Mine integrated intelligent operation management method and system based on BIM + GIS + IoT

Abstract

The invention belongs to the technical field of mine digitization, and particularly discloses a mine integrated intelligent operation management method and system based on BIM + GIS + IoT. According to the method, BIM is adopted to construct a mine system real scene model; a sensor and detection equipment are used for real-time monitoring and statistical analysis of data; the production operation process is managed through order driving based on an ERP system; processing and analyzing the multi-source data by using a Softmax regression model and a cloud model, performing risk assessment and outputting feedback; statistical analysis and production and operation data and a risk assessment result are integrated with the real scene model, total-factor display and reflection of uploaded information are carried out, and a decision is given; data is taken as a medium, and BIM, GIS, IoT and an artificial intelligence algorithm are taken as implementation means. From the global perspective of supply and demand, production, sales and operation, a mine-transportation-factory integrated intelligent operation platform is constructed, and a safe and efficient mineral resource full-life-cycle management mode can be created.

Description

Technical Field

[0001] The present invention belongs to the field of digital production and operation of mines, and specifically relates to an integrated intelligent operation and management method and system for mines based on BIM+GIS+IoT. Background Art

[0002] The mining process involves many links such as mining, transportation, and operation due to the large number of mechanical equipment, complex environment, and large area. The traditional manual recording method makes it difficult to achieve real-time supervision of machinery, which easily leads to waste of resources and safety accidents. In addition, the information transmission is not smooth, resulting in the inability to quickly issue dispatch instructions. Against the background of the rapid development of new intelligent technologies and Industry 4.0, the requirements for the digital transformation of mines are becoming increasingly higher, especially the integrated operation and management of multiple systems. However, the existing information-based mine management methods face many challenges, such as information islands, inadequate utilization of data resources, backward disaster monitoring methods, and decentralized control systems without unified coordination.

[0003] Integrated intelligent mine operation and management technology is a core area of ​​digital transformation in the mining industry. By integrating next-generation information technologies such as the Internet of Things, artificial intelligence, big data, and cloud computing, it aims to achieve intelligent, efficient, and safer mining operations throughout the entire production process. However, existing smart mine solutions, such as digital twin mine technology, utilize digital technology to construct a virtual model of the mine. This model is synchronized with the actual mine in real time, encompassing comprehensive information such as geological structure, orebody distribution, equipment operating status, and personnel location. This provides an intuitive and accurate digital platform for mine operations management, enabling precise decision support, full-lifecycle management of equipment, and visual communication and collaboration. However, since building highly realistic virtual scenes requires high precision and massive amounts of data, initial construction costs are high. Ensuring precise synchronization between the virtual model and the actual mine presents significant challenges in terms of data real-time performance, including data transmission delays, data loss or deviations caused by sensor failures, and other issues. Furthermore, these solutions require high-level technical expertise from development and maintenance personnel. Furthermore, solutions such as automated mine production technology introduce automated machinery and control systems in key processes such as mining, transportation, and ore processing, reducing manual labor and enabling autonomous and precise control of the production process. This improves production efficiency, enhances personnel safety, and reduces labor costs. However, compatibility issues such as interface mismatches and inconsistent communication protocols often arise in equipment across various stages of mining, transportation, and mineral processing, impacting overall system stability. Automated production lines also require lengthy initial installation and commissioning cycles, and any failures during operation require high-level emergency response capabilities from technical personnel. Alternatively, AI-based intelligent mining decision-making solutions utilize machine learning and deep learning algorithms to mine and analyze massive amounts of historical data accumulated by mines (such as production data, geological data, and equipment failure data). These algorithms automatically generate optimization plans and predictive results, assisting or even replacing management in decision-making, thereby enabling deep data insights, rapid decision optimization, and continuous learning and evolution. However, these solutions also present stringent data quality requirements, necessitating significant labor and time-consuming initial data cleaning and organization. Furthermore, the complex internal logic of some AI algorithms, such as deep learning, makes it difficult to intuitively explain the decision-making process. Furthermore, when mining processes and equipment are upgraded or new challenges arise, existing AI algorithms may become ineffective, making retraining models costly and time-consuming, and developing new algorithms requires significant technical expertise.

[0004] Therefore, from the perspective of the overall integration of production, sales, and operations, how to build an integrated intelligent operation and management system of "mine-transportation-factory" has important practical significance and value. Summary of the Invention

[0005] In view of the shortcomings of the existing technology, the present invention proposes an integrated intelligent operation and management method for mines based on BIM+GIS+IoT, and also provides an integrated intelligent operation and management system for mines based on BIM+GIS+IoT.

