Coal mine underground video intelligent monitoring system

Abstract

The invention relates to a coal mine underground video intelligent monitoring system, and aims to improve the real-time performance, intelligence and automation level of coal mine safety production. The system performs comprehensive monitoring and real-time analysis on each monitoring area of a coal mine by integrating a high-definition joint control camera, a laser radar, an AI algorithm model and an efficient server architecture. The system can automatically identify potential safety hazards such as dangerous areas, equipment faults and personnel violation behaviors, and triggers corresponding alarm and emergency response measures. Compared with a traditional manual monitoring system, the AI technology can remarkably improve the accuracy and response speed of danger identification, it is ensured that coal mine managers can take precautionary measures in time, and accidents are reduced. In addition, the system also has data analysis and production management functions, can provide decision support for coal mine safety management, and optimizes the production process. Through the system, intelligent monitoring, automatic alarm and efficient data management of a coal mine are realized, and the overall safety and the production efficiency are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of coal mine safety monitoring, and in particular to an underground coal mine video intelligent monitoring system. Background Art

[0002] Coal mining, a high-risk industry with complex working environments, presents numerous challenges for safe production. Traditional coal mine safety monitoring methods rely on manual inspections and simple camera surveillance systems, which are unable to meet the increasingly complex safety management needs. While existing monitoring equipment can provide basic video surveillance capabilities, due to the large number of operators, complex working environments, and widely distributed equipment, traditional systems suffer from incomplete data collection, untimely alarms, and insufficient monitoring coverage. This is particularly true in high-risk areas such as transportation systems, drilling sites, and tunneling faces, where traditional technologies are unable to conduct real-time analysis and early warning, delaying the implementation of safety measures and posing significant safety risks.

[0003] Therefore, the coal mining industry urgently needs to introduce advanced artificial intelligence (AI) technology to improve the intelligence level of the monitoring system, realize timely detection and processing of various dangerous events in the coal mine production process, and ensure the safe and stable operation of coal mines.

[0004] In recent years, the rapid development of artificial intelligence (AI) has brought new solutions to coal mine safety monitoring. AI technologies, particularly computer vision, deep learning, and pattern recognition, offer significant advantages in target detection, behavior analysis, and image semantic segmentation. AI algorithms enable efficient real-time monitoring and automated early warning of dangerous behaviors, equipment failures, and abnormal events in coal mines.

[0005] While some mining areas have begun to introduce AI technology, most systems are still in the experimental or limited application phase, focusing on basic aspects such as video surveillance and behavior recognition. Existing technologies still have limitations in intelligent analysis and coordinated control in complex environments. Furthermore, many systems suffer from poor device connectivity and insufficient data processing capabilities, making it difficult to accurately identify and respond to complex scenarios. Therefore, improving the accuracy of AI models, enhancing the system's intelligent reasoning capabilities, and increasing the efficiency of collaborative work between devices have become pressing challenges in coal mine safety monitoring.

[0006] This project proposes an AI-based coal mine safety monitoring system, designed to provide intelligent assurance for safe coal mine production through the comprehensive integration of advanced hardware, AI algorithms, and an efficient network architecture. This system not only enables real-time monitoring of different areas and types of equipment in coal mines, but also intelligently analyzes image and video data through deep learning algorithms, automatically identifying potential safety hazards and triggering appropriate alarms and emergency response measures. By combining hardware devices such as AI inference servers, streaming media servers, high-definition linked control cameras, and lidar, the system enables real-time and efficient monitoring in complex coal mine environments, significantly improving the timeliness of safety warnings and emergency response.

[0007] The system is also highly scalable, enabling customized functional modules tailored to the specific needs of coal mines. For example, it optimizes hazardous areas on scraper conveyors, ensures the wearing of protective gear for personnel, and identifies abnormal equipment behavior. By combining multiple monitoring methods with AI-powered intelligent analysis, the system not only enhances overall mine safety but also provides data support and decision-making for intelligent mine management and improved production efficiency. Summary of the Invention

[0008] The present invention provides an intelligent underground video monitoring system for coal mines. By integrating AI algorithms, intelligent hardware devices and efficient network communication technologies, it realizes automated monitoring, behavior analysis, equipment fault detection and alarm linkage functions for coal mine production safety, effectively improving the safety and efficiency of coal mine production.

