Virtual training system and method for coal mine safety monitoring operation

Abstract

The invention discloses a virtual training system and method for coal mine safety monitoring operation, and belongs to the technical field of coal mine safety training. The system comprises a real object operation test and training device, a computer and virtual training software. The software comprises a virtual training scene construction module, a monitoring data acquisition module, a theoretical training module, a real object operation training module and an intelligent evaluation module. According to the invention, multi-dimensional real-time operation data of a student on a real object device is collected, a dynamic time warping algorithm is introduced, an operation time sequence of the student is compared with a standard operation flow sequence to quantify operation normalization, and automatic, precise and intelligent evaluation of practical operation skills is realized in combination with a weighted scoring model. According to the method, the whole business process from equipment installation, sensor calibration to troubleshooting can be completely simulated, the technical problems that an existing training mode breaks away from reality, evaluation means are backward, and operation problems cannot be accurately positioned are effectively solved, and the effectiveness and scientificity of training are improved.

Description

Technical Field

[0001] This invention relates to the field of coal mine safety training technology, and in particular to a virtual training system and method for coal mine safety monitoring and control operations. Background Technology

[0002] Currently, coal mines frequently experience problems such as excessive gas alarms, frequent wire breaks, numerous malfunctions, improper calibration, and inadequate sensor use and maintenance. Coal mine safety monitoring workers are responsible for installing, debugging, inspecting, and maintaining underground safety monitoring systems, ensuring their stable operation. Their work environment is complex, their responsibilities are significant, and high levels of professional knowledge and operational skills are required. Therefore, specialized training for coal mine safety monitoring workers through theoretical learning and hands-on practice is of great importance.

[0003] However, due to the unique and dangerous nature of the underground coal mine environment, key operational skills such as standardized construction, operation and maintenance, and troubleshooting of safety monitoring equipment like sensors and substations cannot be trained safely and in real-time underground, and are difficult to effectively replicate in the actual environment. Currently, coal mine safety monitoring and control operation training mainly adopts two traditional methods: off-the-job training and mentorship.

[0004] The off-the-job training primarily relies on classroom lectures by teachers at coal mine technical schools. This classroom-based training is often overly theoretical and severely disconnected from actual work scenarios. Numerous technical terms and abstract concepts make it difficult for trainees to understand, and the training content fails to closely integrate with the complex and ever-changing monitoring environment underground in coal mines. Furthermore, there is a lack of real safety monitoring equipment for operational training. In addition, limited class time prevents in-depth answers to each trainee's questions, leaving them helpless when encountering problems in actual operation. This seriously affects the training's effectiveness and the safe and efficient conduct of subsequent work, resulting in a disconnect between training and actual coal mine production operations and poor training outcomes.

[0005] In the traditional mentorship model, the skill levels of mentors vary greatly. Some mentors possess outdated knowledge and lack understanding of new equipment and technologies, resulting in the passed-on experience and skills lagging behind current technological advancements. Furthermore, varying degrees of closeness between mentors and apprentices can lead to arbitrary teaching methods, making it difficult to guarantee the apprentice's learning progress. This approach struggles to train trainees in standardized operating procedures and behavioral norms, and it also fails to impart systematic troubleshooting and problem-solving solutions, ultimately hindering the effective guarantee of trainees' learning progress and training outcomes.

[0006] Therefore, given the urgent need for installation and maintenance of coal mine safety monitoring systems, developing a virtual training system for coal mine safety monitoring operations that can simulate actual production operation scenarios and standard operating procedures is of great significance. Currently, most coal mine safety monitoring training products on the market are designed based on the "Coal Mine Safety Monitoring and Control Operation Safety Technology Practical Operation Examination Standard" issued in 2016. This examination standard mainly covers the assessment of safe operation skills such as mine sensor installation, mine sensor calibration, and parameter setting of underground monitoring substations, aiming to meet the basic requirements for special operations personnel to obtain certification and hold their positions.

