A full-stack domestication intelligent ventilation system and method based on a mining honghong

Abstract

The application discloses a full-stack domestic intelligent ventilation system and method based on a mine Hongmeng, integrates domestic high-precision sensors and actuators in a perception layer, arranges 5G / LoRa dual-mode communication substation in a network layer, arranges a domestic ARM architecture edge computing node, a mine Hongmeng operating system, a ventilation network solving engine and an AI decision model in a platform layer, stores a domestic GIS / BIM engine and a three-dimensional digital twin visualization system in an application layer, and connects each layer through a mine Hongmeng HCP / MDTP unified protocol. According to mine ventilation network dynamic parameters and multi-source risk evolution characteristics, the application can maximize the response speed and control precision of the ventilation system to gas abnormalities, fire smoke and other sudden conditions through accurate collection of the perception layer, redundant transmission of the network layer, real-time solving and AI decision of the platform layer and three-dimensional visualization cooperation of the application layer, so as to provide decision and support for deep coal mine safety production.

Description

Technical Field

[0001] This invention relates to the field of intelligent coal mine technology, specifically to a fully domestically produced intelligent ventilation system and method based on HarmonyOS for mining. Background Technology

[0002] With the advancement of intelligent coal mine construction, mine ventilation systems are gradually evolving from traditional manual operation and simple automatic control towards informatization and automation. my country's coal mining has entered a stage of coordinated development of deeper and more intelligent operations. Mine vertical depth continues to increase, ventilation network topologies are becoming longer, wind resistance distribution is becoming more uneven, and the dynamics of gas outbursts are increasing. Most mine ventilation systems still adopt traditional centralized or simple distributed architectures, resulting in inconsistent sensing accuracy and communication reliability, low diversity of control strategies, unreasonable matching between airflow distribution models and actual working conditions, and low overall system response efficiency. Therefore, how to achieve an intelligent transformation from passive response to proactive prediction, closed-loop control, and disaster self-healing, and comprehensively improve the safety, robustness, and autonomous controllability of ventilation systems, has become an urgent problem to be solved in current coal mine safety production.

[0003] Currently, common intelligent ventilation systems are generally designed according to a three-level architecture of sensor + PLC + host computer. Their data processing delay is much greater than the dynamic response time of mine airflow. In addition, the poor protocol compatibility of multi-source heterogeneous equipment and the weak support capabilities of domestic software and hardware make the system integration more complex and the safety risk prevention and control capabilities more prominent.

[0004] Currently, there are three main methods for constructing intelligent ventilation systems: The first is a ventilation monitoring system based on a general industrial IoT platform, which uses third-party sensors and commercial SCADA software to collect basic parameters and perform remote monitoring. However, this approach lacks a dedicated calculation engine and AI decision-making model tailored to the characteristics of mine ventilation networks, making it unable to dynamically predict airflow demand and simulate disturbance propagation paths, thus hindering rapid wind control decisions under disaster conditions. The second method is a ventilation visualization system developed using foreign digital twin platforms, employing commercial GIS engines and cloud-based AI services to construct 3D displays and trend analysis functions. Due to the highly dynamic nature of mine environmental parameters, airflow conditions cannot remain constant; gas migration and network disturbances are constantly changing. Furthermore, the update cycle of the cloud model and the execution delay at the edge are inconsistent, inevitably leading to delayed warnings and inaccurate control. The third method is a local ventilation control system based on domestically produced embedded controllers, which starts with single-point fan speed regulation to achieve coarse airflow adjustment. This method does not consider the impact of network-wide air pressure balance constraints when implementing multi-node collaborative wind control. Meanwhile, there are no clear criteria for identifying key risk conditions such as abnormal gas flow and fire smoke flow, and no mention is made of the emergency strategy generation mechanism. Summary of the Invention

[0005] The purpose of this invention is to provide a fully domestically produced intelligent ventilation system and method based on HarmonyOS for mining. Based on the complex ventilation network structure and multi-source dynamic risk coupling characteristics of deep mines, this invention utilizes a complete set of technologies that organically combine high-precision multi-parameter sensing at the perception layer, dual-mode redundant communication at the network layer, real-time edge computing and AI decision-making at the platform layer, and 3D digital twin visualization at the application layer. This addresses the problems of low perception accuracy, poor communication reliability, insufficient platform intelligence, weak application layer interaction capabilities, and security and supply chain risks caused by the reliance on foreign software and hardware in the entire system.

