A training method and system for artificial intelligence models for smart mines

By dividing smart mines into multiple mining areas and determining model training groups based on security identification similarity and risk, the problem of insufficient data in traditional mine safety management is solved, and the training accuracy and reliability of artificial intelligence models are improved.

CN118469284BActive Publication Date: 2024-11-26HANGZHOU FROG CLOUD NETWORK TECH CO LTD
View PDF 19 Cites 0 Cited by

Patent Information

Application Number
CN202410565948.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-05-09
Publication Date
2024-11-26
Estimated Expiration
2044-05-09

AI Technical Summary

Technical Problem

Traditional mine safety management relies on manual labor and cannot identify security risks in a timely manner, resulting in insufficient training data for artificial intelligence models and difficult to meet the requirements for identification accuracy.

Method used

The smart mine is divided into multiple mining areas, and similar areas are determined through mining data and geological condition data, a model training group is formed, and the training data volume is determined based on security risks, and the model training data volume is increased through data expansion processing.

Benefits of technology

The similarity assessment of the mining area is achieved, ensuring that the training data volume of the model training group meets the requirements, and improving the training reliability of the artificial intelligence model.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118469284B_ABST
    Figure CN118469284B_ABST
Patent Text Reader

Abstract

The present invention provides a training method and system for an artificial intelligence model for a smart mine, which belongs to the field of artificial intelligence technology, and specifically includes: performing safety identification similarity of different mining areas and determining similar mining areas through mining data and geological condition data of different dimensions, dividing similar mining areas into the same model training group, obtaining the safety identification similarity of mining areas of different model training groups and the number of mining areas, and determining the data amount of training data of safety hazard video frames of different model training groups in combination with the safety risks of different model training groups, taking the data amount of training data of safety hazard video frames as the target, performing expansion processing on the original training data to obtain expanded original data, and using the expanded original data to perform training processing on the artificial intelligence model, thereby ensuring the reliability of the training processing of the artificial intelligence model.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of artificial intelligence technology, and in particular relates to a training method and system for an artificial intelligence model for a smart mine. Background Art

[0002] Traditional mine safety management often relies on manual methods. Since it is impossible to detect safety risks in a timely and effective manner, the safety of personnel and equipment in mines cannot be effectively guaranteed. This makes how to automatically identify and handle safety hazards in mines a technical problem that needs to be solved urgently.

[0003] In order to solve the above technical problems, the existing technical solutions often use AI recognition to automatically analyze and process the video frames of the mine, and identify the safety risks of the mine according to the analysis results, so as to ensure that the safety hazards of the mine can be controlled at the first time. However, it is not difficult to find that there are the following technical problems through analysis:

[0004] When training artificial intelligence models, the amount of original training data in which the equipment or mining status of the mine is in a dangerous state or a state of potential danger is very small compared to the original data in a normal state. Therefore, it is impossible to accurately and effectively implement the training of the artificial intelligence model, making it difficult for the recognition accuracy of the artificial intelligence model to meet the requirements.

[0005] In response to the above technical problems, the present invention provides a training method and system for an artificial intelligence model for a smart mine. Summary of the invention

[0006] To achieve the purpose of the present invention, the present invention adopts the following technical solutions:

[0007] According to one aspect of the present invention, a method for training an artificial intelligence model for a smart mine is provided.

[0008] A training method for an artificial intelligence model for a smart mine, characterized by comprising:

[0009] S1 divides the smart mine into multiple mining areas, and uses mining data and geological condition data of different dimensions to identify the safety similarity of different mining areas and determine similar mining areas, and divide similar mining areas into the same model training group;

[0010] S2 determines the risks of mining personnel in different mining areas according to the mining data, and determines the safety risks of different model training groups based on the risks of mining personnel in different mining areas. When the safety risks of the model training groups meet the requirements, proceed to the next step;

[0011] S3 obtains the safety identification similarity of mining areas of different model training groups and the number of mining areas, and determines the data volume of training data of safety hazard video frames of different model training groups in combination with the safety risks of different model training groups;

[0012] S4 takes the data volume of the training data of the safety hazard video frame as a target, performs expansion processing on the original training data to obtain expanded original data, and uses the expanded original data to train the artificial intelligence model.