[0009] An underground video intelligent monitoring system for a coal mine, comprising:

[0010] AI training servers, used to train AI algorithm models, have powerful computing capabilities, are equipped with multiple CPUs, memory and storage devices, and support high-speed network communication interfaces;

[0011] AI inference server, used for inference processing of trained AI algorithm models, equipped with high-performance CPU and AI accelerator card, supports real-time data processing and inference;

[0012] Streaming media server, used to receive and store real-time video data from the coal mine site and perform video stream processing;

[0013] Surveillance cameras, including linked control cameras, high-speed cameras, dust-proof cameras, and lidar cameras, are used for real-time monitoring and data collection in different areas of the coal mine;

[0014] Storage devices, including hard disk recorders, for storing video data and alarm information;

[0015] Network equipment, including 10G aggregation switches, is used to connect servers and monitoring equipment to ensure high-speed and stable data transmission;

[0016] AI algorithm models, covering target detection, behavior recognition, and image semantic segmentation algorithm models, can be customized for various safety monitoring needs of coal mines;

[0017] AI application platform software, including monitoring, alarm, video analysis, device management, and model management functional modules, supports access to multiple monitoring devices and alarm linkage.

[0018] Furthermore, the surveillance camera equipment includes:

[0019] Joint control cameras are used to monitor different areas of the coal mine in real time;

[0020] High-speed cameras, used to capture fast-moving or changing scenes;

[0021] Dust-proof camera, used for high-definition monitoring in coal mine environments where coal dust exists, with fill light and dust removal functions;

[0022] LiDAR cameras are used to identify ultra-high and ultra-wide transport materials and are installed on the top of the tunnel.

[0023] Furthermore, the AI application platform software includes:

[0024] Alarm management module, used to set alarm thresholds and implement multiple alarm modes, such as sound and light alarms, and SMS alarms;

[0025] Video analysis module, used to analyze real-time video data and identify targets and behaviors;

[0026] Equipment management module, used for real-time monitoring and management of various equipment in coal mines;

[0027] Model management module, used to deploy, update and optimize AI algorithm models;

[0028] The statistical dashboard module is used to display the coal mine safety production status, equipment status and alarm statistics in real time.

[0029] Furthermore, the AI algorithm model includes:

[0030] Target detection algorithms for identifying important equipment, personnel, and behaviors in coal mines;

[0031] Behavior recognition algorithms are used to identify the behavior of coal mine workers and determine whether there are any violations;

[0032] Image semantic segmentation algorithm is used to perform pixel-level analysis and segmentation of coal mine monitoring videos and identify specific objects in the area.

[0033] Furthermore, the streaming media server has:

[0034] Data storage function, used to store coal mine monitoring video data, supporting large-scale video data storage and management;

[0035] The data stream processing function can process video streams in real time and supports multiple video encoding formats, including H.264 and H.265.

[0036] Furthermore, the storage device includes:

[0037] Hard disk recorder, used to store video data and alarm records from camera equipment, and equipped with RAID redundancy protection function;

[0038] Storage management function supports regular cleanup of expired data and supports backup and recovery strategies.

[0039] Furthermore, network equipment includes:

[0040] 10G aggregation switch is used to ensure high-speed and stable data transmission between servers and camera equipment;

[0041] Gigabit network interface, used to connect various devices and servers to the network and ensure stable and low latency data transmission.

[0042] Furthermore, the AI application platform supports alarm linkage functions including:

[0043] Equipment control linkage, such as automatic shutdown or activation of emergency equipment when personnel enter a hazardous area;

[0044] Safety and security linkage, such as automatically initiating on-site broadcasting or other safety and security measures when unsafe behavior is detected.