[0006] To achieve the above objectives, the present invention provides the following technical solution: This invention provides a fully domestically produced intelligent ventilation system based on the HarmonyOS operating system for mining. It adopts a four-layer distributed architecture of perception, network, platform, and application, and is equipped with domestically produced hardware and software throughout the entire chain. Specifically, it includes: a perception layer, deploying domestically produced high-precision sensors and actuators. The domestically produced high-precision sensors are used to collect parameters such as wind speed, wind pressure, gas concentration, temperature, humidity, and atmospheric pressure in real time. The actuators include intelligent air doors / windows and variable frequency fan controllers; a network layer, integrating 5G, industrial Ethernet, and LoRa wireless communication technologies. All devices are connected to the HarmonyOS soft bus through the HarmonyOS HCP / MDTP unified protocol, constructing a dual-mode redundant communication transmission channel; a platform layer, based on domestically produced ARM architecture high-performance edge computing nodes, running the HarmonyOS operating system for mining, integrating a self-developed ventilation network calculation engine and AI decision model, supporting localized real-time analysis and closed-loop control; and an application layer, based on a domestically produced GIS / BIM engine, constructing a three-dimensional digital twin visualization system to realize dynamic display of ventilation status, remote air adjustment, and disaster simulation operation.

[0007] Furthermore, the domestically produced high-precision sensor includes a wind speed sensor and a multi-parameter sensor. The wind speed sensor adopts a four-probe cross array layout, with an IP68 protection rating for the housing, a dustproof and waterproof nano-coating on the surface, a built-in temperature compensation module, and an edge-side adaptive filtering algorithm, achieving a measurement error of ≤±0.1m / s. It supports hot-swapping and remote firmware upgrades. The multi-parameter sensor integrates methane, CO, temperature, humidity, and air pressure detection units, covering an atmospheric pressure range of 80.00–120.00kPa, a temperature range of 0–50℃, and a humidity range of 20–95%RH. It communicates via an RS485 interface with a transmission distance of ≥2km at a communication speed of 2400bps.

[0008] Furthermore, the intelligent damper / window adopts a dual-motor redundant structure, driven by a domestic PLC controller. The damper shaft is equipped with a high-precision magnetic encoder with a resolution of 0.1°, forming a closed-loop feedback control system with a minimum adjustment accuracy of 0.5°. The intelligent damper / window is connected to the mining HarmonyOS system through a standardized communication interface, and status information is reported at a 100ms cycle, with an end-to-end delay of control commands ≤50ms. The damper body adopts a Q345 steel frame and composite fireproof and heat-insulating board splicing structure, with an air leakage rate of <3% when closed.

[0009] Furthermore, the communication substations of the network layer support dual-mode redundant networking of 5G and LoRa, and automatically switch to LoRa Mesh self-organizing network mode in areas with weak signals; all transmission links adopt the national cryptographic SM4 encryption algorithm to achieve end-to-end data security protection and prevent data theft or tampering.

[0010] Furthermore, the ventilation network calculation engine of the platform layer adopts an improved Ward-McBride numerical method, completing an iterative calculation of the entire mine's air volume distribution every 30 seconds, with a calculation error of ≤2%. The AI ​​decision model integrates a spatiotemporal convolutional neural network (ST-CNN) and a deep belief network (DBN), and is trained using historical operating data and typical accident cases to achieve air volume demand prediction and gas anomaly trend identification within the next 15 minutes, with an early warning accuracy of ≥92% and a response time of <10 seconds. The AI ​​decision model is deployed on the edge platform, supporting on-demand invocation by multiple services, and ensuring the real-time execution of critical tasks through a priority scheduling mechanism.

[0011] Furthermore, the edge computing nodes of the platform layer use domestically produced ARM-A72 quad-core processors, run the Mining Hong operating system and a domestically produced real-time database, adopt an incremental data upload mechanism, and only transmit data packets whose changes exceed a threshold. The network load is optimized through local caching and cloud synchronization strategies.