[0013] The beneficial effects of the present invention are:

[0014] 1. Similar mining areas are determined through mining data and geological condition data of different dimensions, and similar mining areas are divided into the same model training group, thereby realizing the evaluation of the similarity of mining areas from the two perspectives of mining data and geological condition data. By dividing similar mining areas into the same model training group, the technical problem of using the same artificial intelligence model for different similar mining areas, which causes the model training difficulty to be difficult to meet the requirements, is avoided.

[0015] 2. The amount of training data for safety hazard video frames of different model training groups is determined based on the safety identification similarity of mining areas, the number of mining areas, and the safety risks of model training groups. This not only takes into account the differences in the reliability of training requirements due to differences in safety identification similarity and number of mining areas of different model training groups, but also ensures the data amount requirements for training data for model training groups with higher safety risks by further combining them with the safety risks of model training groups, thereby ensuring the reliability of the training processing of the artificial intelligence model.

[0016] A further technical solution is to divide the smart mine into multiple mining areas, including:

[0017] The distribution of mining areas of the smart mine is obtained, and based on the distribution, the mining areas within a preset area are divided into the same mining area.

[0018] A further technical solution is that the mining data includes the number of mining machines and the mining volumes of different mining machines.

[0019] A further technical solution is that the geological condition data includes a topographic map, a geological map, an ore body distribution map, a stratigraphic sequence and distribution data of different types of minerals in different strata of the mining area.

[0020] A further technical solution is that the method for determining the risk of mining personnel in the mining area is:

[0021] The number of mining personnel in the mining area is determined based on the mining data of the mining area, and the mining personnel risk in the mining area is determined according to the number of mining personnel in the mining area.

[0022] A further technical solution is that the value range of the mining personnel risk in the mining area is between 0 and 1, wherein the higher the mining personnel risk in the mining area, the higher the personnel risk when an accident occurs in the mining area.

[0023] A further technical solution is that the method for determining the security risk of the model training group is:

[0024] Determine the weight values ​​of different mining areas based on the risks of mining personnel in different mining areas;

[0025] The weight sum of the model training group is determined according to the weight values ​​of different mining areas of the model training group, and the safety risk of the model training group is determined based on the weight sum.

[0026] A further technical solution is that the value range of the security risk of the model training group is between 0 and 1, wherein when the security risk of the model training group is within a preset range, it is determined that the security risk of the model training group meets the requirements.

[0027] A further technical solution is that, when the safety risk of the model training group does not meet the requirements, the data amount of the training data of the safety hazard video frames of the model training group is determined based on the preset data amount of the model training group.

[0028] A further technical solution is that the method for determining the amount of training data of the potential safety hazard video frames of the model training group is as follows:

[0029] Determine a baseline data amount of training data of safety hazard video frames of the model training group based on the safety risk of the model training group;

[0030] Determining the training reliability requirement of the model training group by the safety identification similarities of different mining areas of the model training group and the number of mining areas;

[0031] Based on the training reliability requirement of the model training group, the benchmark data volume is corrected to obtain the data volume of the training data of the safety hazard video frames of the model training group.

[0032] A further technical solution is that the expansion method of the original data includes data enhancement, data generation, transfer learning and data expansion.

[0033] In a second aspect, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, characterized in that: when the processor runs the computer program, it executes the above-mentioned training method for an artificial intelligence model for a smart mine.

[0034] Other features and advantages will be described in the following description. The objects and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description and drawings.

[0035] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The above and other features and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings.

[0037] Figure 1 is a flow chart of a training method for an artificial intelligence model for smart mines;

[0038] Figure 2 is a flow chart of a method for determining a safety identification similarity of a mining area;

[0039] Figure 3 It is a flowchart of a method for determining security risks of a model training group. DETAILED DESCRIPTION

[0040] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0041] The following will further elaborate from two perspectives: method-based embodiments and system-based embodiments.

[0042] Method Example

[0043] To solve the above problems, according to one aspect of the present invention, Figure 1 As shown, a training method for an artificial intelligence model for a smart mine is provided, which is characterized by specifically including:

[0044] S1 divides the smart mine into multiple mining areas, and uses mining data and geological condition data of different dimensions to identify the safety similarity of different mining areas and determine similar mining areas, and divide similar mining areas into the same model training group;

[0045] Furthermore, the smart mine is divided into multiple mining areas, including:

[0046] The distribution of mining areas of the smart mine is obtained, and based on the distribution, the mining areas within a preset area are divided into the same mining area.