[0045] Furthermore, the server equipment is equipped with redundant power supplies and fans to ensure stability and high reliability during long-term operation.

[0046] Beneficial effects of the present invention:

[0047] 1. The AI-based coal mine safety monitoring system can analyze and process data collected from various monitoring devices (such as high-definition cameras and lidar) in real time, and perform rapid reasoning and judgment through intelligent algorithms. This enables the system to promptly identify potential hazards in coal mines, such as equipment failures, personnel violations, or other abnormal situations. Compared with traditional manual monitoring systems, AI systems have significant advantages in recognition speed and accuracy, which can greatly improve the timeliness of safety warnings and prevent accidents from occurring or further deteriorating. Through automated detection and real-time alarms, coal mine managers can take emergency measures immediately to protect the lives of miners and reduce the incidence of mine accidents.

[0048] Second, the system integrates multiple AI algorithm models, including target detection, behavior recognition, and image semantic segmentation, and can autonomously identify key equipment, personnel behavior, and other potential risk factors in coal mines. Unlike traditional monitoring systems, AI systems are capable of more sophisticated intelligent analysis and decision-making, not only identifying common safety hazards but also responding to complex environmental changes and emergencies. For example, in the dangerous area of a scraper conveyor, the system can automatically identify and determine whether there is a human intrusion, immediately issue an alarm signal, and even link the control system to an emergency stop. Through intelligent processing, the system greatly reduces the frequency of manual intervention and the possibility of misoperation, and improves the level of automation of safety management in the coal mine production process.

[0049] 3. In addition to improving safety and intelligent monitoring, AI-based coal mine safety monitoring systems can also provide powerful decision-making support for coal mine production management. By collecting and analyzing large amounts of monitoring data, the system can help coal mine managers gain a deeper understanding of potential problems and patterns in the production process. For example, by analyzing personnel behavior data, the system can identify high-incidence periods and high-risk areas for violations, providing a data basis for coal mine safety management. In addition, through functions such as data visualization and statistical dashboards, the system can help managers monitor the operating status of various equipment in real time, promptly detect equipment failures or performance degradation, support equipment maintenance and preventive maintenance in coal mines, and thus optimize the production efficiency and operational management of the entire coal mine. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0051] Figure 1 Schematic diagram of the system architecture of an embodiment of the present invention;

[0052] Figure 2 A schematic diagram of the equipment installation and wiring process according to an embodiment of the present invention;

[0053] Figure 3 A schematic diagram of the system configuration and debugging process according to an embodiment of the present invention;

[0054] Figure 4 A schematic diagram of the system acceptance and delivery process of an embodiment of the present invention;

[0055] Figure 5 A schematic diagram of the system operation and maintenance process according to an embodiment of the present invention;

[0056] Figure 6 Schematic diagram of the AI model update and optimization process according to an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. It is also noted that, to provide a more detailed description, the following embodiments are best and preferred embodiments, and those skilled in the art may employ alternative methods for implementing certain known technologies. Furthermore, the accompanying drawings are intended only to provide a more detailed description of the embodiments and are not intended to limit the present invention.

[0058] It should be noted that references in the specification to "one embodiment," "an embodiment," "an exemplary embodiment," "some embodiments," etc. indicate that the described embodiments may include specific features, structures, or characteristics, but not every embodiment necessarily includes such specific features, structures, or characteristics. In addition, when specific features, structures, or characteristics are described in conjunction with an embodiment, it is within the knowledge of persons skilled in the relevant art to implement such features, structures, or characteristics in conjunction with other embodiments (whether or not explicitly described).