[0012] Furthermore, the application layer's three-dimensional digital twin visualization system dynamically renders wind speed vector fields, gas concentration cloud maps, and personnel location information, and supports hierarchical permission management; authorized users can initiate one-click reverse wind operations through mobile or PC terminals, and the command automatically verifies safety conditions and executes preset action sequences after dual authentication.

[0013] Furthermore, the system pre-sets at least six emergency scenario templates, including fire, gas leak, and roof collapse / blockage. When an abnormal operating condition is detected, the system automatically activates the emergency plan, combines the current ventilation network topology and real-time parameters, performs airflow disturbance simulation using the improved Ward-McBride numerical method, evaluates various airflow control strategies, generates the optimal solution, and triggers a sequence of actions such as main fan reversal, damper locking, and fan speed adjustment to complete the emergency airflow control response within 5 minutes.

[0014] Furthermore, the fully domestically produced hardware and software includes domestically produced chips, PLCs, MCUs, operating systems, GIS / BIM engines, and real-time databases, without dependence on foreign technologies, achieving independent control over core technologies.

[0015] This invention provides another technical solution: a control method for a fully domestically produced intelligent ventilation system based on HarmonyOS for mining, comprising the following steps: S1, real-time acquisition of wind speed, wind pressure, gas concentration, temperature, humidity and atmospheric pressure parameters through domestically produced high-precision sensors in the sensing layer; S2, uploading the acquired data to the edge computing node of the platform layer via a dual-mode redundant communication channel in the network layer, and completing unified equipment management through the HarmonyOS HCP / MDTP protocol; S3, the platform layer performs iterative calculation of the air volume distribution of the entire mine based on the improved Ward-McBride numerical method, and calls the AI ​​decision model to predict air volume demand and identify abnormal gas trends; S4, the application layer three-dimensional digital twin system receives the output results of the platform layer, dynamically renders the ventilation status, and generates a wind control strategy according to a preset emergency template or manual instructions, linking the actuator to complete the damper adjustment, fan speed adjustment or main fan reversal action.

[0016] Compared with the prior art, the beneficial effects of the present invention are: ① The present invention provides a fully domestically produced intelligent ventilation system based on HarmonyOS for mining, which can be equipped with a four-probe cross array wind speed sensor and a multi-parameter integrated sensing unit. The protection level and communication parameters can be configured according to the environmental differences of different measuring points in the mine. It can achieve wind speed measurement error ≤ ±0.1m / s and reliable transmission of multi-parameter data over 2km, significantly improving the quality and adaptability of the original data of the sensing layer. ② The present invention provides a fully domestically produced intelligent ventilation system based on HarmonyOS for mining. According to the distribution characteristics of communication blind spots in mines, it can achieve automatic switching of communication interruption and data anti-tampering in complex roadway environments through 5G / LoRa dual-mode redundant networking and SM4 end-to-end encryption mechanism, which greatly enhances the robustness and security of network layer transmission. ③ The present invention provides a fully domestically produced intelligent ventilation system based on HarmonyOS for mining, which can realize iterative calculation of the air volume distribution of the entire network every 30 seconds and AI-driven 15-minute air volume demand prediction. It can dynamically generate air control strategies according to the real-time air network status, and can support early warning response time <10 seconds and air control decision accuracy ≥92%, effectively improving the localized closed-loop control capability of the platform layer. ④ The present invention provides a fully domestically produced intelligent ventilation system based on HarmonyOS for mining, which can build a three-dimensional digital twin platform based on a domestic GIS / BIM engine, realize millisecond-level dynamic rendering of wind speed vector field and gas concentration cloud map, support remote ventilation adjustment with hierarchical permissions and automatic verification of one-click reverse ventilation command, significantly enhance the human-computer interaction efficiency of the application layer and the intuitiveness of disaster response. ⑤ The present invention provides a fully domestically produced intelligent ventilation system based on HarmonyOS for mining, which can realize the full-stack domestic deployment from sensors, controllers, operating systems, databases to visualization engines. It can flexibly replace domestic component models according to the security level requirements of the supply chain, eliminate the technical dependence on imported software and hardware, and provide an inherently safe and sustainable technological foundation for coal mine ventilation systems. Attached Figure Description

[0017] Figure 1 This is a diagram showing the overall system structure of the present invention; Figure 2 This is a flowchart of the disaster emergency response process of the present invention; Figure 3 This is a flowchart of the method of the present invention. Detailed Implementation

[0018] The present invention will now be described in further detail with reference to embodiments. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit the invention.