[0047] Specifically, the mining data includes the number of mining machines and the mining volume of different mining machines.

[0048] Furthermore, the geological condition data include topographic maps, geological maps, ore body distribution maps, stratigraphic sequences of the mining area, and distribution data of different types of minerals in different strata.

[0049] Specifically, Figure 2 As shown, the method for determining the safety identification similarity of the mining area is:

[0050] Determine the number of mining machines in the mining area and the mining volume of different mining machines based on the mining data, and determine the similarity of mining data between the mining area and other mining areas in combination with the mining distribution of different mining machines in the mining area;

[0051] Determine the similarity of geological conditions between the mining area and other mining areas through geological condition data of different dimensions of the mining area;

[0052] The safety identification similarity between the mining area and other mining areas is determined based on the mining data similarity and the geological condition similarity.

[0053] Further, when the security identification similarity between the mining area and other mining areas meets the requirement, the mining area and other mining areas are determined to be similar mining areas.

[0054] In another embodiment, the method for determining the safety identification similarity of the mining area is:

[0055] Determine the number of mining machines in the mining area based on the mining data, and determine whether the mining area is similar to other mining areas based on the number of mining machines, if so, proceed to the next step, if not, determine that the mining area and other mining areas are not similar mining areas;

[0056] Acquire the mining volume of different mining machines in the mining area, and determine whether the mining area is similar to other mining areas based on the mining volume, if so, proceed to the next step, if not, determine that the mining area and other mining areas are not similar mining areas;

[0057] Determine the similarity of the geological conditions of the mining area in different dimensions with other mining areas through the geological condition data of different dimensions of the mining area, and judge whether there is a dimension of geological conditions whose similarity meets the requirements. If so, proceed to the next step; if not, determine that the mining area and other mining areas do not belong to similar mining areas;

[0058] Determine the similarity of mining data between the mining area and other mining areas through the number of mining machines, the mining volume of different mining machines, and the mining distribution of different mining machines in the mining area; determine the similarity of geological conditions between the mining area and other mining areas through the similarity of geological conditions between the mining area and other mining areas in different dimensions;

[0059] The safety identification similarity between the mining area and other mining areas is determined based on the mining data similarity and the geological condition similarity.

[0060] In another embodiment, the method for determining the safety identification similarity of the mining area is:

[0061] Determine the number of mining machines in the mining area based on the mining data, and determine the similarity of mining data between the mining area and other mining areas through the number of mining machines, the mining volume of different mining machines, and the mining distribution of different mining machines in the mining area;

[0062] Determine whether the similarity of the mining data meets the requirement, if yes, proceed to the next step, if no, determine that the mining area and other mining areas do not belong to similar mining areas;

[0063] Determine the similarity of the geological conditions of the mining area in different dimensions with other mining areas through the geological condition data of different dimensions of the mining area, and judge whether there is a dimension of geological conditions whose similarity meets the requirements. If so, proceed to the next step; if not, determine that the mining area and other mining areas do not belong to similar mining areas;

[0064] The dimensions of the geological conditions that meet the similarity requirements are taken as similar dimensions, and it is determined whether the number of the similar dimensions meets the requirements. If so, the next step is entered; if not, it is determined that the mining area and other mining areas do not belong to similar mining areas;

[0065] Determine the similarity of geological conditions between the mining area and other mining areas by the similarity of geological conditions between the mining area and other mining areas in different dimensions, and judge whether the similarity of geological conditions meets the requirements. If yes, proceed to the next step; if not, determine that the mining area and other mining areas do not belong to similar mining areas;

[0066] The safety identification similarity between the mining area and other mining areas is determined based on the mining data similarity and the geological condition similarity.

[0067] S2 determines the risks of mining personnel in different mining areas according to the mining data, and determines the safety risks of different model training groups based on the risks of mining personnel in different mining areas. When the safety risks of the model training groups meet the requirements, proceed to the next step;

[0068] Furthermore, the method for determining the risk of mining personnel in the mining area is:

[0069] The number of mining personnel in the mining area is determined based on the mining data of the mining area, and the mining personnel risk in the mining area is determined according to the number of mining personnel in the mining area.

