Mine intelligent ventilation management method and system

Abstract

The present application belongs to the technical field of mine ventilation control, and particularly relates to a mine intelligent ventilation control method and system, comprising the following steps: S1. Data sensing and fusion: through a multi-protocol data gateway, multi-source heterogeneous data from ventilation facilities, environmental sensors and personnel positioning systems is collected to generate a standardized data stream; S2. Digital twin construction and updating: based on the standardized data stream, a digital twin of the mine ventilation system is driven to be constructed; S3. Intelligent deduction and decision-making; S4. Strategy execution and closed-loop control; S5. Edge autonomy. Through the multi-protocol data gateway, heterogeneous data is integrated and a standardized data stream is generated, providing a unified and high-quality data basis for subsequent analysis and solving the information island problem. Based on real-time and feedback data, the digital twin is dynamically calibrated, which significantly improves the simulation accuracy and makes it more realistic to reflect the underground ventilation conditions.

Description

Technical Field

[0001] This invention belongs to the field of mine ventilation control technology, specifically relating to a mine intelligent ventilation control method and system. Background Technology

[0002] With the development of automation technology, remote monitoring and automatic control technologies have been gradually introduced into mine ventilation systems. Existing technologies typically involve deploying various sensors (such as wind speed, air pressure, and gas concentration sensors) and actuators (such as electric air doors and variable frequency fans) underground for centralized monitoring and control at a ground-based dispatch center. However, these systems still have many inherent shortcomings, making it difficult to achieve truly intelligent management and control.

[0003] First, existing systems often suffer from the problem of "data silos." Ventilation systems, environmental monitoring systems, and personnel positioning systems are usually built independently, making it difficult for the system to grasp a holistic overview of the underground ventilation environment, equipment status, and personnel distribution, resulting in insufficient decision-making basis.

[0004] Secondly, the control strategies of the system are mostly based on preset fixed thresholds or simple logic, lacking foresight and adaptability. When the underground environment changes or a sudden disaster occurs (such as gas outburst or fire), the existing system cannot quickly simulate the execution effect of different control strategies (such as reverse ventilation, increased resistance, and opening of backup roadways), making it difficult to evaluate and select the optimal solution among multiple possible options. The decision-making process relies on human experience, resulting in low efficiency and high risk.

[0005] Finally, the control models of existing systems are usually static or semi-static, which cannot use real-time sensor data and execution feedback to dynamically calibrate and update ventilation network model parameters (such as roadway resistance). This causes the model simulation accuracy to decrease over time, failing to truly reflect the underground ventilation conditions, and the reliability of decisions made based on this is questionable. Summary of the Invention

[0006] To address the aforementioned technical problems, one objective of this invention is to provide a mine intelligent ventilation control method capable of multi-source data fusion, intelligent inference and decision-making based on high-fidelity digital twins, closed-loop optimization capability, and autonomous and reliable operation even during communication interruptions; another objective of this invention is to provide a mine intelligent ventilation control system.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0008] A method for intelligent ventilation control in mines includes the following steps:

[0009] S1. Data Sensing and Fusion: Collect multi-source heterogeneous data from ventilation facilities, environmental sensors and personnel positioning systems through a multi-protocol data gateway, and generate standardized data streams;

[0010] S2. Digital Twin Construction and Update: Based on the standardized data stream, a digital twin of the mine ventilation system is constructed. During the initial run, real-time sensor data is used to initialize and calibrate the drag coefficient parameters in the digital twin, and the initialized digital twin is output. In subsequent runs, the drag coefficient parameters are dynamically calibrated by combining real-time sensor data and feedback data, and the dynamically updated digital twin is output.