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0020] Please see Figure 1 This invention provides a fully domestically produced intelligent ventilation system based on the HarmonyOS mining operating system. It adopts a four-layer distributed architecture of perception, network, platform, and application, and is equipped with domestically produced hardware and software across the entire chain. Specifically, it includes: a perception layer deploying domestically produced high-precision sensors and actuators. The high-precision sensors are used to collect parameters such as wind speed, wind pressure, gas concentration, temperature, humidity, and atmospheric pressure in real time. The actuators include intelligent air doors / windows and variable frequency fan controllers; a network layer integrating 5G, industrial Ethernet, and LoRa wireless communication technologies. All devices are connected to the HarmonyOS mining soft bus via the HarmonyOS HCP / MDTP unified protocol, constructing a dual-mode redundant communication transmission channel; a platform layer based on domestically produced ARM architecture high-performance edge computing nodes running the HarmonyOS mining operating system, integrating a self-developed ventilation network calculation engine and AI decision model, supporting localized real-time analysis and closed-loop control; and an application layer based on a domestically produced GIS / BIM engine to construct a three-dimensional digital twin visualization system, enabling dynamic display of ventilation status, remote air adjustment, and disaster simulation operations.

[0021] In this embodiment of the invention, the domestically produced high-precision sensor includes a wind speed sensor and a multi-parameter sensor. The wind speed sensor adopts a four-probe cross array layout, the housing has an IP68 protection rating, the surface is coated with a dustproof and waterproof nano-coating, and it has a built-in temperature compensation module. With the edge-side adaptive filtering algorithm, the measurement error is ≤±0.1m / s, and it supports hot-swapping and remote firmware upgrades. The multi-parameter sensor integrates methane, CO, temperature, humidity, and air pressure detection units, and the measurement range covers atmospheric pressure 80.00–120.00kPa, temperature 0–50°C, and humidity 20–95%RH. It communicates via an RS485 interface, and the transmission distance is ≥2km at a communication speed of 2400bps.

[0022] In this embodiment of the invention, the intelligent damper / window adopts a dual-motor redundant structure, driven by a domestic PLC controller. The damper shaft is equipped with a high-precision magnetic encoder with a resolution of 0.1°, forming a closed-loop feedback control system with a minimum adjustment accuracy of 0.5°. The intelligent damper / window is connected to the mining HarmonyOS system through a standardized communication interface, and status information is reported at a 100ms cycle, with an end-to-end delay of control commands ≤50ms. The damper body adopts a Q345 steel frame and a composite fireproof and heat-insulating board splicing structure, with an air leakage rate of <3% when closed.

[0023] In this embodiment of the invention, the communication substation of the network layer supports dual-mode redundant networking of 5G and LoRa, and automatically switches to LoRa Mesh self-organizing network mode in areas with weak signal; all transmission links adopt the national cryptographic SM4 encryption algorithm to achieve end-to-end data security protection and prevent data theft or tampering.

[0024] In this embodiment of the invention, the ventilation network calculation engine at the platform layer adopts the improved Ward-McBride numerical method, completing an iterative calculation of the air volume distribution of the entire mine every 30 seconds, with a calculation error of ≤2%; the AI ​​decision model integrates a spatiotemporal convolutional neural network (ST-CNN) and a deep belief network (DBN), and is trained using historical operating data and typical accident cases to achieve air volume demand prediction and gas anomaly trend identification within the next 15 minutes, with an early warning accuracy of ≥92% and a response time of <10 seconds; the AI ​​decision model is deployed on the edge platform, supporting on-demand invocation by multiple services, and ensuring the real-time execution of critical tasks through a priority scheduling mechanism.