[0059] In general, terms can be understood, at least in part, from their use in context. For example, depending at least in part on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending at least in part on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0060] Example 1

[0061] 1. Required Materials and Equipment

[0062] (1) Hardware equipment

[0063] Server Class

[0064] AI training server: Powerleader PR420KIG2 training server, with specifications of 4U rack server, has powerful computing power (computing power of 2.5pFP16 for the whole machine). It is equipped with 4 CPUs with 48 cores and a main frequency of 2.6GHz, 32 single 32GB memories (supporting 32 DDR4 memory slots), 2 480GB SSD hard drives and 2 3.84TNVMESSD for data storage. In terms of network, there are 8 200GE interfaces (including 100G modules) and 4 10GE interfaces (including modules) to ensure high-speed data transmission. Equipped with 8 AI chips, the total chip memory is 256GBHBM, the chip integrates 200Gb interface capability, and there are redundant power supplies and fans to ensure stable operation, and supports Feiteng, Huawei Kunpeng, and Loongson technology routes.

[0065] AI inference server: Powerleader PR210K inference server, equipped with two 2.6GHz CPUs with 64 physical cores each, and 256GB of 3200MHz DDR4 memory. Storage includes two 480GB SATA SSD hard drives and four 2.4TB SAS 2.5-inch hard drives, along with a RAID card supporting RAID 0 / 1 / 10. The network interface is a 4GE electrical port + a 410GE optical port (with a 10G optical module, single-mode dual-fiber), equipped with seven domestically produced AI accelerator cards (single-card computing power of 140TOPS@INT8, single-card memory capacity of 24GB), fully equipped with redundant hot-swappable power supplies (single power supply power of 2000W), and equipped with out-of-band fault detection capabilities.

[0066] Streaming media server: KunLun2280 streaming media server, using two Huawei Kunpeng 920 chips (CPU main frequency 2.6GHz, single physical core count 32 cores), one inference card Atlas300Ipro (cache 24G), 32G*4 memory, 100T SATA hard drive for large-scale video data storage, one RAID card with 2G cache, two 10GE optical ports (with 10G optical modules, single-mode dual-fiber) and one standard guide rail.

[0067] Network equipment: 10G aggregation switch, used to connect various servers and other network devices to build a high-speed and stable network environment.

[0068] Camera equipment

[0069] Joint control cameras: 9 units, used for real-time monitoring of different areas of the coal mine.

[0070] High-speed camera: 1 unit, mainly used to accurately capture some fast-moving or changing scenes, such as tail rope operation monitoring.

[0071] Anjin KBA12(B) dust-proof camera: 1 unit, maximum resolution 4 megapixels, backward compatible with 1080P, frame rate supports 25FPS, bit rate supports adaptive bit rate (default bit rate 4Mbps), supports H264 / H265 video encoding, comes with built-in fill light or can be attached with a dedicated fill light, supports fixed focus, electric focus or auto focus, supports GB / T28181 national standard protocol or ONVIF protocol, RTSP video streaming protocol, has strong light suppression and wide dynamic adjustment functions, has a wiper dust removal function, and includes a matching power supply (2h backup battery) and bracket.

[0072] LiDAR camera: 1 set, used for identifying ultra-high and ultra-wide transport materials, installed on the top of the tunnel, and must ensure a clear bird's-eye view of the carriages in the train.

[0073] Storage device: 64-channel Hikvision hard disk recorder, used to store surveillance video data.

[0074] Cables and accessories

[0075] Mine flame-retardant communication cable: 1500 meters, used to connect underground equipment, ensuring stable communication and flame-retardant properties, in line with coal mine safety requirements.

[0076] Polyethylene insulated braided shielded PVC sheathed communication cable for coal mines: 1000 meters, used for signal transmission, with good insulation and shielding properties.

[0077] PDU cabinet sockets: 4, providing stable power output for servers and other equipment.

[0078] White Category 6 Gigabit network cable (12 meters): 20 sets, used for network connection between devices.

[0079] (2) Software resources

[0080] AI algorithm model: covers a variety of algorithm models such as target detection, behavior recognition, and image semantic segmentation, and is customized, developed, and optimized for different monitoring scenarios (including monitoring of dangerous areas of scraper conveyors and detection of personnel wearing labor protection equipment).