[0070] Specifically, the value range of the mining personnel risk in the mining area is between 0 and 1, wherein the higher the mining personnel risk in the mining area, the higher the personnel risk when an accident occurs in the mining area.

[0071] It should be noted that the method for determining the security risk of the model training group is:

[0072] Determine the weight values ​​of different mining areas based on the risks of mining personnel in different mining areas;

[0073] The weight sum of the model training group is determined according to the weight values ​​of different mining areas of the model training group, and the safety risk of the model training group is determined based on the weight sum.

[0074] Furthermore, the value range of the security risk of the model training group is between 0 and 1, wherein when the security risk of the model training group is within a preset range, it is determined that the security risk of the model training group meets the requirements.

[0075] It can be understood that when the safety risk of the model training group does not meet the requirements, the data amount of the training data of the safety hazard video frames of the model training group is determined based on the preset data amount of the model training group.

[0076] Specifically, Figure 3As shown, the method for determining the security risk of the model training group is:

[0077] Based on the risks of miners in different mining areas, the mining areas are divided into risky mining areas and general mining areas;

[0078] Determine the comprehensive safety risk of the risky mining area by the number of the risky mining areas and the risks of the mining personnel in different risky mining areas, and determine the comprehensive safety risk of the general mining area based on the number of the general mining areas and the risks of the mining personnel in different general mining areas;

[0079] The safety risk of the model training group is determined according to the comprehensive safety risk of the risky mining area and the comprehensive safety risk of the general mining area.

[0080] In another embodiment, the method for determining the security risk of the model training group is:

[0081] Determine the average value of the risks of miners in different mining areas of the model training group based on the risks of miners in different mining areas, and judge whether the average value of the risks of miners in the mining areas is greater than a preset risk threshold. If so, determine that the safety risk of the model training group does not meet the requirements. If not, proceed to the next step.

[0082] The mining area is divided into risk mining areas and general mining areas, and it is determined whether the number of risk mining areas of the model training group meets the requirements. If so, the next step is entered; if not, it is determined that the safety risk of the model training group does not meet the requirements;

[0083] Determine the comprehensive safety risk of the risky mining area by the number of risky mining areas and the risk of mining personnel in different risky mining areas, and judge whether the comprehensive safety risk of the risky mining area meets the requirements. If so, proceed to the next step; if not, determine that the safety risk of the model training group does not meet the requirements;

[0084] The comprehensive safety risk of the general mining areas is determined based on the number of general mining areas and the risks of mining personnel in different general mining areas, and the safety risk of the model training group is determined according to the comprehensive safety risk of the risky mining areas and the comprehensive safety risk of the general mining areas.

[0085] In another embodiment, the method for determining the security risk of the model training group is:

[0086] Determine the average value of the risks of miners in different mining areas of the model training group based on the risks of miners in different mining areas, and judge whether the average value of the risks of miners in the mining areas is greater than a preset risk threshold. If so, determine that the safety risk of the model training group does not meet the requirements. If not, proceed to the next step.

[0087] Determine the weight values ​​of different mining areas based on the risks of mining personnel in different mining areas, determine the weight sum of the model training group according to the weight values ​​of different mining areas of the model training group, and determine whether the safety risk of the model training group meets the requirements based on the weight sum, if not, determine that the safety risk of the model training group does not meet the requirements, if yes, proceed to the next step;

[0088] The mining area is divided into a risk mining area and a general mining area, and it is determined whether the weight of the risk mining area meets the requirements. If so, the next step is entered; if not, it is determined that the safety risk of the model training group does not meet the requirements;

[0089] Determine the comprehensive safety risk of the risky mining area by the number of the risky mining areas and the risks of the mining personnel in different risky mining areas, and determine the comprehensive safety risk of the general mining area based on the number of the general mining areas and the risks of the mining personnel in different general mining areas;

[0090] The safety risk of the model training group is determined according to the comprehensive safety risk of the risky mining area and the comprehensive safety risk of the general mining area.

[0091] S3 obtains the safety identification similarity of mining areas of different model training groups and the number of mining areas, and determines the data volume of training data of safety hazard video frames of different model training groups in combination with the safety risks of different model training groups;

[0092] Specifically, the method for determining the data amount of the training data of the safety hazard video frames of the model training group is:

[0093] Determine a baseline data amount of training data of safety hazard video frames of the model training group based on the safety risk of the model training group;

[0094] Determining the training reliability requirement of the model training group by the safety identification similarities of different mining areas of the model training group and the number of mining areas;

[0095] Based on the training reliability requirement of the model training group, the benchmark data volume is corrected to obtain the data volume of the training data of the safety hazard video frames of the model training group.