[0025] In this embodiment of the invention, the edge computing nodes of the platform layer use domestic ARM-A72 quad-core processors, run the Mining Hong operating system and a domestic real-time database, adopt an incremental data upload mechanism, and only transmit data packets whose changes exceed a threshold. The network load is optimized through local caching and cloud synchronization strategies.

[0026] In this embodiment of the invention, the three-dimensional digital twin visualization system at the application layer dynamically renders wind speed vector fields, gas concentration cloud maps, and personnel location information, and supports hierarchical permission management; authorized users can initiate one-click reverse wind operations through mobile or PC terminals, and the command automatically verifies safety conditions and executes a preset action sequence after dual authentication.

[0027] In this embodiment of the invention, the system presets no fewer than six emergency scenario templates, such as fire, gas leak, and roof collapse. When an abnormal operating condition is detected, the system automatically activates the emergency plan, combines the current ventilation network topology and real-time parameters, performs airflow disturbance simulation using the improved Ward-McBride numerical method, evaluates various airflow control strategies, generates the optimal solution, and links a sequence of actions such as main fan reversal, damper locking, and fan speed adjustment to complete the emergency airflow control response within 5 minutes.

[0028] In this embodiment of the invention, the fully domestically produced hardware and software includes domestically produced chips, domestically produced PLCs, domestically produced MCUs, domestically produced operating systems, domestically produced GIS / BIM engines, and domestically produced real-time databases, without dependence on foreign technologies, achieving independent control over core technologies.

[0029] To further illustrate the above embodiments, the present invention also provides: a method for a fully domestically produced intelligent ventilation system based on HarmonyOS for mining, comprising the following steps: Step 1: Real-time acquisition of wind speed, wind pressure, gas concentration, temperature, humidity, and atmospheric pressure parameters via domestically produced high-precision sensors in the sensing layer. The actuator receives control commands from the platform layer to adjust the opening of intelligent dampers / windows and the frequency of the variable frequency fan controller. The wind speed sensor employs a four-probe cross-array layout, with an IP68 protection rating, a dustproof and waterproof nano-coating, and a built-in temperature compensation module. Combined with an edge-side adaptive filtering algorithm, the measurement error is ≤±0.1m / s, and it supports hot-swapping and remote firmware upgrades. The multi-parameter sensor integrates methane, CO, temperature, humidity, and air pressure detection units, covering atmospheric pressure 80.00–120.00kPa, temperature 0–50°C, and humidity 20–95%RH. Communication is via an RS485 interface, with a transmission distance ≥2km at a communication speed of 2400bps.

[0030] Step 2: By integrating 5G, Industrial Ethernet and LoRa wireless communication technologies at the network layer, all devices are connected to the mining HarmonyOS soft bus through the HarmonyOS HCP / MDTP unified protocol to build a dual-mode redundant communication transmission channel. Among them, the communication substation supports 5G and LoRa dual-mode redundant networking, and automatically switches to LoRa Mesh self-organizing network mode in weak signal areas. All transmission links adopt the national cryptographic SM4 encryption algorithm to achieve end-to-end data security protection and prevent data theft or tampering.

[0031] Step 3: Using edge computing nodes based on domestically produced ARM-A72 quad-core processors at the platform layer, a mining-grade HarmonyOS operating system is run, integrating a self-developed ventilation network calculation engine and an AI decision-making model. This supports localized real-time analysis and closed-loop control. The ventilation network calculation engine employs an improved Ward-McBride numerical method, completing an iterative calculation of the entire mine's airflow distribution every 30 seconds with a calculation error ≤2%. The AI ​​decision-making model integrates a spatiotemporal convolutional neural network (ST-CNN) and a deep belief network (DBN), trained using historical operating data and typical accident cases. This enables prediction of airflow demand within the next 15 minutes and identification of abnormal gas trends, with a warning accuracy ≥92% and a response time <10 seconds. The AI ​​decision-making model is deployed on the edge platform, supporting on-demand invocation by multiple services and ensuring real-time execution of critical tasks through a priority scheduling mechanism. The edge computing nodes run the HarmonyOS operating system and a domestically produced real-time database, employing an incremental data upload mechanism that only transmits data packets whose changes exceed a threshold. Network load is optimized through local caching and cloud synchronization strategies.