[0081] Basic software: including an operating system suitable for the server (Linux system can be used) and a database management system (MySQL can be used), providing basic support for the operation of the AI application platform.

[0082] AI application platform software: includes seven subsystems: AI-assisted coal flow transportation, drilling site monitoring, electromechanical system, auxiliary transportation, tunneling system, security system and camera anomaly detection, as well as functional modules such as alarm management, video analysis, equipment management, model management, statistical dashboard, configuration management, and permission management.

[0083] 2. Specific operation steps

[0084] like Figure 1 As shown in the figure, the system architecture shows the configuration and communication relationship of each hardware device and software module:

[0085] (1) Preliminary planning and design

[0086] On-site investigation

[0087] Professional and technical personnel were organized to conduct a comprehensive survey of all underground areas of the Wugou Coal Mine, including coal flow transportation routes, drilling sites, electromechanical equipment installation points, auxiliary transportation tunnels, excavation working faces, key security areas, etc.

[0088] Record the environmental conditions of each area in detail, such as light intensity, coal dust concentration, humidity, temperature, etc.; measure the spatial dimensions, equipment layout and possible installation locations of each area; and evaluate the feasibility and path of network cabling.

[0089] Demand Analysis

[0090] Communicate with coal mine managers and front-line workers to understand their functional requirements for the AI application platform, such as the equipment and behaviors they want to monitor, alarm methods, and threshold settings.

[0091] Based on the production scale, process flow and safety requirements of the coal mine, determine the specific goals and functions that the AI application platform needs to achieve.

[0092] System Design

[0093] Based on the results of on-site investigation and demand analysis, the overall architecture of the AI application platform is designed, including the selection and layout of hardware equipment, and the module division and functional design of the software system.

[0094] Plan the network topology, determine the connection method and communication protocol between servers, cameras, switches and other devices, and ensure the stability and reliability of the network.

[0095] Develop data storage and management plans, determine the storage methods and storage cycles for video data, alarm data, etc., as well as data backup and recovery strategies.

[0096] like Figure 2 The device installation and cabling process describes the installation steps and connection methods for cameras, servers, and storage devices:

[0097] (2) Equipment installation and wiring

[0098] Server Installation

[0099] In the coal mine computer room, AI training servers, AI inference servers, and streaming media servers are installed in standard cabinets according to the design plan and fixed with guide rails to ensure that the servers are firmly installed.

[0100] Connect the power cord of the server and check whether the power supply is normal. Ensure that the redundant power supply can work normally and provide stable power support for the server.

[0101] Use a network cable to connect the server's network interface to the 10 Gigabit aggregation switch. Configure the network according to the server's IP address plan to ensure that the server can access the network normally.

[0102] Camera equipment installation

[0103] Select appropriate cameras for installation based on different monitoring needs and scenarios. For example, a high-definition linked control camera can be installed at the nose and tail rotating parts of a scraper conveyor, a flameproof and intrinsically safe camera and fill light can be installed directly above the belt, and a lidar camera can be installed at the top of the tunnel.

[0104] Determine the camera's installation position and angle to ensure it covers the target monitoring area. The installation height is generally adjusted based on the monitoring range and target size. For example, for human behavior monitoring, the camera can be installed at a height of 2-3 meters; for belt monitoring, the installation height should ensure a clear view of the entire belt.

[0105] Use a bracket to fix the camera in the installation location to ensure that it is firmly installed to prevent the camera from shifting or shaking due to vibration or other factors.

[0106] Connect the camera's power cord and network cable, connect the power cord to the supporting power supply (for cameras with backup batteries, ensure that the backup battery is charged normally), and connect the network cable to the corresponding port of the switch to check whether the camera is powered on and connected to the network normally.

[0107] Network cabling

[0108] Lay the flame-retardant communication cables for mines and the polyethylene insulated, braided, shielded, and polyvinyl chloride sheathed communication cables for coal mines according to the design plan. The cables should be laid along the tunnel walls or cable trays to avoid interference with other equipment or pipelines.