[0011] S3. Intelligent simulation and decision-making: Utilize the digital twin output from step S2 to perform ventilation network calculations, simulate the effects of various ventilation control plans under different disaster types and occurrence points in parallel; evaluate and compare the simulation results, and generate the optimal ventilation control strategy based on the evaluation conclusions;

[0012] S4. Strategy Execution and Closed-Loop Control: The optimal ventilation control strategy is sent to the actuators of the corresponding ventilation facilities for execution;

[0013] And collect feedback data after the instruction is executed, and send the feedback data back to step S2;

[0014] S5. Edge Autonomy: Define the digital twin output by the current running step S2 and a specific area downhole as a digital twin sub-model, and send it along with the optimal ventilation control strategy to the edge controller deployed in that area;

[0015] When a communication interruption between the edge controller and the central processing system is detected, the edge controller executes autonomous ventilation control logic based on the locally stored digital twin sub-model and the optimal ventilation control strategy.

[0016] In step S1, a standardized data stream is generated by cleaning, compensating and time-series alignment of multi-source heterogeneous data.

[0017] In step S2, the specific process of dynamic calibration is as follows: the simulation calculation results of the digital twin are compared with the real-time sensor and feedback data, and the drag coefficient in the model is automatically adjusted through optimization algorithm so that the simulation value is close to the measured value.

[0018] In step S3, the evaluation comparison is performed using safety, effectiveness, and economy as multi-objective functions for optimization evaluation.

[0019] The actuators of the ventilation system in step S4 include damper controllers, window regulators, and main ventilation fan frequency converters; the specific underground area in step S5 refers to an area composed of one or more roadways connected in the ventilation network topology.

[0020] The edge controller in step S5 is an industrial control device with local computing, storage and communication capabilities, used to take over control of the actuators of ventilation facilities within its jurisdiction when communication is interrupted.

[0021] The edge controller periodically sends and receives heartbeat signals with the central processing system. If no heartbeat response is received within a preset time threshold, it is determined to be an interruption.

[0022] It also includes the following steps:

[0023] S6. Dynamic Navigation: Based on the simulation results obtained from the parallel simulation in step S3, and combined with the personnel location information contained in the standardized data stream generated in step S1, the optimal disaster avoidance route is dynamically generated and pushed.

[0024] A mine intelligent ventilation control system for performing the method includes:

[0025] The data perception and fusion layer is used to perform step S1;

[0026] The digital twin and decision layer, used to execute steps S2 and S3, has a multi-objective optimization evaluator inside to evaluate and compare simulation results;

[0027] The execution control layer, used to execute step S4, includes the actuators of a central processing system deployed on the well and various ventilation facilities connected in communication with it;

[0028] An edge autonomous layer, used to perform step S5, includes an edge controller deployed in a specific area downhole;

[0029] The execution control layer is equipped with a data feedback acquisition channel to send feedback data back to the digital twin and decision layer; the digital twin and decision layer is communicatively connected to the edge autonomous layer and is used to distribute digital twin sub-models and control strategies.

[0030] It also includes a dynamic navigation module for performing step S6.

[0031] The actuators of the ventilation system include a damper controller, a window regulator, and a main ventilation fan frequency converter; the edge controller has a built-in communication monitoring module and a lightweight simulation module. The communication monitoring module is used to monitor the link status through a heartbeat mechanism, and the lightweight simulation module can perform real-time fine-tuning of the digital twin sub-model based on local sensor data.

[0032] Compared with the prior art, the beneficial effects of this invention are:

[0033] By integrating heterogeneous data through a multi-protocol data gateway and generating standardized data streams, a unified and high-quality data foundation is provided for subsequent analysis, solving the problem of information silos. Dynamic calibration of the digital twin (model) based on real-time and feedback data significantly improves simulation accuracy, enabling it to more realistically reflect downhole ventilation conditions. Furthermore, parallel simulation and multi-objective optimization are employed to evaluate and generate control strategies, transforming the decision-making process from experience-based to scientific computation, thereby improving the accuracy and efficiency of disaster response.

[0034] This method forms a complete closed-loop control system, enabling the system to continuously self-optimize based on execution feedback and dynamically adapt to changes in downhole conditions. Most importantly, by designing an edge autonomy mechanism, it can maintain regional ventilation safety even under special conditions of communication interruption, based on locally stored models and strategies.