[0007] However, these products have significant shortcomings: First, their training curriculum is incomplete, failing to delve into key operational content such as troubleshooting solutions for safety monitoring equipment, installation and setup standards for various sensors, and equipment operation methods. Second, these products do not utilize real safety monitoring equipment for safety operation training, preventing trainees from gaining hands-on experience consistent with production sites. Third, and most importantly, existing products cannot effectively and intelligently analyze and evaluate the compliance of trainees' actual operations, failing to accurately pinpoint specific problems in operational procedures, standardization, and efficiency. This results in serious deficiencies in the practicality and relevance of the training, failing to provide scientific and accurate guidance for on-site coal mine safety monitoring operations, and failing to achieve the refined training goal of promoting learning through assessment. Summary of the Invention

[0008] The technical problem to be solved by the present invention is: In order to overcome the above-mentioned technical problems, the present invention provides a virtual training system and method for coal mine safety monitoring and control operations.

[0009] The technical solution adopted by the present invention to solve its technical problem is: a virtual training system for coal mine safety monitoring and control operations, characterized in that it includes a physical operation examination and training device and a computer; The physical operation and testing device is used to simulate coal mine safety monitoring equipment; the computer is equipped with virtual training software; the physical operation and testing device interacts with the virtual training software in the computer via a communication protocol. The physical operation training device includes a monitoring and control physical operation device and a mining sensor calibration training device. The monitoring and control physical operation device collects and outputs real-time monitoring data and operating parameters through multiple sensors. The mining sensor calibration training device is used to provide standard gas samples required for calibration of mining gas sensors and monitor the gas flow rate during the calibration process. The virtual training software includes a virtual training scenario construction module, a monitoring and data acquisition module, and an intelligent evaluation module; the real-time monitoring data and operating parameters collected by the physical operation and testing device are synchronously output to the virtual training software through the monitoring and data acquisition module for training and assessment of trainees. The virtual training scenario construction module is used to construct an interactive three-dimensional virtual environment in a coal mine. The monitoring and data acquisition module is used to collect multi-dimensional real-time operation data of trainees from the physical operation and testing device, driving the three-dimensional virtual environment to dynamically change according to the collected multi-dimensional real-time operation data, realizing the virtual-real linkage between the simulation equipment model, the virtual scene, and the physical operation and testing device, and to construct more realistic virtual training scenarios and fault cases, thereby enhancing the relevance of training. The intelligent evaluation module includes a sequence comparison unit and a scoring calculation unit, which, by recording all the trainees' operation logs and based on preset scoring rules, evaluates the trainees' operational standardization, process correctness, and test question answers. The system automatically quantifies and scores troubleshooting time and the rationality of emergency response to test and assess training effectiveness. The sequence comparison unit is configured to use the Dynamic Time Warping (DTW) algorithm to compare the time series of trainees' safety monitoring operations with the time series of standard operating procedures for safety monitoring. The scoring calculation unit is configured to output a practical evaluation score based on the comparison results of the sequence comparison unit using a weighted scoring model. The weighted scoring model is configured to convert the comparison results of the sequence comparison unit into operation step scores, and then perform a weighted summation based on preset step weights to generate the final practical evaluation score.

[0010] The physical hands-on training device synchronizes real-time operational data, such as equipment status and operating parameters, from trainees' actual training sessions to the virtual training software. This ensures hands-on operation consistent with the production site, meeting the needs of practical skills training. The virtual training software simulates the actual operation procedures and various fault phenomena of the coal mine safety monitoring system, fully demonstrating the entire workflow of standard inspection and operation for monitoring and control work. It can be used by trainees to learn the working principles and operating procedures of various safety monitoring equipment in coal mines, accurately mastering the technical skills and behavioral norms for equipment installation, commissioning, and maintenance, achieving experiential and case-based training for safety monitoring workers.

[0011] The physical hands-on training device interacts with the virtual training software via Modbus and MQTT communication protocols.

[0012] The monitoring and control physical operation device includes a monitoring substation, a ring network access device, a mining methane sensor, a remote control switch, a junction box for intrinsically safe mining circuits, and an infrared remote controller for system operation. It collects real-time monitoring data and operating parameters from each sensor and outputs them to the monitoring and control data acquisition module.

[0013] The mining sensor calibration training device includes a flow control component and a gas cylinder component. The flow control component monitors the real-time gas flow rate during sensor calibration. The gas cylinder component provides a stable flow rate of air and standard methane gas samples for sensor zeroing and accuracy calibration. The gas cylinder component includes a standard methane cylinder, an air cylinder, and a pressure reducing valve connected to the standard methane and air cylinders. The pressure reducing valve controls the output flow rates of the standard methane and air cylinders to provide standard methane and air gas samples for methane sensor calibration training. The gas flow sensor monitors the real-time gas flow rate during sensor calibration to determine whether the trainee is operating according to the standard required gas sample flow rate.