[0109] During cable laying, pay attention to the cable bending radius and fixed spacing to ensure that the cable is not damaged. For example, the bending radius of the flame-retardant communication cable for mining should be no less than 15 times the outer diameter of the cable, and the fixed spacing should not be too large to prevent the cable from sagging.

[0110] Use white Category 6 Gigabit network cables to connect cameras, servers and other devices to the switch, ensuring that the cables are firmly connected to avoid looseness or poor contact.

[0111] Storage device installation

[0112] Install the 64-channel Hikvision DVR in a suitable location in the equipment room, connect the power cord and network cable, and ensure that it can work normally and access the network.

[0113] Install the hard disk into the DVR, configure it reasonably according to storage requirements and hard disk capacity, and set the storage path and storage period of the video data.

[0114] like Figure 3 The flowchart of system configuration and debugging shows how to configure the server, AI application platform, and system debugging:

[0115] (3) System configuration and debugging

[0116] Server Configuration

[0117] Install an operating system for the server and select an appropriate operating system version based on the server's hardware configuration and software requirements, such as a specific Linux distribution.

[0118] Install the drivers required by the server, such as the AI accelerator card driver and network card driver, to ensure that the hardware devices can work properly.

[0119] Configure the server's network parameters, including IP address, subnet mask, gateway, DNS, etc., to ensure that the server can communicate normally with other devices.

[0120] Install database management systems and related service software to provide data storage and management support for the operation of the AI application platform.

[0121] AI application platform configuration

[0122] Deploy AI application platform software, install the developed software system on the server, and perform initial configuration.

[0123] Import various AI algorithm models and select appropriate models for deployment and configuration based on different monitoring scenarios and requirements. For example, for monitoring hazardous areas on scraper conveyors, deploy a model based on target detection algorithms; for identifying human behavior, deploy a behavior recognition algorithm model.

[0124] Configure functional modules such as alarm management, video analysis, device management, model management, statistical dashboards, configuration management, and permission management on the platform. Set parameters such as alarm thresholds, alarm methods (such as audible and visual alarms, SMS alarms, etc.), device linkage rules (such as shutting down devices or activating emergency lighting when an anomaly is detected), and user permission levels.

[0125] Configure the edge camera video stream and application rules for model inference. Users can define the target detection range, corresponding events, and linkage rules based on actual conditions.

[0126] System debugging

[0127] Perform a comprehensive system debug to check the device's operating status and network communications. Use the server's management interface and monitoring software to check the server's CPU, memory, hard disk usage, and other indicators to ensure normal server operation.

[0128] Check the camera's video image quality and adjust the camera's focus, aperture, brightness and other parameters to ensure that the video image is clear, without blur, distortion, backlight and other problems.

[0129] Test the equipment linkage function, simulating various abnormal situations, such as human intrusion and equipment failure, to check whether the system can accurately identify and trigger the corresponding alarms and linkage operations. For example, when simulating a human intrusion into the dangerous area of a scraper conveyor, check whether the underground broadcasting equipment can promptly issue a voice alarm and whether the scraper conveyor control system can normally stop the equipment.

[0130] Optimize and adjust the system based on the debugging results. If the recognition accuracy of a certain monitoring area is low, adjust the camera position, angle, or model parameters in that area. If the device linkage response time is too long, check for network latency or device control program issues and optimize them.

[0131] like Figure 4 As shown in the figure, the flowchart of system acceptance and delivery shows the detailed steps of functional acceptance, performance acceptance and document delivery:

[0132] (IV) System acceptance and delivery

[0133] Functional acceptance

[0134] Conduct comprehensive testing and acceptance of all functions of the AI application platform in accordance with the system design plan and contract requirements. Check whether the alarm management function can accurately record and push alarm information, whether the video analysis function can correctly identify targets and behaviors, whether the device management function can effectively monitor and manage devices, and whether the model management function can deploy, update, and optimize models.

[0135] Verify whether the system's equipment linkage function is normal, and whether the relevant equipment can perform linkage operations according to the preset rules when an abnormal situation occurs.