[0006] The BIM+GIS+IoT-based integrated intelligent mine operation and management method of the present invention is implemented as follows: It includes full system model construction, data collection and analysis, production operation, safety risk warning, visualization and control steps. The contents of each step are as follows: A. Construction of the full system model: Use BIM to build a real-life model of the mine system and reproduce the mine work area and surrounding environment in 3D; B. Data Collection and Analysis: Use sensors and testing equipment to monitor equipment, vehicles, and production processes in real time, and organize the monitoring data, including removing abnormal data and filling missing values ​​using the average method; C. Production and Operation: Driven by orders, we rely on the ERP system to conduct comprehensive quality management of the entire production and operation process and generate production and operation data; D. Safety Risk Warning: The dataset is used to train a deep learning Softmax regression model based on a neural network. The trained Softmax regression model is used to process and analyze the previously organized data. A cloud model is also used to process and analyze production and operation data. Finally, the risk assessment results generated after processing and analysis are output as feedback. E. Visualization and Control: Each component of the mine system reality model is assigned a specific label, and each label has independent read and write functions. The previously organized data, production and operation data, and risk assessment results are input into the storage space under the corresponding label. Then, an integrated full-factor display is displayed through the display device, and any part of the integrated full-factor display can be intuitively and arbitrarily extracted for browsing. The information uploaded by each control module is then reflected on the mine system reality model, and the response decision is issued to each control module for response and regulation. In the full system model building step, BIM and GIS data are stored in the corresponding IFC and City GML formats respectively, and then the relevant information can be directly viewed in the data collection and analysis step and the visualization and control step; In the data collection and analysis step, after data collection and processing using radio frequency identification technology, the data is uploaded to the IoT and then displayed on the display device in the visualization and control step. In the data collection and analysis step, the data collected by the sensor is directly sent to the security risk warning step to prepare for subsequent risk assessment. In the visualization and control step, the production and operation plan is adjusted in the production operation step according to the order fluctuation situation; the risk assessment results output by the safety risk warning step are displayed in the visualization and control step, and relevant instructions are issued according to the results; In the production operation step, production data and financial information are transmitted to the security risk warning step for risk assessment.

[0007] Furthermore, the above-mentioned step A includes BIM multi-scenario modeling and terrain environment modeling sub-steps, which are as follows: A10. BIM multi-scenario modeling: This includes the modeling of mines and related facilities, transportation systems and related facilities, and factory areas and related facilities. Each step uses BIM to construct digital models of mines and related facilities, transportation systems and related facilities, and factory areas and related facilities, and then forms the corresponding real-life models. A20. Terrain environment modeling: Scan the work area and surrounding environment through drone oblique photography to generate a real-life model of the terrain environment.

[0008] Furthermore, the specific process of step B is: using various types of sensors and detection equipment to collect real-time data on the equipment operating temperature of the mine, transportation system and factory area, the running path and load of the mine car, traffic flow, process processing status, waste index monitoring, air humidity, pollutant indicators, weather warnings, noise level, carbon emissions, operator positioning and physical health monitoring, and then organize the above-mentioned collected data. Finally, the organized data is transmitted to the visualization and control steps and the safety risk warning steps respectively through the IoT platform.

[0009] Furthermore, the C step includes production plan adjustment, supply chain management, quality management, order management, warehouse management, logistics and transportation, financial management, and personnel management sub-steps. Each sub-step directly relies on the ERP system to conduct comprehensive quality management of the entire production operation process and form financial information and production information data.

[0010] Furthermore, in the step D, the data set is the data collected and sorted in the early stage, and the data set is divided into a sample training set, a validation set, and a test set in a ratio of 8:1:1 to train the Softmax regression model; the input value of the Softmax regression model is a vector composed of multi-source data formed in the data collection and analysis step, and the output value is a plurality of numerical values ​​corresponding to different fault types, which are expressed as , n is the total sample size.

[0011] Furthermore, in the step D, for the above-mentioned sorted data, each neuron of the Softmax regression model represents the output of a fault classification. i OutputG i for: , Where: w ij For the i The first neuron j weights, X j is a vector input composed of multi-source data, b i is the offset; For the input of the aforementioned multi-source data, the corresponding output can be obtained G i , and then the Softmax regression model calculates the corresponding probability of such failure , G represents the probability of each type of fault, and the fault type with the highest probability is taken as the output result y 0, where: , Where: k is the total number of fault types.

[0012] Furthermore, the specific process of using the cloud model to process and analyze the aforementioned production and operation data is as follows: First, we conduct multi-dimensional and index-based discussions on financial information through manual experience, and then have relevant personnel score and obtain a comprehensive score matrix. The comprehensive score matrix and production information are used as data inputs for the Softmax regression model, and the range method in the entropy method is used to standardize the data to obtain the weight of each indicator. W j : , Where: y ij For the i The first sample j The standardized ratio of the indicators, m is the number of samples, n is the number of indicators, where k=1 / ln n; Secondly, to facilitate the visualization of the results, the fixed interval method was adopted and the cloud model was used to divide the risk levels into 5 levels: I, II, III, IV and V. The intervals of each risk level are [0, 20), [20, 40), [40, 60), [60, 80), [80, 100]. Then, determine the unified cloud digital characteristics of the cloud model Ex = ( φ max + φ min ) / 2,En = ( φ max - φ min ) / 2.355, He = α , where a is set based on the actual visualization situation after debugging. φ max is the upper limit of the level interval constraint, φ min is the lower limit of the level interval constraint; then the cloud model membership matrix is ​​established through the cloud generator μ : , in, En It satisfies the normal distribution N ( En , He 2 ), expressed as En’~N ( En , He 2 ); Finally, by weight W j and membership matrix μ The comprehensive membership distribution is obtained together , and according to the maximum principle, find C max The risk level interval corresponding to the value is used to obtain the corresponding risk level.