[0014] The virtual training software also includes a theoretical training module, which includes a 3D equipment simulation model, interactive model animations, and multimedia teaching materials, used to provide theoretical training on the overall composition of the coal mine safety monitoring system, equipment operation methods, and fault diagnosis.

[0015] The virtual training software also includes a physical operation training module, which guides trainees to interactively learn practical skills based on the standard operating procedure logic and common system fault troubleshooting solutions of the coal mine safety monitoring and control system.

[0016] The virtual training scenario construction module constructs a three-dimensional virtual environment in a coal mine, designed according to the training needs of coal mine safety monitoring workers. This includes three-dimensional area scene models of coal mining faces, tunneling faces, and return airways; a two-dimensional system topology model of the coal mine safety monitoring system; and three-dimensional equipment models of the safety monitoring and control equipment within the system. The three-dimensional virtual environment is designed using 3ds Max software to model the three-dimensional area scene of coal mining faces, tunneling faces, and return airways in the coal mine. The modeling mainly includes two categories: The first category is the environment model, such as coal seams and roadways in different types of working faces, as well as inherent production equipment such as coal mining machines, tunneling machines, and belt conveyors, simulating the real environment of coal mine safety monitoring operations. The second category is the three-dimensional equipment models of the safety monitoring and control equipment. Based on the actual installation and setting requirements of underground sensors and substations, the equipment models are placed in the scene to form the final three-dimensional model, which is then imported into the virtual engine for virtual training software development. The system is designed with interactive operation based on the actual coal mine safety monitoring and control process. When trainees learn different business knowledge in the virtual scene, the corresponding videos and animations are automatically played, allowing operators to learn in a realistic virtual environment.

[0017] A practical evaluation method for a virtual training system for coal mine safety monitoring and control operations provided by the present invention includes the following steps: Step 1: Record all student operation logs to create a multidimensional dataset containing at least two types of data: operation event data, time-series data, answer data, and equipment status parameters. This dataset represents the student's multidimensional real-time operation data during the practical training process. All data is then integrated into a single multidimensional feature vector V. t =(Operation event code, timing data value, answer score, sensor status parameter 1, sensor status parameter 2, etc.).

[0018] The aforementioned operation event data includes the actual actions performed by the student at operation step t. The system preprocesses and extracts features from the student's operation video via camera, and matches the student's operation sequence to the corresponding operation step sequence number. For example, sequence number 1: cleaning the sensor and air chamber dust, sequence number 2: setting the sensor to enter online calibration mode, etc. If no corresponding step is detected, it means that the student has missed a key step. The above time series data represents the time taken by the student to complete the operation, and is statistically analyzed in seconds. It is used to analyze the time taken by the student to complete the operation and to judge the student's proficiency in the operation. The above answer data represents the trainees' performance in different virtual training scenarios; it is used to assess the trainees' mastery of safety standards and regulations. A correct answer earns 1 point, and an incorrect answer earns 0 points. The above equipment status data are the equipment status parameters of the trainee at operation step t, including the equipment status log during the sensor calibration operation, which includes sensor function codes, changes in self-monitored values, calibration start time, calibration gas flow monitoring value, etc.

[0019] Step 2: Based on the Dynamic Time Warping (DTW) algorithm, the operation steps of trainees in the process of sensor installation, calibration, and monitoring substation parameter setting are compared with the preset standard operation instruction library. The path similarity is calculated, and the operation sequence is checked to see if it is consistent with the standard sequence. If there is an error in the operation sequence, omission of key steps, or failure to meet the standard of operation, an alarm will be triggered and points will be deducted.

[0020] First, construct the distance matrix, where the standard operation instruction library sequence Q = ( , ,… , ), each It is a multidimensional feature vector, representing the first feature vector in the standard process. The ideal multidimensional state of the steps. The student's operation process sequence C = ( , ,… , ), each It is a multidimensional feature vector, representing the student's position in the th... The multi-dimensional state of actual operation at each point in time. First, construct a The matrix D, each element in the matrix This indicates the first instruction in the standard operation instruction library sequence. Points The first in the sequence of student operation process Points The weighted Euclidean distance between them is calculated using the following formula: =

[0021] in, and They are vectors and In the Eigenvalues ​​in dimension, This represents the weight of the k-th dimension. Higher weights can be assigned to key operational events or state parameters of specific devices, thus enabling more targeted differentiation of operational norms.