[0136] Performance acceptance

[0137] Test and evaluate the system's performance indicators, including video image clarity, frame rate, bit rate, system response time, processing capacity, model recognition accuracy, recall rate, etc. Ensure that the system's performance indicators meet design requirements and industry standards.

[0138] Test the stability of the system under high concurrency conditions, simulate the situation where a large number of users access the system at the same time and multiple monitoring points generate alarms at the same time, and check whether the system can operate normally without crashes or data loss.

[0139] Document delivery

[0140] Deliver system-related documents to the coal mine, including system design documents, operation manuals, maintenance manuals, test reports, training materials, etc., to ensure that coal mine staff can fully understand the system's functions, operation methods, and maintenance points.

[0141] Training and Delivery

[0142] System training is provided to coal mine management, operators, and maintenance personnel. The training content includes system function introduction, operation methods, daily maintenance, troubleshooting, etc. Through theoretical explanations and practical operation demonstrations, the trainees can master the use and maintenance skills of the system.

[0143] After completing the delivery of the system, the system will be officially handed over to the coal mine for use, and technical support and after-sales service will be provided for a certain period of time.

[0144] like Figure 5 The flowchart of system operation and maintenance describes how to perform daily monitoring, data management and analysis, and regular maintenance of equipment:

[0145] (V) System operation and maintenance

[0146] Daily operation monitoring

[0147] Assign dedicated personnel to be responsible for the daily operation and monitoring of the system. Through the statistical dashboard and monitoring interface of the AI application platform, real-time information such as the coal mine safety production status, access equipment statistics, various alarm statistics, and statistical analysis of alarm handling results can be obtained.

[0148] Regularly check the server's operating status and performance indicators, such as CPU usage, memory usage, hard disk read and write speed, etc., to promptly identify potential server problems.

[0149] Check the working status of the camera to see if the video image is normal and whether there is any blur, freeze, or loss of the image.

[0150] Data management and analysis

[0151] Regularly clear outdated video and alarm data to free up server storage space, and back up important data to prevent data loss. Backup data can be stored on external storage devices or cloud storage to ensure data security and recoverability.

[0152] Analyze and mine the vast amounts of data generated during system operation. Using data visualization tools and data analysis algorithms, we can identify potential problems and patterns in coal mine production, providing support for safe production and management decisions. For example, we can analyze personnel behavior data to identify peak periods and areas for violations, enabling targeted improvements.

[0153] Equipment maintenance and care

[0154] Clean and maintain the camera regularly, check whether the lens is covered with coal dust and whether the wiper dust removal function is normal; check whether the camera installation position is loose, and adjust it in time if there is any problem.

[0155] Perform hardware inspection and maintenance on the server, including cleaning dust inside the server, checking whether the hardware device connections are secure, and replacing aging or damaged hardware components.

[0156] Check the operating status of network devices, such as switch port status and network bandwidth usage, to ensure smooth network operation. Address network faults such as network outages and packet loss promptly.

[0157] like Figure 6 As shown in the figure, the flowchart of AI model update and optimization shows how to optimize the model based on operating data and finally deploy it:

[0158] Model update and optimization

[0159] Collect abnormal data and misjudgment data generated during system operation, label and analyze these data, and upload the labeled data to the AI training center.

[0160] Update and optimize AI algorithm models based on new data and business needs to improve their recognition accuracy and adaptability. Use methods such as incremental learning and transfer learning to train and adjust models without impacting the normal operation of the system.

[0161] Regularly test and validate the updated model to ensure improved performance, and then deploy the optimized model to the system.

[0162] The present invention encompasses any alternatives, modifications, equivalents, and solutions that fall within the spirit and scope of the present invention. To provide a thorough understanding of the present invention, specific details are described in detail below in connection with the preferred embodiments of the present invention, but those skilled in the art will be able to fully understand the present invention without these detailed descriptions. Furthermore, to avoid unnecessary confusion regarding the essence of the present invention, well-known methods, processes, procedures, components, and circuits have not been described in detail.