[0013] Furthermore, the E step includes the virtual reality of the entire system, real-time monitoring data of each indicator, and control system sub-steps, which are as follows: E10, Full System Virtual Reality: Integrate the collated data, production and operation data, and risk assessment results with the mine system real-life model to display all elements of the mine in an integrated manner, allowing users to intuitively and freely extract any part of the information for browsing; E20, Real-time monitoring data of various indicators: Integrate the dynamic parameters of real-time monitoring with the real-scene model of the mining system, and then display them in an integrated and comprehensive manner through display devices, showing the dynamic changes of various parameters under monitoring in real time; E30, control system: reflects the information uploaded by each control module on the real-life model of the mining system, and transmits the response decision to each control module for response regulation.

[0014] The integrated intelligent operation and management system for mines based on BIM+GIS+IoT is implemented as follows: it includes a full-system model building module, a data acquisition and analysis module, a production and operation module, a safety risk warning module, and a visualization and control module. The full system model building module is used to build a real-scene model of the mine system through BIM and reproduce the mine work area and surrounding environment in 3D; The data acquisition and analysis module is used to use sensors and detection equipment to monitor equipment, vehicles, and production processes in real time, and to organize the monitoring data, including eliminating abnormal data and filling missing values ​​using the average method; The production operation module is used to perform comprehensive quality management of the entire production operation process through order-driven and ERP system to generate production and operation data; The safety risk warning module is used to train a deep learning Softmax regression model based on a neural network through a data set, use the trained Softmax regression model to process and analyze the aforementioned sorted data, and simultaneously use a cloud model to process and analyze production and operation data, and finally output the risk assessment results formed by the processing and analysis as feedback; The visualization and control module is used to assign specific tags to each component of the mine system real-life model, and each tag has independent read and write functions. The aforementioned organized data, production and operation data, and risk assessment results are input into the storage space under the corresponding tag, and then an integrated full-factor display is performed through the display device. The information of any part of the integrated full-factor display can be intuitively and arbitrarily extracted for browsing; the information uploaded by each control module is then reflected on the mine system real-life model, and the response decision is issued to each control module for response regulation; The full system model building module stores BIM and GIS data in corresponding IFC and City GML formats respectively, and then directly views the relevant information in the data acquisition and analysis module and the visualization and control module; The data acquisition and analysis module collects and processes data using radio frequency identification technology, uploads the data to the IoT, and then displays it on the display device in the visualization and control module. The data acquisition and analysis module directly sends the data collected by the sensor to the security risk warning module to prepare for subsequent risk assessment. The visualization and control module adjusts the production and operation plan in the production operation module according to order fluctuations; the visualization and control module is used to display the risk assessment results output by the safety risk warning module and issue relevant instructions based on the results; The production operation module transmits production data and financial information to the security risk warning module for risk assessment.

[0015] Furthermore, in the safety risk warning module, the data set is the data collected and sorted in the early stage, and the data set is divided into a sample training set, a validation set and a test set in a ratio of 8:1:1 to train the Softmax regression model; the input value of the Softmax regression model is a vector composed of multi-source data formed by the data acquisition and analysis module, and the output value is a plurality of numerical values ​​corresponding to different fault types, which are expressed as , n is the total sample size.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention uses data as a medium for interaction, integrating BIM (Building Information Modeling), GIS (Geographic Information System), IoT (Internet of Things) and artificial intelligence technologies. From the perspective of global integrated management of supply and demand, production, sales and operations, it constructs a real-life model of the mining system covering the entire process of "mine-transportation-factory". Sensors are used to obtain the operating status information of relevant facilities, and the real-life model is digitally filled through data collection and analysis. Relying on the ERP system to regulate daily production through order-driven, the Softmax regression model and cloud model are applied to conduct all-round risk control of production and operation, breaking through the barriers of multi-source data such as geology, production, sales, and personnel, solving the problems of system dispersion and data isolation in traditional mine management, realizing real-time sharing and collaboration of data in production, transportation, operation and other links, improving resource allocation efficiency, forming a full life cycle management model for mineral resource development, and promoting the green transformation and sustainable development of mines.

[0017] 2. This invention uses a sensor network to collect key parameters such as equipment operating temperature, mine car path, pollutant indicators, and personnel health in real time, combines it with the IoT platform to achieve efficient data transmission, and uses dynamic data filling of BIM and GIS to ensure that the virtual model is accurately synchronized with the actual mine status, significantly reducing the risk of decision-making lag caused by delays or data loss in traditional digital twin technology.