[0032] Step 4: Construct a 3D digital twin visualization system based on a domestic GIS / BIM engine at the application layer to achieve dynamic display of ventilation status, remote airflow adjustment, and disaster simulation operations. The 3D digital twin visualization system dynamically renders wind speed vector fields, gas concentration cloud maps, and personnel location information, supporting hierarchical access control. Authorized users can initiate one-click reversal operations via mobile or PC terminals. After dual authentication, the command automatically verifies safety conditions and executes a preset action sequence. The system presets at least six emergency scenario templates, including fire, gas leak, and roof collapse / blockage. When abnormal conditions are detected, the system automatically activates the emergency plan, combining the current ventilation network topology and real-time parameters. It uses an improved Ward-McBride numerical method to simulate airflow disturbances, evaluates various airflow control strategies, generates the optimal solution, and triggers a sequence of actions such as main fan reversal, damper locking, and fan speed adjustment, completing the emergency airflow control response within 5 minutes.

[0033] In the above embodiments, the system of the present invention solves the problems of insufficient sensor accuracy and sluggish response of traditional systems by coordinating the deployment of a full stack of domestically produced high-precision multi-parameter sensors and intelligent actuators in the perception layer. This enables wind speed measurement error to be controlled within ±0.1m / s, gas concentration detection resolution to reach 0.01%, and supports a minimum adjustment accuracy of 0.5° for the damper. Because the network layer integrates 5G / Industrial Ethernet / LoRa and relies on the Kuanghong HCP / MDTP protocol to build a dual-mode redundant channel, it solves the problems of easy interruption of underground communication, protocol fragmentation, and easy data theft, ensuring communication availability of over 99.99% and end-to-end data security. Furthermore, because the platform layer runs Kuanghong OS on domestically produced ARM edge nodes, it achieves high accuracy. The self-developed calculation engine and AI model solve the problems of control delay and uncontrollable models caused by cloud dependence, compressing the air volume distribution calculation cycle to the 30-second level and shortening the AI ​​early warning response time to within 10 seconds. Because the application layer is based on the domestic GIS / BIM engine to build a three-dimensional digital twin system, it solves the problems of ventilation status being invisible, intangible, inaccurate, and impossible to simulate, making the success rate of remote ventilation adjustment operation ≥99.5%, and disaster simulation covering no less than 6 typical working conditions. Finally, the deployment of domestic software and hardware across the entire chain, from chips, operating systems, databases, communication protocols to the upper-level engine, has achieved independent control, completely avoiding the risk of foreign technology supply disruptions and meeting the dual strategic needs of coal mine inherent safety and digital transformation.

[0034] Working Principle: This invention provides a fully domestically produced intelligent ventilation system and method based on HarmonyOS for mining. The system can calculate the actual air volume requirement based on the quantity, process, status, and concentrated blasting time of personnel, equipment, and work locations in the underground work area. Combined with a three-dimensional ventilation simulation system, the system provides the operating frequency of the ventilation fans in the work area. Through real-time feedback data of underground environmental parameters, the system compares and judges the relationship between the actual air volume requirement and the actual air supply. The system uses the start-stop or stepless frequency conversion speed regulation control method of the ventilation fans to correct and adjust the actual air supply. It also has upper and lower limit functions for operating frequency to ensure that the actual air supply of the system is controlled within a basically safe and reasonable range, realizing real-time monitoring and control of the ventilation system in the work area.

[0035] Secondly, to facilitate efficient collaboration between sensor data acquisition at the perception layer, dual-mode redundant communication scheduling at the network layer, real-time inference between the platform layer's calculation engine and AI model, and 3D visualization rendering at the application layer, this invention also integrates domestically produced chips, PLCs, MCUs, operating systems, GIS / BIM engines, and real-time databases, forming a full-stack trusted foundation from the underlying hardware to the upper-layer applications. All components are interconnected through unified API specifications, national cryptographic algorithms, and BOM-level traceability management, ensuring 100% independent control of core technologies.