[0022] Step 3: Construct the cumulative distance matrix. Create a cumulative distance matrix C, whose dimensions are also... Matrix elements Indicates starting from the path origin arrive The minimum cumulative distance among all paths.

[0023] initialization: = ; First row of the matrix: = ; First column of the matrix: = ; For the other elements of the matrix Its value is determined by the following formula. =

[0024] This recursive relationship means that, upon reaching... The minimum cumulative cost of a point is equal to the local cost of that point plus the cost to its left. The one with the smallest cumulative cost among these three adjacent cells guarantees the continuity and monotonicity of the path, ensuring that each point matches and the path points move forward with each step.

[0025] Finding the optimal regularization path: Optimal regularization path =( , ,… , (From the end point) Back to the beginning A sequence of points, where each point =( The backtracking rule is: start from the endpoint. To begin, each time select one of the three adjacent cells to the left, top, or top left that makes The smallest cell is taken as the previous point in the path, until the starting point (1,1) is reached. Finally, the DTW distance between the two sequences is calculated from the endpoint element of the cumulative distance matrix. This indicates that the smaller the value, the higher the overall similarity between the two sequences.

[0026] Step 4: By measuring the Euclidean distance The scoring is converted to calculate the student's operational score for each step, which is used to assess the student's operational standardization and skill mastery level. The system employs a weighted summation algorithm to automatically evaluate and score trainees' practical exams, generating detailed diagnostic reports with corresponding deductions and improvement measures. The specific calculation formula is as follows:

[0027] The "Score" represents the student's practical exam score in safety monitoring and surveillance operations. Each step of the operation is scored to reflect whether the trainee has followed the standard operating procedures and code of conduct. The total number of steps is used to average the scores of all operation steps.

[0028] Compared with the prior art, the virtual training system and method for coal mine safety monitoring and control operations provided by the present invention have the following significant advantages: 1. This invention applies a dynamic time warping algorithm to the practical assessment of coal mine safety monitoring operations. Through this algorithm, the system can treat the entire operation process of trainees with different operating rhythms and speeds as a time series, and perform global comparison and similarity calculation with the time series of standard operating procedures. This technical approach fundamentally solves the industry problem of inaccurate and unfair evaluations caused by differences in operation timing in traditional assessment methods. Combined with a multi-dimensional data collection system including operational event data, time series data, answer data, and equipment status data, and a weighted scoring model based on algorithm output, the system can transform subjective and vague operational evaluations into objective and accurate data scores, greatly avoiding interference from human factors and achieving objectivity and fairness in the assessment results.

[0029] 2. The intelligent assessment module of this invention not only provides the final assessment score but also accurately identifies specific problems trainees encounter in operational procedures, behavioral standardization, operational efficiency, and decision-making logic. The system can clearly pinpoint which steps are omitted, which operational sequence is incorrect, and which key parameters are not up to standard, and provide corresponding improvement measures. This in-depth diagnostic capability overcomes the shortcomings of traditional training that only provides results without understanding the causes, effectively guiding trainees to make targeted improvements and practice. It truly achieves the refined training goal of promoting learning through assessment and practice through evaluation, significantly improving training efficiency and quality.

[0030] 3. By employing physical operation and training equipment completely identical to the underground coal mine production environment, such as real monitoring substations, mining sensors, and calibration training devices, combined with a constructed 3D virtual environment of an underground coal mine, the system provides trainees with a hands-on operation platform highly consistent with the production site. Through standard communication protocols such as Modbus and MQTT, real-time, bidirectional dynamic driving of physical operation data and the virtual scene is achieved. This architecture, combining virtual and real elements and providing interconnected feedback, ensures that trainees gain a realistic feel for operating the equipment while simulating the complex and ever-changing underground environment and difficult-to-reproduce fault cases through the virtual scene. It combines immersion and interactivity, greatly enhancing the authenticity and relevance of the training while ensuring safety.