[0018] 3. The present invention links the ERP system with the visualization platform to dynamically adjust the production plan based on order demand, optimize the supply chain, warehousing and logistics links, and combines it with comprehensive quality management to achieve precise control of resource consumption and production progress, thereby reducing waste and improving overall operational efficiency.

[0019] 4. This invention adopts the Softmax regression model to conduct deep learning analysis on multi-source data (equipment status, environmental indicators, production information), combines manual experience scoring and cloud model risk grading to achieve multi-dimensional risk assessment, thereby providing effective safety risk warnings, timely preventing potential risks in equipment, environment, etc., and reducing operating management costs and labor costs.

[0020] 5. The present invention deeply integrates real-time data, risk warnings and 3D real-scene models, supports detailed information extraction and dynamic parameter display of any scene, and managers can quickly locate problems through an intuitive visual interface and directly issue control instructions through an integrated control system, which not only shortens emergency response time but also improves decision-making efficiency.

[0021] In summary, the present invention organically combines mining, processing, production, sales and other links, reduces the communication costs between different systems, and is conducive to creating a new industry with high efficiency, low risk, environmental friendliness and people-oriented, which can form a typical demonstration. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is the flow chart of the integrated intelligent operation and management method of mines based on BIM+GIS+IoT of the present invention. DETAILED DESCRIPTION

[0023] The present invention will be further described below with reference to the accompanying drawings and embodiments, but the present invention is not limited in any way. Any changes or improvements made based on the teachings of the present invention fall within the scope of protection of the present invention.

[0024] like Figure 1 As shown in the figure, the integrated intelligent operation and management method of mines based on BIM+GIS+IoT of the present invention includes the steps of full system model construction, data collection and analysis, production operation, safety risk warning, visualization and control. The contents of each step are as follows: A. Construction of the full system model: Use BIM to build a real-life model of the mine system and reproduce the mine work area and surrounding environment in 3D; B. Data Collection and Analysis: Use sensors and testing equipment to monitor equipment, vehicles, and production processes in real time, and organize the monitoring data, including removing abnormal data and filling missing values ​​using the average method; C. Production and Operation: Driven by orders, we rely on the ERP system to conduct comprehensive quality management of the entire production and operation process and generate production and operation data; D. Safety Risk Warning: The dataset is used to train a deep learning Softmax regression model based on a neural network. The trained Softmax regression model is used to process and analyze the previously organized data. A cloud model is also used to process and analyze production and operation data. Finally, the risk assessment results generated after processing and analysis are output as feedback. E. Visualization and Control: Each component of the mine system reality model is assigned a specific label, and each label has independent read and write functions. The previously organized data, production and operation data, and risk assessment results are input into the storage space under the corresponding label. Then, an integrated full-factor display is displayed through the display device, and any part of the integrated full-factor display can be intuitively and arbitrarily extracted for browsing. The information uploaded by each control module is then reflected on the mine system reality model, and the response decision is issued to each control module for response and regulation. In the full system model building step, BIM and GIS data are stored in the corresponding IFC and City GML formats respectively. The two formats can be directly converted to each other to achieve data sharing. Afterwards, relevant information can be directly viewed in the data collection and analysis step and the visualization and control step; In the data collection and analysis step, after data collection and processing using radio frequency identification technology, the data is uploaded to the IoT and then displayed on the display device in the visualization and control step. In the data collection and analysis step, the data collected by the sensor is directly sent to the security risk warning step to prepare for subsequent risk assessment. In the visualization and control step, the production and operation plan is adjusted in the production operation step according to the order fluctuation situation; the risk assessment results output by the safety risk warning step are displayed in the visualization and control step, and relevant instructions are issued according to the results; In the production operation step, production data and financial information are transmitted to the security risk warning step for risk assessment.

[0025] Step A includes BIM multi-scenario modeling and terrain environment modeling steps, as follows: A10. BIM multi-scenario modeling: This includes the modeling of mines and related facilities, transportation systems and related facilities, and factory areas and related facilities. Each step uses BIM to construct digital models of mines and related facilities, transportation systems and related facilities, and factory areas and related facilities, and then forms the corresponding real-life models. A20. Terrain environment modeling: Scan the work area and surrounding environment through drone oblique photography to generate a real-life model of the terrain environment.

[0026] The B step includes the mining system, transportation system, plant system, environmental information, personnel positioning and health monitoring sub-steps. The mining system sub-steps include equipment operation monitoring and real-time site update steps. The transportation system sub-steps include mine car operation monitoring, human-machine collaboration, and intelligent transportation system steps (ITS). The plant system sub-steps include process detection, automated production, and waste treatment steps. The environmental information sub-steps include pollutant monitoring, weather warning, carbon emission monitoring, and noise monitoring steps. Each step uses sensors or detection equipment to perform corresponding monitoring and obtain data.