[0035] By combining ventilation control with dynamic navigation, the system generates and pushes the optimal evacuation route based on the simulated disaster path and personnel location information, thus extending from wind control to personnel protection and forming a complete safety guarantee closed loop. Attached Figure Description

[0036] Figure 1 This is a connection block diagram of Embodiment 1 of the present invention. Detailed Implementation

[0037] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0038] Example 1

[0039] This embodiment provides a method for intelligent ventilation control in mines, such as... Figure 1 As shown, the specific implementation process is as follows:

[0040] S1. Data Sensing and Fusion: Through a multi-protocol data gateway deployed underground, multi-source heterogeneous data from devices such as intelligent dampers, variable frequency fans, methane sensors, carbon monoxide sensors, temperature sensors, air pressure sensors, and personnel positioning base stations (systems) are collected in real time. The gateway's built-in data processing unit cleans (removes outliers), compensates (interpolates data), and performs time-series alignment on the raw data, generating a standardized data stream with a unified timestamp and format, which is then transmitted to the surface data center via the industrial ring network.

[0041] S2. Construction and updating of digital twin: Based on the above standardized data flow, a digital twin of the mine ventilation system is constructed using three-dimensional ventilation simulation software. This model accurately reflects the roadway topology, the location of ventilation facilities, and characteristic parameters.

[0042] During initial operation, real-time sensor data is used to initialize and calibrate the drag coefficient parameters in the digital twin, outputting the initialized digital twin. In subsequent operations, the drag coefficient parameters are dynamically calibrated by combining real-time sensor data and feedback data, outputting a dynamically updated digital twin.

[0043] Specifically, a parameter inversion algorithm based on the least squares method can be used to compare the simulation calculation results of the digital twin (such as wind speed in the tunnel) with real-time sensor data and feedback data from step S4, dynamically calibrate the drag coefficient in the model, make the simulation value close to the measured value, and output the digital twin.

[0044] S3. Intelligent Simulation and Decision-Making: Utilizing the digital twin output from step S2, ventilation network calculations are performed, simulating in parallel the effects of various ventilation control plans (such as adjusting window opening, reversing ventilation, and activating backup roadways) under different disaster types and occurrence points, including gas outbursts and fires. The simulation results of each plan are evaluated and compared using a multi-objective optimization algorithm, with safety (e.g., whether gas concentration exceeds limits), effectiveness (e.g., smoke extraction time), and economy (e.g., power consumption) as optimization objectives to generate the optimal ventilation control strategy.

[0045] S4. Strategy Execution and Closed-Loop Control: The optimal ventilation control strategy is sent to the corresponding actuators (such as damper motors controlled by programmable logic controllers and main ventilation fans driven by frequency converters) for execution via industrial communication protocols.

[0046] Real-time collection of feedback data such as wind pressure and air volume after command execution, and feeding back the feedback data to the digital twin construction and update process in step S2 to form a closed-loop optimization.

[0047] S5. Edge Autonomy: Cut the part of the digital twin output by the current running step S2 that is related to a certain mining face into a digital twin sub-model, and send it along with the optimal ventilation control strategy to the edge controller deployed in the explosion-proof box in that area.

[0048] The edge controller maintains communication with the central processing system above ground through periodic heartbeat packets. When the heartbeat packets are lost more than a preset number (i.e., communication is interrupted), the edge controller executes preset autonomous ventilation control logic based on the locally stored digital twin sub-model and strategy, such as maintaining the current damper status or executing the last valid instruction.

[0049] Furthermore, in step S1, data cleaning uses mathematical statistics criteria to remove outliers, compensation uses interpolation to fill in missing data, and time alignment uses a time synchronization mechanism.

[0050] Furthermore, in step S2, the dynamic calibration process specifically involves: establishing an error function between the simulated wind speed of the digital twin and the measured wind speed of the sensor; using an optimization algorithm to iteratively adjust the frictional resistance coefficient and local resistance coefficient in the model until the error function is minimized, thereby achieving approximation of the measured value to the simulated value.