[0036] To reduce the impact of complex downhole electromagnetic environments, dust deposition, temperature and humidity fluctuations, and equipment aging on sensing accuracy, this invention also incorporates a four-probe cross array layout, IP68 protection and nano-coating, a temperature compensation module, an edge-side adaptive filtering algorithm, dual-motor redundant drive, high-precision magnetic encoder closed-loop feedback, a wind resistance dynamic correction factor, a nonlinear wind pressure compensation term, and a multi-objective optimization evaluation function, among other structural and algorithmic collaborative mechanisms, to comprehensively improve the system's sensing reliability, execution stability, and decision-making accuracy under harsh working conditions.

[0037] To improve ventilation control efficiency and personnel survival probability in the early stages of a disaster, this invention pre-sets at least six emergency scenario templates in the system, including fire, gas leak, and roof collapse. When abnormal conditions are detected, the system uses an improved Ward-McBride numerical method to simulate airflow disturbances, evaluates various ventilation control strategies, generates the optimal solution, and triggers a sequence of actions such as main fan reversal, door locking, and fan speed adjustment to complete the emergency ventilation control response within 5 minutes. This reduces the mine emergency ventilation control response time from the traditional 15–30 minutes to less than 5 minutes.

[0038] In addition, to ensure basic ventilation control functions are maintained during maintenance, emergency response, or communication interruptions, this invention adds on-site local control and LoRa Mesh self-organizing network local closed-loop operation functions. All key actuators support local command parsing and action execution in offline mode, and the continuity and traceability of control strategies during disasters are maintained through an edge node caching mechanism.

[0039] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A fully domestically produced intelligent ventilation system based on HarmonyOS for mining, characterized in that, It adopts a four-layer distributed architecture of "sensing-network-platform-application" and is equipped with domestically produced software and hardware throughout the entire chain, specifically including: -Perception layer: Deploys domestically produced high-precision sensors and actuators. The domestically produced high-precision sensors are used to collect parameters such as wind speed, wind pressure, gas concentration, temperature and humidity, and atmospheric pressure in real time. The actuators include intelligent dampers / windows and variable frequency fan controllers. - Network layer: Integrating 5G, industrial Ethernet and LoRa wireless communication technologies, all devices are connected to the mining HarmonyOS soft bus through the unified protocol of HarmonyOS HCP / MDTP, and a dual-mode redundant communication transmission channel is constructed. -Platform layer: Based on domestic ARM architecture high-performance edge computing nodes, running mining HarmonyOS operating system, integrating self-developed ventilation network calculation engine and AI decision model, supporting local real-time analysis and closed-loop control; - Application Layer: A 3D digital twin visualization system is built based on a domestic GIS / BIM engine to realize dynamic display of ventilation status, remote air adjustment and disaster simulation operation.

2. The fully domestically produced intelligent ventilation system based on HarmonyOS for mining, as described in claim 1, is characterized in that: The domestically produced high-precision sensors include wind speed sensors and multi-parameter sensors; The wind speed sensor adopts a four-probe cross array layout, the housing has an IP68 protection rating, the surface is coated with a dustproof and waterproof nano-coating, it has a built-in temperature compensation module, and with the edge-side adaptive filtering algorithm, the measurement error is ≤±0.1m / s, and it supports hot-swapping and remote firmware upgrades. The multi-parameter sensor integrates methane, CO, temperature, humidity and air pressure detection units, and the measurement range covers atmospheric pressure 80.00-120.00 kPa, temperature 0-50℃, humidity 20-95%RH. It communicates via RS485 interface and has a transmission distance of ≥2km at a communication speed of 2400bps.

3. The fully domestically produced intelligent ventilation system based on HarmonyOS for mining, as described in claim 1, is characterized in that... The intelligent damper / window adopts a dual-motor redundant structure, driven by a domestic PLC controller. The damper shaft is equipped with a high-precision magnetic encoder with a resolution of 0.1°, forming a closed-loop feedback control system with a minimum adjustment accuracy of 0.5°. The intelligent air door / air window is connected to the mining HarmonyOS system through a standardized communication interface. Status information is reported at a 100ms cycle, and the end-to-end delay of control commands is ≤50ms. The damper body adopts a Q345 steel frame and composite fireproof and heat insulation board splicing structure, and the air leakage rate is <3% when closed.