[0027] The specific process of step B is: using various types of sensors and detection equipment to collect real-time data on the equipment operating temperature of mines, transportation systems and factory areas, the running path and load of mine cars, traffic flow, process processing status, waste index monitoring, air humidity, pollutant indicators, weather warnings, noise levels, carbon emissions, operator positioning and physical health monitoring, and then organize the above-mentioned collected data. Finally, the organized data is transmitted to the visualization and control steps and the safety risk warning steps respectively through the IoT platform.

[0028] The C step includes production plan adjustment, supply chain management, quality management, order management, warehouse management, logistics and transportation, financial management, and personnel management. Each step directly relies on the ERP system to conduct comprehensive quality management of the entire production operation process and form financial information and production information data.

[0029] The D step includes data analysis and processing, fuzzy evaluation, evaluation result output and early warning sub-steps. The data analysis and processing sub-step uses a deep learning Softmax regression model based on a neural network to process and analyze production-type multi-source data, and improves the adaptability of the model by continuously training the model accuracy; the fuzzy evaluation sub-step uses a combination of manual analysis and qualitative indicators to conduct risk assessment on production, finance and other information to ensure smooth production operation; the evaluation result output and early warning sub-step outputs and feeds back the risk ratings of different control modules to improve the effectiveness of early warnings, and issue advance warnings for high-risk production events to achieve the effect of "treating existing diseases and preventing future diseases."

[0030] In step D, the data set is the data collected and sorted in the early stage. The data set is divided into a sample training set, a validation set, and a test set in a ratio of 8:1:1 to train the Softmax regression model. The input value of the Softmax regression model is a vector composed of multi-source data formed in the data collection and analysis step, and the output value is multiple values ​​corresponding to different fault types, expressed as , n is the total sample size.

[0031] Furthermore, in the step D, for the above-mentioned sorted data, each neuron of the Softmax regression model represents the output of a fault classification. i Output G i for: , Where: w ij For the i The first neuron j weights, X j is a vector input composed of multi-source data, b i is the offset; For the input of the aforementioned multi-source data, the corresponding output can be obtained G i , and then the Softmax regression model calculates the corresponding probability of such failure , G represents the probability of each type of fault, and the fault type with the highest probability is taken as the output result y 0, where: , Where: k is the total number of fault types.

[0032] The specific process of using the cloud model to process and analyze the aforementioned production and operation data is as follows: First, we conduct multi-dimensional and index-based discussions on financial information through manual experience, and then have relevant personnel score and obtain a comprehensive score matrix. The comprehensive score matrix and production information are used as data inputs for the Softmax regression model, and the range method in the entropy method is used to standardize the data to obtain the weight of each indicator. W j : , Where: y ij For the i The first sample j The standardized ratio of the indicators, m is the number of samples, n is the number of indicators, where k = 1 / ln n; secondly, to facilitate the visualization of the results, the fixed interval method is adopted and the cloud model is used to divide the risk levels into 5 levels: I, II, III, IV and V. The intervals of each risk level are [0, 20), [20, 40), [40, 60), [60, 80), [80, 100]; Then, determine the unified cloud digital characteristics of the cloud model Ex = (φ max + φ min ) / 2, En = ( φ max - φ min ) / 2.355, He = α , where a is set based on the actual visualization situation after debugging. φ max is the upper limit of the level interval constraint, φ min is the lower limit of the level interval constraint; then the cloud model membership matrix is ​​established through the cloud generator μ : , in, En It satisfies the normal distribution N ( En , He 2 ), expressed as En’~N ( En , He 2 ); Finally, by weight W j and membership matrix μ The comprehensive membership distribution is obtained together , and according to the maximum principle, find C max The risk level interval corresponding to the value is used to obtain the corresponding risk level.

[0033] The cloud model is composed of ( Ex, En, He ) characterizes; wherein: Ex Indicates the central value of the evaluation data; En Indicates the fuzziness of the evaluation results and can reflect the degree of discreteness of cloud droplets; He The entropy represents entropy, He The larger it is, the thicker the cloud droplets are.

[0034] The cloud model is a model based on random mathematics and fuzzy mathematics. It can effectively overcome the ambiguity of risk assessment, reduce the subjectivity of evaluation, realize the correspondence between quantitative indicator data and qualitative levels, and achieve the purpose of quantitative and qualitative transformation.

[0035] The E step includes the virtual reality of the entire system, real-time monitoring data of each indicator, and control system sub-steps, as follows: E10, Full System Virtual Reality: Integrate the collated data, production and operation data, and risk assessment results with the mine system real-life model to display all elements of the mine in an integrated manner, allowing users to intuitively and freely extract any part of the information for browsing; E20, Real-time monitoring data of various indicators: Integrate the dynamic parameters of real-time monitoring with the real-scene model of the mining system, and then display them in an integrated and comprehensive manner through display devices, showing the dynamic changes of various parameters under monitoring in real time; E30, control system: reflects the information uploaded by each control module on the real-life model of the mining system, and transmits the response decision to each control module for response regulation.