[0031] 4. This invention breaks through the limitations of existing training products that only target certification subjects, constructing a training system covering the entire business process of coal mine safety monitoring and control. The system integrates multiple modules such as theoretical training, equipment deployment, sensor calibration, and fault simulation troubleshooting, supporting trainees to learn the entire business process from equipment installation and debugging to daily operation, maintenance, and even emergency handling of complex faults. Through systematic case-based and experiential training, it can comprehensively improve the fault diagnosis and emergency response capabilities of safety monitoring workers in complex environments, thereby fundamentally enhancing the safety monitoring system maintenance capabilities of coal mining enterprises and ensuring safe coal mine production.

[0032] In summary, this invention, by deeply integrating physical simulation, virtual reality, and intelligent evaluation algorithms, has successfully created an intelligent virtual training platform with comprehensive training content, realistic training process, and accurate training evaluation. It effectively solves the core pain points in current coal mine safety monitoring and control operation training, such as the disconnect between training and practice, outdated evaluation methods, and poor training results, and has extremely high promotion and application value. Attached Figure Description

[0033] Figure 1 This is a schematic diagram of the system architecture of the virtual training system for coal mine safety monitoring and control operations provided in this embodiment of the invention.

[0034] Figure 2 This is a flowchart illustrating the method for practical assessment and judgment by the intelligent evaluation module in this embodiment of the invention.

[0035] Figure 3 This is a schematic diagram of the overall process of the virtual training method for coal mine safety monitoring and control operations provided in this embodiment of the invention.

[0036] Figure 4 This is a flowchart of the virtual training method for simulating and troubleshooting faults in a security monitoring system, as described in this invention. Detailed Implementation

[0037] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that the following embodiments are only for explaining the present invention and should not be construed as limiting the present invention.

[0038] Example 1: This embodiment provides a virtual training system for coal mine safety monitoring and control operations, such as... Figure 1 As shown, it includes a physical hands-on testing and training device, a computer, and virtual training software installed on the computer.

[0039] The physical operation training device is a key hardware component connecting virtual and real-world operations, and it includes: Monitoring and Control Device: This device integrates the host computer software for the coal mine safety monitoring system, monitoring substations, ring network access devices, mine-use methane sensors, remote control switches, intrinsically safe junction boxes for mine circuits, wind speed sensors, mine-use start / stop sensors, and infrared remote controllers for system operation. Its core function is to collect real-time monitoring data and operating parameters from various sensors and synchronously output them to virtual training software and the host computer software via subsequent monitoring and control data acquisition modules, providing a realistic equipment interactive experience for training and assessment.

[0040] Mining Sensor Calibration Training Device: This device is specifically designed for training in the calibration of various mining gas sensors, especially methane sensors. It includes a flow control component and a gas cylinder component. The flow control component monitors the real-time gas flow rate during the calibration process to determine if the trainee is operating according to the standard gas sample flow rate requirements. The gas cylinder component provides suitable flow rates and stable air and standard methane samples for zero-point calibration and accuracy calibration of the sensor. The gas cylinder component includes a standard methane cylinder, an air cylinder, and a pressure reducing control valve connected to the standard methane and air cylinders. By operating the pressure reducing control valve, it is possible to control the output flow rates of gas from the standard methane and air cylinders, providing standard methane and air samples for methane sensor calibration training.

[0041] The virtual training software is the intelligent core of the system, running within the computer, and includes the following core modules: Virtual Training Scenario Construction Module: Used to build an interactive 3D virtual environment for underground coal mines. Specifically, 3D scene models of coal mining faces, tunneling faces, and return airways are created in 3ds Max software. At the same time, a 2D system topology model of the coal mine safety monitoring system and 3D equipment models of various sensors, substations, and other safety monitoring equipment in the system are constructed. Finally, these are imported into the virtual engine to form complete virtual training courseware.

[0042] Monitoring and Data Acquisition Module: This module serves as a data bridge for achieving virtual-real linkage. It is electrically connected to the coal mine safety monitoring system and the physical operation training device, communicating with the computer using the standard Modbus communication protocol. It collects real-time data on trainee operations, such as sensor function codes, changes in self-monitored values, calibration start time, stable duration, and calibration gas flow rate. When data changes, it is immediately sent to the data exchange area, and then notified to the virtual training software and the monitoring system's host computer software via the MQTT communication protocol. The equipment's real-time data and status are converted into drive signals recognizable by the virtual simulation scene, driving dynamic changes in the scene and simulating the actual underground operations of safety monitoring workers. This achieves virtual-real linkage between the simulation equipment model, the virtual scene, and the physical operation training device. Simultaneously, the virtual training software automatically generates a multidimensional dataset containing operation steps, answer status, calibration flow rate, calibration concentration, and calibration time, enabling intelligent evaluation of trainee operation procedures.