[0163] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A coal mine underground video intelligent monitoring system, characterized in that: include: AI training servers, used to train AI algorithm models, have powerful computing capabilities, are equipped with multiple CPUs, memory and storage devices, and support high-speed network communication interfaces; AI inference server, used for inference processing of trained AI algorithm models, equipped with high-performance CPU and AI accelerator card, supports real-time data processing and inference; Streaming media server, used to receive and store real-time video data from the coal mine site and perform video stream processing; Surveillance cameras, including linked control cameras, high-speed cameras, dust-proof cameras, and lidar cameras, are used for real-time monitoring and data collection in different areas of the coal mine; Storage devices, including hard disk recorders, for storing video data and alarm information; Network equipment, including 10G aggregation switches, is used to connect servers and monitoring equipment to ensure high-speed and stable data transmission; AI algorithm models, covering target detection, behavior recognition, and image semantic segmentation algorithm models, can be customized for various safety monitoring needs of coal mines; AI application platform software, including monitoring, alarm, video analysis, device management, and model management functional modules, supports monitoring device access and alarm linkage.

2. The coal mine underground video intelligent monitoring system according to claim 1, characterized in that: The monitoring camera equipment includes: Joint control cameras are used to monitor different areas of the coal mine in real time; High-speed cameras, used to capture fast-moving or changing scenes; Dust-proof camera, used for high-definition monitoring in coal mine environments where coal dust exists, with fill light and dust removal functions; LiDAR cameras are used to identify ultra-high and ultra-wide transport materials and are installed on the top of the tunnel.

3. The coal mine underground video intelligent monitoring system according to claim 1, characterized in that: The AI application platform software includes: Alarm management module, used to set alarm thresholds and implement multiple alarm modes, such as sound and light alarms, and SMS alarms; Video analysis module, used to analyze real-time video data and identify targets and behaviors; Equipment management module, used for real-time monitoring and management of various equipment in coal mines; Model management module, used to deploy, update and optimize AI algorithm models; The statistical dashboard module is used to display the coal mine safety production status, equipment status and alarm statistics in real time.

4. The coal mine underground video intelligent monitoring system according to claim 1, characterized in that: The AI algorithm model includes: Target detection algorithms for identifying important equipment, personnel, and behaviors in coal mines; Behavior recognition algorithms are used to identify the behavior of coal mine workers and determine whether there are any violations; Image semantic segmentation algorithm is used to perform pixel-level analysis and segmentation of coal mine monitoring videos and identify specific objects in the area.

5. The coal mine underground video intelligent monitoring system according to claim 1, characterized in that: The streaming media server has: Data storage function, used to store coal mine monitoring video data, supporting large-scale video data storage and management; The data stream processing function can process video streams in real time and supports multiple video encoding formats, including H.264 and H.

265.

6. The coal mine underground video intelligent monitoring system according to claim 1, characterized in that: The storage device includes: Hard disk recorder, used to store video data and alarm records from camera equipment, and equipped with RAID redundancy protection function; Storage management function supports regular cleanup of expired data and supports backup and recovery strategies.

7. The coal mine underground video intelligent monitoring system according to claim 1, characterized in that: The network equipment includes: 10G aggregation switch is used to ensure high-speed and stable data transmission between servers and camera equipment; Gigabit network interface, used to connect various devices and servers to the network and ensure stable and low latency data transmission.

8. The coal mine underground video intelligent monitoring system according to claim 1, characterized in that: The alarm linkage functions supported by the AI application platform include: Equipment control linkage, such as automatic shutdown or activation of emergency equipment when personnel enter a hazardous area; Safety and security linkage, such as automatically initiating on-site broadcasting or other safety and security measures when unsafe behavior is detected.

9. The coal mine underground video intelligent monitoring system according to claim 1, characterized in that: The server equipment is configured with redundant power supplies and fans to ensure stability and high reliability during long-term operation.

Images