[0036] It should be noted that the visualization and control step issues model modification information through the control system. Based on this modification information, the full-system model construction step digitally models the newly constructed, renovated, and expanded facilities and buildings, the current status of the site construction, and the surrounding terrain environment through BIM (possibly through drone oblique photography technology), providing detailed and accurate management information on the structure, equipment, and process of the structures. The mine, transportation system, plant area, and terrain environment are visualized in 3D, and the 3D model is transmitted back to the visualization and control step for management personnel to review. The full-system model construction step uses GIS to collect images from multiple angles, gather geospatial data and information, and conduct management analysis. The mine, transportation system, plant area, and terrain environment are geographically located, and this data is passed to the data collection and analysis step, paving the way for data filling and building an intuitive virtual digital twin corresponding to the real world.

[0037] It's important to note that the production operations and safety risk warning steps are connected through production information (physical information about equipment operation) and financial information (currency management, upstream and downstream customer credit information, and operational information generated during fund settlement). IoT, through various sensors and detection equipment, collects real-time data on equipment operating temperature, mine car routes and loads, traffic flow, process status, waste index monitoring, air humidity, pollutant indicators, weather warnings, noise levels, carbon emissions, operator location, and vital health monitoring. This data is then transmitted to the visualization and control steps via the Industrial Internet of Things platform, enabling real-time perception of the entire system's operating status. Currently, communication protocols that enable interoperable data are available. For example, IoT can collect field data through sensors and other technologies, converting the analog output of these sensors into digital signals using digital-to-analog converters, facilitating the subsequent processing and application of various types of information. At the same time, the data set is uploaded to the safety risk warning step. The safety risk warning step adopts the Softmax regression model in the deep learning model. By classifying the similarity of the data features between the data samples and the fault samples of each category, the probability distribution of various types of faults in the data samples can be clarified. The value of the fault probability can be used as the basis for evaluating the equipment operation status.

[0038] It should be noted that the production operation process is integrated with ERP. The visualization and control step transmits order information directly to the production plan adjustment step of the production operation step. ERP automatically adjusts procurement plans and production volumes, and manages quality throughout the entire process, from raw materials to finished product delivery. Orders are archived and stored, raw material yards are rationally planned, and finished product warehouses are stored through inventory management methods and connected with shipments. The entire transportation cycle from the finished product delivery point to the destination is monitored, and real-time transportation information for each transport trip is obtained, including real-time positioning of transport vehicles, driver driving status monitoring, estimated time of arrival at the destination, and transportation route planning in emergency situations. At the same time, financial funds are synchronized, and a dedicated person is responsible for managing personnel throughout the entire process to ensure the coordination of logistics, business flow, and information flow. The safety risk warning step classifies risk warning information from different control modules into three colors: green (no risk, normal operation), yellow (potential risk, equipment maintenance), and red (high risk, downtime for maintenance). The information is uploaded to the visualization and control module for presentation. The visualization and control step then publishes information to different control modules based on different risk levels, guiding each control module to implement targeted control.

[0039] The integrated intelligent operation and management system for mines based on BIM+GIS+IoT includes a full-system model building module, a data acquisition and analysis module, a production and operation module, a safety risk warning module, and a visualization and control module. The full system model building module is used to build a real-scene model of the mine system through BIM and reproduce the mine work area and surrounding environment in 3D; The data acquisition and analysis module is used to use sensors and detection equipment to monitor equipment, vehicles, and production processes in real time, and to organize the monitoring data, including eliminating abnormal data and filling missing values ​​using the average method; The production operation module is used to perform comprehensive quality management of the entire production operation process through order-driven and ERP system to generate production and operation data; The safety risk warning module is used to train a deep learning Softmax regression model based on a neural network through a data set, use the trained Softmax regression model to analyze and process the aforementioned sorted data, and simultaneously use a cloud model to process and analyze production and operation data, and finally output the risk assessment results formed after processing and analysis as feedback; The visualization and control module is used to assign specific tags to each component of the mine system real-life model, and each tag has independent read and write functions. The aforementioned organized data, production and operation data, and risk assessment results are input into the storage space under the corresponding tag, and then an integrated full-factor display is performed through the display device. The information of any part of the integrated full-factor display can be intuitively and arbitrarily extracted for browsing; the information uploaded by each control module is then reflected on the mine system real-life model, and the response decision is issued to each control module for response regulation; The full system model building module stores BIM and GIS data in corresponding IFC and City GML formats respectively, and then directly views the relevant information in the data acquisition and analysis module and the visualization and control module; The data acquisition and analysis module collects and processes data using radio frequency identification technology, uploads the data to the IoT, and then displays it on the display device in the visualization and control module. The data acquisition and analysis module directly sends the data collected by the sensor to the security risk warning module to prepare for subsequent risk assessment. The visualization and control module adjusts the production and operation plan in the production operation module according to order fluctuations; the visualization and control module is used to display the risk assessment results output by the safety risk warning module and issue relevant instructions based on the results; The production operation module transmits production data and financial information to the security risk warning module for risk assessment.

