Intelligent mine centralized control system
Patent Information
- Application Number
- CN202522321109.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Utility models(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-03
- Publication Date
- 2026-10-13
- Estimated Expiration
- 2035-11-03
AI Technical Summary
网络依赖性强，矿山通常处于内网封闭环境，且井下通信条件复杂，一旦网络中断或延迟，云端推理服务随即失效，导致集控系统失去智能决策能力，影响生产连续性；
1、系统由本地服务器、数据采集模块、数据处理模块、智能问答模块、前端展示模块和模型更新模块组成，各模块通过本地局域网互联，形成离线数据闭环，现场无需互联网即可实现AI问答与决策；
Smart Images
Figure CN224851527U_ABST
Abstract
Description
Technical Field
[0001] This utility model relates to the field of mine automation control technology, specifically to an intelligent mine centralized control system. Background Technology
[0002] With the rapid development of modern mining technology, intelligent mine control systems have become the core support for mines to achieve efficient, safe, and intelligent operation. These systems typically utilize artificial intelligence technology, especially cloud-based online AI services, to perform real-time analysis and decision support on information such as mine equipment status, production data, and environmental monitoring.
[0003] However, traditional intelligent mine centralized control systems have the following problems: The network is highly dependent on mines, which are usually located in closed intranet environments and have complex underground communication conditions. Once the network is interrupted or delayed, the cloud inference service will fail immediately, causing the centralized control system to lose its intelligent decision-making ability and affecting the continuity of production. Data security and privacy risks exist because sensitive information such as power distribution data, equipment status, and production plans need to be uploaded via the Internet, which poses a risk of interception, tampering, or leakage. This does not comply with the protection requirements for critical information infrastructure under the Cybersecurity Law and the Data Security Law. Utility Model Content
[0004] The purpose of this invention is to provide an intelligent mine centralized control system to solve the problems mentioned in the background art.
[0005] To achieve the above objectives, this utility model provides the following technical solution: an intelligent mine centralized control system, comprising: Local server, data acquisition module, data processing module, intelligent question answering module, front-end display module, and model update module; The local server has a built-in GPU acceleration card and is equipped with a GPT language model finely tuned from mining professional corpus, which is used to complete intelligent question answering and decision reasoning in an environment without the Internet. The data acquisition module is connected to the mine's power distribution equipment via an RS485 bus and collects current, voltage, power, and switch status in real time via the Modbus-RTU protocol. The data processing module and the data acquisition module are connected via Ethernet and are used to clean, normalize, and structure the acquired data for storage, and to form a real-time data stream. The intelligent question-answering module connects to the local server and data processing module via a RESTful API. It receives users' natural language questions and inputs the real-time data stream and the questions into the GPT model to obtain structured answers. The front-end display module is connected to the intelligent question-answering module via WebSocket, and is used to present answers to users in the form of charts, text, or alarms. The model update module is connected to the local server and is used to update the weights of the GPT model through local retraining or encrypted offline packages. The modules are interconnected via a local area network to form an offline intelligent centralized control system that can operate in environments without the internet.
[0006] Furthermore, the data acquisition module also supports at least one of Modbus-TCP, OPC UA, or MQTT protocols.
[0007] Furthermore, the front-end display module is built on the Vue.js framework and maintains bidirectional real-time data refresh with the intelligent question-answering module via WebSocket.
[0008] Furthermore, the model update module supports a completely offline approach, retraining the GPT language model using local historical labeled data and replacing the original weight file through hot loading.
[0009] Furthermore, the model update module also supports a temporary encrypted network connection method, which downloads the official model parameter package through an encrypted channel verified by digital signature, and automatically completes the version upgrade after verifying its integrity.
[0010] Furthermore, the GPT model is ChatGLM-6B, which has been fine-tuned with a mining-specific corpus and has functions such as equipment failure prediction, energy consumption analysis, and scheduling optimization.
[0011] Furthermore, the model update module is equipped with a physical one-way import gate, which allows only data writing and blocks data back transmission during temporary encrypted network upgrades, ensuring physical isolation of the internal network.
[0012] The technical effects and advantages provided by this utility model in the above technical solution are as follows: 1. The system consists of a local server, a data acquisition module, a data processing module, an intelligent question-and-answer module, a front-end display module, and a model update module. The modules are interconnected through a local area network to form an offline data closed loop, enabling AI question-and-answer and decision-making without the need for internet access on-site. 2. Data flow is restricted to the mining area's local area network, and the model update module is configured with a physical one-way import gate to ensure that data can only enter and not exit, thus ensuring physical isolation of the internal network and preventing ransomware and sensitive information from being leaked. 3. The local server has a built-in GPU acceleration card and deploys the ChatGLM-6B model, which has been fine-tuned with professional mining corpus, to achieve natural language interaction and structured reasoning. Attached Figure Description
[0013] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this utility model. For those skilled in the art, other drawings can be obtained based on these drawings.
[0014] Figure 1 This is a system module relationship diagram of this utility model.
[0015] Explanation of reference numerals in the attached figures: 1. Local server; 2. Data acquisition module; 3. Data processing module; 4. Intelligent question answering module; 5. Front-end display module; 6. Model update module. Detailed Implementation
[0016] To enable those skilled in the art to better understand the technical solution of this utility model, the present utility model will be further described in detail below with reference to the accompanying drawings.