4. The fully domestically produced intelligent ventilation system based on HarmonyOS for mining, as described in claim 1, is characterized in that... The network layer communication substation supports 5G and LoRa dual-mode redundant networking, and automatically switches to LoRa Mesh self-organizing network mode in weak signal areas. All transmission links employ the national standard SM4 encryption algorithm to achieve end-to-end data security protection and prevent data theft or tampering.

5. A fully domestically produced intelligent ventilation system based on HarmonyOS for mining, as described in claim 1, is characterized in that... The ventilation network calculation engine of the platform layer adopts the improved Ward-McBride numerical method, which completes an iterative calculation of the air volume distribution of the entire mine every 30 seconds, with a calculation error of ≤2%. The AI ​​decision-making model integrates spatiotemporal convolutional neural networks and deep belief networks. Through training with historical operating data and typical accident cases, it can predict air volume demand and identify abnormal gas trends within the next 15 minutes, with an early warning accuracy of ≥92% and a response time of <10 seconds. The AI ​​decision-making model is deployed on an edge platform, supports on-demand invocation by multiple services, and ensures the real-time execution of critical tasks through a priority scheduling mechanism.

6. The fully domestically produced intelligent ventilation system based on HarmonyOS for mining, as described in claim 1, is characterized in that... The edge computing nodes of the platform layer use domestically produced ARM-A72 quad-core processors, run the Mining Hong operating system and a domestically produced real-time database, adopt an incremental data upload mechanism, and only transmit data packets whose changes exceed a threshold. The network load is optimized through local caching and cloud synchronization strategies.

7. A fully domestically produced intelligent ventilation system based on HarmonyOS for mining, as described in claim 1, is characterized in that... The application layer's 3D digital twin visualization system dynamically renders wind speed vector fields, gas concentration cloud maps, and personnel location information, and supports hierarchical access control. Authorized users can initiate a "one-click anti-virus" operation via mobile or PC terminals. After the command is dual-authenticated, the security conditions are automatically verified and a preset action sequence is executed.

8. A fully domestically produced intelligent ventilation system based on HarmonyOS for mining, as described in claim 1, is characterized in that... The system has pre-set templates for no fewer than 6 types of emergency scenarios, including fire, gas leak, and roof collapse. When an abnormal operating condition is detected, the system automatically activates the emergency plan. Combining the current ventilation network topology and real-time parameters, it performs airflow disturbance simulation using the improved Ward-McBride numerical method, evaluates various air control strategies, generates the optimal solution, and triggers a series of actions such as main fan reversal, damper locking, and fan speed adjustment to complete the emergency air control response within 5 minutes.

9. A fully domestically produced intelligent ventilation system based on HarmonyOS for mining, as described in claim 1, is characterized in that... The fully domestically produced hardware and software includes domestically produced chips, PLCs, MCUs, operating systems, GIS / BIM engines, and real-time databases, with no reliance on foreign technologies, achieving independent control over core technologies.

10. A control method for a fully domestically produced intelligent ventilation system based on HarmonyOS for mining, as described in any one of claims 1-9, characterized in that, Includes the following steps: S1: Real-time collection of wind speed, wind pressure, gas concentration, temperature, humidity and atmospheric pressure parameters through domestically produced high-precision sensors in the sensing layer; S2: The collected data is uploaded to the edge computing node of the platform layer via the dual-mode redundant communication channel of the network layer, and the unified management of the equipment is completed through the MiningHong HCP / MDTP protocol; S3: The platform layer completes the iterative calculation of the ventilation distribution of the entire mine based on the improved Ward-McBride numerical method, and calls the AI ​​decision model to predict ventilation demand and identify abnormal gas trends. S4: The application layer 3D digital twin system receives the output results from the platform layer, dynamically renders the ventilation status, and generates a wind control strategy based on a preset emergency template or manual instructions, and links the actuators to complete the damper adjustment, fan speed regulation or main fan reversal action.

Images