[0040] In the safety risk warning module, the data set is the data collected and sorted in the early stage. The data set is divided into a sample training set, a validation set, and a test set in a ratio of 8:1:1 to train the Softmax regression model. The input value of the Softmax regression model is a vector composed of multi-source data formed by the data acquisition and analysis module, and the output value is multiple values ​​corresponding to different fault types, expressed as , n is the total sample size.

[0041] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. The integrated intelligent operation and management method for mines based on BIM+GIS+IoT is characterized by: It includes full system model construction, data collection and analysis, production operation, safety risk warning, visualization and control steps. The contents of each step are as follows: A. Construction of the full system model: Use BIM to build a real-life model of the mine system and reproduce the mine work area and surrounding environment in 3D; B. Data Collection and Analysis: Use sensors and testing equipment to monitor equipment, vehicles, and production processes in real time, and organize the monitoring data, including removing abnormal data and filling missing values ​​using the average method; C. Production and Operation: Driven by orders, we rely on the ERP system to conduct comprehensive quality management of the entire production and operation process and generate production and operation data; D. Safety Risk Warning: The dataset is used to train a deep learning Softmax regression model based on a neural network. The trained Softmax regression model is used to process and analyze the previously organized data. A cloud model is also used to process and analyze production and operation data. Finally, the risk assessment results generated after processing and analysis are output as feedback. E. Visualization and Control: Each component of the mine system reality model is assigned a specific tag, and each tag has independent read and write functions. The previously organized data, production and operation data, and risk assessment results are input into the storage space under the corresponding tag. Then, an integrated full-factor display is displayed through the display device, and any part of the integrated full-factor display can be intuitively and arbitrarily extracted for browsing; Then, the information uploaded by each control module is reflected on the real-life model of the mining system, and the response decision is passed to each control module for response regulation; In the full system model building step, BIM and GIS data are stored in the corresponding IFC and CityGML formats respectively, and then the relevant information can be directly viewed in the data collection and analysis step and the visualization and control step; In the data collection and analysis step, after data collection and processing using radio frequency identification technology, the data is uploaded to the IoT and then displayed on the display device in the visualization and control step. In the data collection and analysis step, the data collected by the sensor is directly sent to the security risk warning step to prepare for subsequent risk assessment. In the visualization and control step, the production and operation plan is adjusted in the production operation step according to the order fluctuation situation; the risk assessment results output by the safety risk warning step are displayed in the visualization and control step, and relevant instructions are issued according to the results; In the production operation step, production data and financial information are transmitted to the security risk warning step for risk assessment.

2. The integrated intelligent operation and management method for mines based on BIM+GIS+IoT according to claim 1 is characterized by: Step A includes BIM multi-scenario modeling and terrain environment modeling steps, as follows: A10. BIM multi-scenario modeling: This includes the modeling of mines and related facilities, transportation systems and related facilities, and factory areas and related facilities. Each step uses BIM to construct digital models of mines and related facilities, transportation systems and related facilities, and factory areas and related facilities, and then forms the corresponding real-life models. A20. Terrain environment modeling: Scan the work area and surrounding environment through drone oblique photography to generate a real-life model of the terrain environment.

3. The integrated intelligent operation and management method for mines based on BIM+GIS+IoT according to claim 1 is characterized by: The specific process of step B is: using various types of sensors and detection equipment to collect real-time data on the equipment operating temperature of mines, transportation systems and factory areas, the running path and load of mine cars, traffic flow, process processing status, waste index monitoring, air humidity, pollutant indicators, weather warnings, noise levels, carbon emissions, operator positioning and physical health monitoring, and then organize the above-mentioned collected data. Finally, the organized data is transmitted to the visualization and control steps and the safety risk warning steps respectively through the IoT platform.

4. The integrated intelligent operation and management method for mines based on BIM+GIS+IoT according to claim 1, 2 or 3, is characterized by: The C step includes production plan adjustment, supply chain management, quality management, order management, warehouse management, logistics and transportation, financial management, and personnel management. Each step directly relies on the ERP system to conduct comprehensive quality management of the entire production operation process and form financial information and production information data.

5. The integrated intelligent operation and management method for mines based on BIM+GIS+IoT according to claim 4 is characterized by: In step D, the data set is the data collected and sorted in the early stage. The data set is divided into a sample training set, a validation set, and a test set in a ratio of 8:1:1 to train the Softmax regression model. The input value of the Softmax regression model is a vector composed of multi-source data formed in the data collection and analysis step, and the output value is multiple values ​​corresponding to different fault types, expressed as , n is the total sample size.

6. The integrated intelligent operation and management method for mines based on BIM+GIS+IoT according to claim 5 is characterized by: In the D step, for the above-mentioned sorted data, each neuron of the Softmax regression model represents the output of a fault classification. i Output G i for: ， Where: w ij For the i The first neuron j weights, X j is a vector input composed of multi-source data, b i is the offset; For the input of the aforementioned multi-source data, the corresponding output G i , and then the Softmax regression model calculates the corresponding probability of such failure , G represents the probability of each type of fault, and the fault type with the highest probability is taken as the output result y 0, where: ， Where: k is the total number of fault types.