[0051] Furthermore, in step S3, the specific indicators for multi-objective optimization evaluation include: safety indicators (such as gas concentration not exceeding the limit and temperature within the safe range within the disaster spread range), effectiveness indicators (such as disaster airflow reversal time), and economic indicators (such as total power consumption of the ventilation system).

[0052] By using a weighted summation method or an optimal solution screening mechanism, the overall optimal ventilation control strategy is evaluated from multiple scenarios simulated in parallel.

[0053] Furthermore, the ventilation actuator in step S4 specifically includes: an explosion-proof electric push rod controller for controlling the opening and closing of the damper, a stepper motor controller for adjusting the angle of the air window, and an explosion-proof frequency converter for controlling the speed of the main ventilation fan.

[0054] Furthermore, the specific underground area in step S5 refers to a coal mining face and its associated intake and return airways connected on the ventilation network diagram.

[0055] Furthermore, the edge controller employs an embedded industrial control computer, equipped with a multi-core processor, memory, and storage, and features a built-in operating system and programming environment. This allows it to take over control of actuators such as damper controllers within its jurisdiction during communication interruptions. Communication interruption monitoring involves the edge controller periodically sending heartbeat packets to the wellhead server; if no response is received for several consecutive cycles, autonomous mode is triggered.

[0056] Furthermore, it also includes step S6, dynamic navigation. In step S6, based on the fire smoke diffusion path deduced in step S3 (such as a carbon monoxide concentration distribution cloud map), and combined with the real-time personnel location information (personnel location coordinates) obtained in step S1, a path planning algorithm is used to dynamically generate the shortest disaster avoidance path around the danger zone, and push it to the miner's handheld terminal and emergency broadcast through the underground wireless access point and broadcast system.

[0057] Example 2

[0058] This embodiment provides a system for performing the above method, specifically including:

[0059] The data perception and fusion layer, consisting of a monitoring system and a data cleaning server, is used to execute step S1;

[0060] The digital twin and decision-making layer consists of servers that run simulation software and optimization algorithm modules, which are used to execute steps S2 and S3;

[0061] The execution control layer consists of an industrial control computer, a programmable logic controller, and connected damper actuators and frequency converters in the wellhead dispatch room. It is used to execute step S4.

[0062] The edge autonomous layer consists of an industrial control computer and communication module inside an explosion-proof box in the well, which is used to execute step S5.

[0063] The execution control layer transmits feedback data, such as wind turbine current, back to the digital twin and decision layer via industrial communication protocols; the digital twin and decision layer distributes sub-models and strategies to the edge controller via the industrial ring network.

[0064] Furthermore, the dynamic navigation module can be implemented by a web server with a browser and server architecture. The server calls the disaster spread interface and personnel positioning interface provided by the digital twin and the decision-making level to generate a dynamic disaster avoidance route map, and pushes it to the intrinsically safe mobile phone for mining and the underground emergency information screen through the network communication protocol.

[0065] Furthermore, the aforementioned ventilation actuators specifically include damper controllers, window regulators, and main ventilation fan frequency converters. The communication monitoring module built into the edge controller continuously monitors the link status with the well server by detecting commands; if the response time exceeds a threshold, an early warning is triggered.

[0066] Furthermore, the lightweight simulation module built into the edge controller uses a simplified ventilation network solution algorithm, which can fine-tune the branch airflow in the digital twin sub-model in real time based on locally deployed laser anemometer data, thereby improving the accuracy of autonomous control.

[0067] The above description only illustrates preferred embodiments of the present invention, but the present invention is not limited to the above embodiments.