[0017] This utility model provides, for example Figure 1 The intelligent mine centralized control system shown includes: Local server 1, data acquisition module 2, data processing module 3, intelligent question answering module 4, front-end display module 5, and model update module 6; Local server 1 has a built-in GPU acceleration card and is equipped with a GPT language model finely tuned from a mining professional corpus, which is used to complete intelligent question answering and decision reasoning in an environment without the Internet. Data acquisition module 2 is connected to the mine's power distribution equipment via an RS485 bus and collects current, voltage, power, and switch status in real time via the Modbus-RTU protocol; The data processing module 3 is connected to the data acquisition module 2 via Ethernet. It is used to clean, normalize, and structure the acquired data for storage, and to form a real-time data stream. The intelligent question answering module 4 connects to the local server 1 and the data processing module 3 via a RESTful API. It receives users' natural language questions and inputs the real-time data stream and the questions into the GPT model to obtain structured answers. The front-end display module 5 is connected to the intelligent question-answering module 4 via WebSocket, and is used to present the answers to the user in the form of charts, text or alarms. The model update module 6 is connected to the local server 1 and is used to update the weights of the GPT model through local retraining or encrypted offline packages. The modules are interconnected via a local area network to form an offline intelligent centralized control system that can operate in environments without internet access.
[0018] Data acquisition module 2 also supports at least one of Modbus-TCP, OPC UA or MQTT protocols.
[0019] The front-end display module 5 is built on the Vue.js framework and maintains bidirectional real-time data refresh with the intelligent question-answering module 4 via WebSocket.
[0020] Model update module 6 supports a completely offline approach, using local historical labeled data to retrain the GPT language model and replacing the original weight file through hot reloading.
[0021] Model update module 6 also supports a temporary encrypted network connection method, which downloads the official model parameter package through an encrypted channel verified by digital signature, and automatically completes the version upgrade after verifying its integrity.
[0022] The GPT model is ChatGLM-6B, which has been fine-tuned with a mining-specific corpus and has functions such as equipment failure prediction, energy consumption analysis, and scheduling optimization.
[0023] The model update module 6 is equipped with a physical one-way import gate, which allows only data writing and blocks data back transmission during temporary encrypted network upgrades, ensuring physical isolation of the internal network.
[0024] This system consists of a local server 1, a data acquisition module 2, a data processing module 3, an intelligent question-and-answer module 4, a front-end display module 5, and a model update module 6. The modules are interconnected through a local area network to form an offline data closed loop, enabling AI question-and-answer and decision-making without the need for internet access on-site. The data flow is confined to the mining area's local area network, and the model update module 6 is equipped with a physical one-way import gate to ensure that data can only enter and not leave, ensuring physical isolation of the internal network and preventing ransomware and leakage of sensitive information. The local server has a built-in GPU acceleration card and deploys the ChatGLM-6B model, which has been fine-tuned with professional mining corpus, to achieve natural language interaction and structured reasoning.
[0025] The foregoing description only illustrates certain exemplary embodiments of the present invention. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.
Claims
1. A smart mine centralized control system, characterized in that, include: Local server (1), data acquisition module (2), data processing module (3), intelligent question answering module (4), front-end display module (5), and model update module (6); The local server (1) has a built-in GPU acceleration card and is equipped with a GPT language model finely tuned by mining professional corpus, which is used to complete intelligent question answering and decision reasoning in an environment without the Internet. The data acquisition module (2) is connected to the mine power distribution equipment via RS485 bus and collects current, voltage, power and switch status in real time via Modbus-RTU protocol; The data processing module (3) and the data acquisition module (2) are connected via Ethernet and are used to clean, normalize, and structure the acquired data to form a real-time data stream. The intelligent question answering module (4) is connected to the local server (1) and the data processing module (3) through a RESTful API. It is used to receive users' natural language questions and input the real-time data stream and the questions into the GPT model to obtain structured answers. The front-end display module (5) is connected to the intelligent question-answering module (4) via WebSocket, and is used to present the answer to the user in the form of charts, text or alarms; The model update module (6) is connected to the local server (1) and is used to update the weights of the GPT model through local retraining or encrypted offline package. The modules are interconnected via a local area network to form an offline intelligent centralized control system that can operate in environments without the internet.
2. The intelligent mine centralized control system according to claim 1, characterized in that: The data acquisition module (2) also supports at least one of Modbus-TCP, OPC UA or MQTT protocols.
3. The intelligent mine centralized control system according to claim 1, characterized in that: The front-end display module (5) is built on the Vue.js framework and maintains bidirectional real-time data refresh with the intelligent question-answering module (4) through WebSocket.
4. The intelligent mine centralized control system according to claim 1, characterized in that: The model update module (6) supports a completely offline mode, using local historical labeled data to retrain the GPT language model and replace the original weight file through hot loading.
5. The intelligent mine centralized control system according to claim 1, characterized in that: The model update module (6) also supports a temporary encrypted network connection method, which downloads the official model parameter package through an encrypted channel verified by digital signature, and automatically completes the version upgrade after verifying the integrity.
6. The intelligent mine centralized control system according to claim 1, characterized in that: The GPT model is ChatGLM-6B, which has been fine-tuned with a mining-specific corpus and has functions such as equipment failure prediction, energy consumption analysis, and scheduling optimization.
7. The intelligent mine centralized control system according to claim 1, characterized in that: The model update module (6) is equipped with a physical one-way import gate, which allows only data writing and blocks data back transmission during temporary encrypted network upgrades, ensuring physical isolation of the internal network.