[0096] In another embodiment, the method for determining the amount of training data of the potential safety hazard video frames of the model training group is:

[0097] S31 obtains the number of mining areas of the model training group, and determines whether the number of mining areas of the model training group is greater than the preset number of areas, if not, proceeds to the next step, if yes, determines the data amount of the training data of the safety hazard video frames of the model training group according to the preset data amount;

[0098] S32 determines whether the number of mining areas in the model training group is greater than the second preset number of areas, if so, proceeds to the next step, if not, proceeds to step S34;

[0099] S33 determines whether the average value of the safety identification similarity of different mining areas of the model training group is less than a preset similarity threshold, and if so, determines the data amount of the training data of the safety hazard video frames of the model training group according to the preset data amount, and if not, proceeds to the next step;

[0100] S34 determines the training reliability requirement of the model training group by the safety identification similarities of different mining areas of the model training group and the number of mining areas, and judges whether the training reliability requirement of the model training group is greater than a preset reliability threshold. If so, the data amount of the training data of the safety hazard video frame of the model training group is determined according to the preset data amount. If not, proceed to the next step.

[0101] S35 determines a baseline data volume of training data of safety hazard video frames of the model training group based on the safety risk of the model training group, and corrects the baseline data volume based on the training reliability requirement of the model training group to obtain the data volume of training data of safety hazard video frames of the model training group.

[0102] S4 takes the data volume of the training data of the safety hazard video frame as a target, performs expansion processing on the original training data to obtain expanded original data, and uses the expanded original data to train the artificial intelligence model.

[0103] Furthermore, the expansion methods of expanding the original data include data enhancement, data generation, transfer learning and data expansion.

[0104] System Example

[0105] In a second aspect, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, characterized in that: when the processor runs the computer program, it executes the above-mentioned training method for an artificial intelligence model for a smart mine.

[0106] Through the above embodiments, the present invention achieves the following beneficial effects:

[0107] 1. Similar mining areas are determined through mining data and geological condition data of different dimensions, and similar mining areas are divided into the same model training group, thereby realizing the evaluation of the similarity of mining areas from the two perspectives of mining data and geological condition data. By dividing similar mining areas into the same model training group, the technical problem of using the same artificial intelligence model for different similar mining areas, which causes the model training difficulty to be difficult to meet the requirements, is avoided.

[0108] 2. The amount of training data for safety hazard video frames of different model training groups is determined based on the safety identification similarity of mining areas, the number of mining areas, and the safety risks of model training groups. This not only takes into account the differences in the reliability of training requirements due to differences in safety identification similarity and number of mining areas of different model training groups, but also ensures the data amount requirements for training data for model training groups with higher safety risks by further combining them with the safety risks of model training groups, thereby ensuring the reliability of the training processing of the artificial intelligence model.

[0109] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0110] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0111] The above description is only one or more embodiments of this specification and is not intended to limit this specification. For those skilled in the art, one or more embodiments of this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of this specification shall be included in the scope of the claims of this specification.

