tunnel.

7. The method according to claim 5, wherein The non-lane graph structure data is generated using the following steps: Acquire non-tunnel CAD drawing data of a mine, where the non-tunnel CAD drawing data includes graphic information of the non-tunnel of the mine; Convert non-laneway CAD drawing data into non-laneway data in GIS exchange format; Based on the non-lane data in the GIS exchange format, non-lane graph structure data is generated.

8. The method according to any one of claims 5 to 7, wherein: The non-roadway nodes include: ore body contour lines and chute; The tunnel design node is the tunnel design centerline.

9. A mine tunnel design model training device, characterized in that: include: a training data acquisition module, configured to acquire graph structure data of the mine for model training, wherein the graph structure data includes: non-roadway data and roadway data, wherein the non-roadway data includes non-roadway node data and non-roadway edge data, and the roadway data includes roadway node data and roadway edge data, wherein the non-roadway edge data represents the spatial relationship between each non-roadway node, and the roadway edge data represents the spatial relationship between each non-roadway node and each roadway node; a training data input module, configured to input the non-laneway data into a graph neural network to be trained, and obtain laneway design data output by the graph neural network, wherein the laneway design data includes laneway design node data and laneway design edge data, wherein the laneway design edge data represents the spatial relationship between each non-laneway node and each laneway design node; A model training module is used to adjust the network parameters of the graph neural network based on the error between the lane design data and the lane data.

10. A mine tunnel design device, characterized in that: include: A non-roadway data acquisition module is used to acquire non-roadway graph structure data of the mine to be designed, wherein the non-roadway graph structure data includes: non-roadway node data and non-roadway edge data, wherein the non-roadway edge data represents the spatial relationship between each non-roadway node; A lane design module is used to input the non-lane graph structure data into a graph neural network to obtain lane design data output by the graph neural network. The lane design data serves as lane graph structure data. The lane design data includes lane design node data and lane design edge data. The lane design edge data represents the spatial relationship between each non-lane node and each lane design node.

11. An electronic device, characterized in that: The method comprises a processor and a machine-readable storage medium, wherein the machine-readable storage medium stores machine-executable instructions that can be executed by the processor, and the processor is prompted by the machine-executable instructions to implement the method described in any one of claims 1 to 4, or to implement the method described in any one of claims 5 to 8.

12. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 4 is implemented, or the method according to any one of claims 5 to 8 is implemented.

Images