Claims

1. A method for intelligent ventilation control in mines, characterized in that, Includes the following steps: S1. Data Sensing and Fusion: Collect multi-source heterogeneous data from ventilation facilities, environmental sensors and personnel positioning systems through a multi-protocol data gateway, and generate standardized data streams; S2. Digital Twin Construction and Update: Based on the standardized data stream, a digital twin of the mine ventilation system is constructed. During the initial run, real-time sensor data is used to initialize and calibrate the drag coefficient parameters in the digital twin, and the initialized digital twin is output. In subsequent runs, the drag coefficient parameters are dynamically calibrated by combining real-time sensor data and feedback data, and the dynamically updated digital twin is output. S3. Intelligent simulation and decision-making: Utilize the digital twin output from step S2 to perform ventilation network calculations, simulate the effects of various ventilation control plans under different disaster types and occurrence points in parallel; evaluate and compare the simulation results, and generate the optimal ventilation control strategy based on the evaluation conclusions; S4. Strategy Execution and Closed-Loop Control: The optimal ventilation control strategy is issued to the actuators of the corresponding ventilation facilities for execution; and feedback data after the command execution is collected and the feedback data is fed back to step S2; S5. Edge Autonomy: Define the digital twin output by the current running step S2 and a specific area downhole as a digital twin sub-model, and send it along with the optimal ventilation control strategy to the edge controller deployed in that area; When a communication interruption between the edge controller and the central processing system is detected, the edge controller executes autonomous ventilation control logic based on the locally stored digital twin sub-model and the optimal ventilation control strategy. In step S1, a standardized data stream is generated by cleaning, compensating and time-series alignment of multi-source heterogeneous data. In step S2, the specific process of dynamic calibration is as follows: the simulation calculation results of the digital twin are compared with the real-time sensor and feedback data, and the drag coefficient in the model is automatically adjusted through the optimization algorithm so that the simulation value is close to the measured value. In step S3, the evaluation comparison is an optimization evaluation based on multiple objective functions, namely safety, effectiveness, and economy. The edge controller in step S5 is an industrial control device with local computing, storage and communication capabilities, used to take over control of the actuators of ventilation facilities within its jurisdiction when communication is interrupted. The edge controller periodically sends and receives heartbeat signals with the central processing system. If no heartbeat response is received within a preset time threshold, it is determined to be an interruption.

2. The intelligent ventilation control method for mines according to claim 1, characterized in that, The actuators of the ventilation system in step S4 include damper controllers, window regulators, and main ventilation fan frequency converters; the specific underground area in step S5 refers to an area composed of one or more roadways connected in the ventilation network topology.

3. A method for intelligent ventilation control in mines according to claim 1 or 2, characterized in that, It also includes the following steps: S6. Dynamic Navigation: Based on the simulation results obtained from the parallel simulation in step S3, and combined with the personnel location information contained in the standardized data stream generated in step S1, the optimal disaster avoidance route is dynamically generated and pushed.

4. A mine intelligent ventilation control system for performing the method as described in claim 3, characterized in that, include: The data perception and fusion layer is used to perform step S1; The digital twin and decision layer, used to execute steps S2 and S3, has a multi-objective optimization evaluator inside to evaluate and compare simulation results; The execution control layer, used to execute step S4, includes the actuators of a central processing system deployed on the well and various ventilation facilities connected in communication with it; An edge autonomous layer, used to perform step S5, includes an edge controller deployed in a specific area downhole; The execution control layer is equipped with a data feedback acquisition channel to send feedback data back to the digital twin and decision-making layer. The digital twin and the decision-making layer are communicatively connected to the edge autonomous layer for distributing digital twin sub-models and control strategies.

5. The intelligent ventilation control system for mines according to claim 4, characterized in that, It also includes a dynamic navigation module for performing step S6.

6. The intelligent ventilation control system for mines according to claim 4, characterized in that, The actuators of the ventilation system include a damper controller, a window regulator, and a main ventilation fan frequency converter; the edge controller has a built-in communication monitoring module and a lightweight simulation module. The communication monitoring module is used to monitor the link status through a heartbeat mechanism, and the lightweight simulation module can perform real-time fine-tuning of the digital twin sub-model based on local sensor data.

Images