Claims

1. A training method for an artificial intelligence model for smart mines, It is characterized in that Specifically include: The smart mine is divided into multiple mining areas, and the similarity of safety identification of different mining areas and similar mining areas are determined through mining data and geological condition data of different dimensions, and similar mining areas are divided into the same model training group; Determine the risks of mining personnel in different mining areas according to the mining data, and determine the safety risks of different model training groups based on the risks of mining personnel in different mining areas. When the safety risks of the model training groups meet the requirements, proceed to the next step; Obtain the safety identification similarity of mining areas and the number of mining areas of different model training groups, and determine the data volume of training data of safety hazard video frames of different model training groups in combination with the safety risks of different model training groups; Taking the data volume of the training data of the safety hazard video frame as a target, the original training data is expanded to obtain expanded original data, and the artificial intelligence model is trained using the expanded original data; The method for determining the data amount of the training data of the potential safety hazard video frames of the model training group is as follows: S31 obtains the number of mining areas of the model training group, and determines whether the number of mining areas of the model training group is greater than the preset number of areas, if not, proceeds to the next step, if yes, determines the data amount of the training data of the safety hazard video frames of the model training group according to the preset data amount; S32 determines whether the number of mining areas in the model training group is greater than the second preset number of areas, if so, proceeds to the next step, if not, proceeds to step S34; S33 determines whether the average value of the safety identification similarity of different mining areas of the model training group is less than a preset similarity threshold, and if so, determines the data amount of the training data of the safety hazard video frames of the model training group according to the preset data amount, and if not, proceeds to the next step; S34 determines the training reliability requirement of the model training group by the safety identification similarities of different mining areas of the model training group and the number of mining areas, and judges whether the training reliability requirement of the model training group is greater than a preset reliability threshold. If so, the data amount of the training data of the safety hazard video frame of the model training group is determined according to the preset data amount. If not, proceed to the next step. S35: determining a baseline data volume of training data of safety hazard video frames of the model training group based on the safety risk of the model training group, and correcting the baseline data volume based on the training reliability requirement of the model training group to obtain the data volume of training data of safety hazard video frames of the model training group; The value range of the security risk of the model training group is between 0 and 1. When the security risk of the model training group is within a preset range, it is determined that the security risk of the model training group meets the requirements. When the security risk of the model training group does not meet the requirements, the data amount of the training data of the security hazard video frames of the model training group is determined based on the preset data amount of the model training group.

2. The training method of the artificial intelligence model for smart mines according to claim 1, It is characterized in that The smart mine is divided into multiple mining areas, including: The distribution of mining areas of the smart mine is obtained, and based on the distribution, the mining areas within a preset area are divided into the same mining area.

3. The training method of the artificial intelligence model for smart mines according to claim 1, It is characterized in that The mining data includes the number of mining machines and the mining volumes of different mining machines.

4. The training method of the artificial intelligence model for smart mines according to claim 1, It is characterized in that The geological condition data include topographic maps, geological maps, ore body distribution maps, stratigraphic sequences and distribution data of different types of minerals in different strata of the mining area.

5. The training method of the artificial intelligence model for smart mines according to claim 1, It is characterized in that The method for determining the safety identification similarity of the mining area is: Determine the number of mining machines in the mining area and the mining volume of different mining machines based on the mining data, and determine the similarity of mining data between the mining area and other mining areas in combination with the mining distribution of different mining machines in the mining area; Determine the similarity of geological conditions between the mining area and other mining areas through geological condition data of different dimensions of the mining area; The safety identification similarity between the mining area and other mining areas is determined based on the mining data similarity and the geological condition similarity.

6. The training method of the artificial intelligence model for smart mines according to claim 5, It is characterized in that When the security identification similarity between the mining area and other mining areas meets the requirement, the mining area and other mining areas are determined to be similar mining areas.

7. The training method of the artificial intelligence model for smart mines according to claim 1, It is characterized in that The method for determining the risk of mining personnel in the mining area is: The number of mining personnel in the mining area is determined based on the mining data of the mining area, and the mining personnel risk in the mining area is determined according to the number of mining personnel in the mining area.

8. The training method of the artificial intelligence model for smart mines according to claim 1, It is characterized in that The value range of the mining personnel risk in the mining area is between 0 and 1, wherein the higher the mining personnel risk in the mining area, the higher the personnel risk when an accident occurs in the mining area.

9. The training method of the artificial intelligence model for smart mines according to claim 1, It is characterized in that The method for determining the security risk of the model training group is: Determine the weight values ​​of different mining areas based on the risks of mining personnel in different mining areas; The weight sum of the model training group is determined according to the weight values ​​of different mining areas of the model training group, and the safety risk of the model training group is determined based on the weight sum.

10. A computer system, include: A memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, characterized in that the processor executes a training method for an artificial intelligence model for a smart mine as described in any one of claims 1 to 9 when running the computer program.

Citation Information

Patent Citations

  • Model training method, data similarity determining method, apparatuses and devices

    CN107609461A

  • Risk model training method and apparatus, risk identification method and apparatus, equipment, and medium

    CN108520343A

  • Oil and gas exploitation monitoring method, life determining model obtaining method and related devices

    CN110952973A

  • Data set equalization training method based on artificial intelligence

    CN111461262A

  • Financial risk monitoring method based on artificial intelligence and related equipment

    CN112950383A