7. The integrated intelligent operation and management method for mines based on BIM+GIS+IoT according to claim 5 is characterized by: The specific process of using the cloud model to process and analyze the aforementioned production and operation data is as follows: First, we conduct multi-dimensional and index-based discussions on financial information through manual experience, and then have relevant personnel score and obtain a comprehensive score matrix. The comprehensive score matrix and production information are used as data input for the cloud model, and the range method in the entropy method is used to standardize the data to obtain the weight of each indicator. W j : ， Where: y ij For the i The first sample j The standardized ratio of the indicators, m is the number of samples, n is the number of indicators, where k =1 / ln n ; Secondly, to facilitate the visualization of the results, the fixed interval method was adopted and the cloud model was used to divide the risk levels into 5 levels: I, II, III, IV and V. The intervals of each risk level are [0, 20), [20, 40), [40, 60), [60, 80), [80, 100]. Then, determine the unified cloud digital characteristics of the cloud model Ex = ( φ max + φ min ) / 2, En = ( φ max - φ min ) / 2.355, He = α , where a is set based on the actual visualization situation after debugging. φ max is the upper limit of the level interval constraint, φ min is the lower limit of the grade interval constraint; Then the cloud model membership matrix is ​​established through the cloud generator μ : ， in, En It satisfies the normal distribution N ( En , He 2 ), expressed as En'~N ( En , He 2 ); Finally, by weight W j and membership matrix μ The comprehensive membership distribution is obtained together , and according to the maximum principle, find C max The risk level interval corresponding to the value is used to obtain the corresponding risk level.

8. The integrated intelligent operation and management method for mines based on BIM+GIS+IoT according to claim 1 is characterized by: The E step includes the virtual reality of the entire system, real-time monitoring data of each indicator, and control system sub-steps, as follows: E10, Full System Virtual Reality: Integrate the collated data, production and operation data, and risk assessment results with the mine system real-life model to display all elements of the mine in an integrated manner, allowing users to intuitively and freely extract any part of the information for browsing; E20, Real-time monitoring data of various indicators: Integrate the dynamic parameters of real-time monitoring with the real-scene model of the mining system, and then display them in an integrated and comprehensive manner through display devices, showing the dynamic changes of various parameters under monitoring in real time; E30, control system: reflects the information uploaded by each control module on the real-life model of the mining system, and transmits the response decision to each control module for response regulation.

9. The integrated intelligent mine operation and management system based on BIM+GIS+IoT is characterized by: It includes full system model building module, data collection and analysis module, production operation module, safety risk warning module, visualization and control module, The full system model building module is used to build a real-scene model of the mine system through BIM and reproduce the mine work area and surrounding environment in 3D; The data acquisition and analysis module is used to use sensors and detection equipment to monitor equipment, vehicles, and production processes in real time, and to organize the monitoring data, including eliminating abnormal data and filling missing values ​​using the average method; The production operation module is used to perform comprehensive quality management of the entire production operation process through order-driven and ERP system to generate production and operation data; The safety risk warning module is used to train a deep learning Softmax regression model based on a neural network through a data set, use the trained Softmax regression model to process and analyze the aforementioned sorted data, and simultaneously use a cloud model to process and analyze production and operation data, and finally output the risk assessment results formed after processing and analysis as feedback; The visualization and control module is used to assign specific tags to each component of the mine system real-life model, and each tag has independent read and write functions. The aforementioned collated data, production and operation data, and risk assessment results are input into the storage space under the corresponding tag, and then an integrated full-factor display is performed through the display device. The user can intuitively and arbitrarily extract any part of the integrated full-factor display for browsing; Then, the information uploaded by each control module is reflected on the real-life model of the mining system, and the response decision is passed to each control module for response regulation; The full system model building module stores BIM and GIS data in corresponding IFC and CityGML formats respectively, and then directly views the relevant information in the data acquisition and analysis module and the visualization and control module; The data acquisition and analysis module collects and processes data using radio frequency identification technology, uploads the data to the IoT, and then displays it on the display device in the visualization and control module. The data acquisition and analysis module directly sends the data collected by the sensor to the security risk warning module to prepare for subsequent risk assessment. The visualization and control module adjusts the production and operation plan in the production operation module according to order fluctuations; the visualization and control module is used to display the risk assessment results output by the safety risk warning module and issue relevant instructions based on the results; The production operation module transmits production data and financial information to the security risk warning module for risk assessment.

10. The integrated intelligent mine operation and management system based on BIM+GIS+IoT according to claim 9 is characterized by: In the safety risk warning module, the data set is the data collected and sorted in the early stage. The data set is divided into a sample training set, a validation set, and a test set in a ratio of 8:1:1 to train the Softmax regression model. The input value of the Softmax regression model is a vector composed of multi-source data formed by the data acquisition and analysis module, and the output value is multiple values ​​corresponding to different fault types, expressed as ，n is the total sample size.

Images