[0043] The theoretical training module systematically imparts theoretical knowledge based on training outlines such as the "Coal Mine Safety Monitoring and Control Operation Safety Technology Training Outline and Assessment Requirements." This module further includes a system awareness learning module, an equipment deployment training module, and a common troubleshooting module. The system cognitive learning module employs interactive equipment simulation models and operational video tutorials to facilitate cognitive learning of the hardware equipment. It includes a topology learning module and an equipment simulation learning module. The topology learning module uses a two-dimensional topology diagram to facilitate cognitive learning of the overall system composition, key equipment, and inter-equipment relationships within the coal mine safety monitoring system. The equipment simulation learning module utilizes various methods, including 2D and 3D animations, text explanations, and audio-visual materials, to provide synchronous instruction alongside the corresponding 3D equipment models. It explains the internal and external structure of the equipment, the location and function of major components, and operational methods. Simultaneously, for safety monitoring equipment such as sensors and substations, the system provides multimedia instruction through audio-visual explanations regarding usage methods and maintenance precautions. For example, it teaches sensor installation, wiring, and IP address settings.

[0044] Equipment Deployment Training Module: This module is used to conduct training on the deployment of safety monitoring system equipment. Based on the relevant provisions of the "Coal Mine Safety Regulations" and "AQ1029-2019 Coal Mine Safety Monitoring System and Testing Instrument Use and Management Specifications," and combined with local and enterprise standards, it draws equipment deployment diagrams outlining the deployment requirements of safety monitoring equipment in different areas of the coal mine (such as coal mining, tunneling faces, and intake and return airways). This diagrams are provided for safety monitoring workers to learn, helping them clearly understand the reasonable equipment deployment methods and the standards followed. Specific training content includes the setting requirements of methane sensors and other sensors, alarm concentration, power-off concentration, power-on concentration, power-off range, installation location, and installation standards. This enables the correct deployment and installation of safety monitoring equipment underground according to the standards.

[0045] Common Troubleshooting Module: This module trains safety monitoring workers to troubleshoot and resolve system faults. When faults such as exceeding limits or disconnections occur in the safety monitoring system, it enables them to quickly and accurately locate the fault point and take correct troubleshooting measures. The system lists common faults such as methane sensor disconnection, substation communication interruption, abnormal circuit breaker interlocking control, and ring network switch failure in a catalog format. For each fault phenomenon, clicking on it displays the corresponding location in the overall equipment 3D model, accompanied by corresponding 2D and 3D animations, images, and videos to describe the fault phenomenon and corresponding solutions. Users can intelligently search the fault troubleshooting knowledge base based on keywords such as fault category and fault phenomenon to accurately locate the fault point and cause.

[0046] The hands-on training module runs on a Linux operating system. Trainees operate the physical training device and complete safety operations such as installing mine sensors, calibrating mine methane sensors, setting monitoring substation parameters, and operating the host computer software of the safety monitoring system within the virtual training software. During the training, trainees must complete all assessment items within a fixed time according to the practical operation assessment standards and procedures for their specific positions. They will also interactively learn the behavioral norms and technical skills for coal mine safety monitoring operations based on the guidance and prompts. At the end of the assessment time, the system provides real-time scores for each step and the corresponding assessment results to determine whether the assessment was passed.

[0047] Example 2: This embodiment provides a virtual training method, also known as a practical assessment method, using the virtual training system for coal mine safety monitoring and control operations described in Embodiment 1. Based on standard operating procedures and operational specifications, corresponding discrimination models are established to achieve intelligent assessment of the training process. Specifically, the virtual training method flow of this invention is as follows: Figure 2 As shown, it includes the following steps: Step 1: Record all student operation logs to create a multidimensional dataset containing operation event data, time-series data, answer data, and device status parameters. Integrate all data into a single multidimensional feature vector V. t =(Operation event code, timing data value, answer score, sensor status parameter 1, sensor status parameter 2, etc.).

[0048] The operation event data in the above steps includes the actual operation content performed by the student at operation step t. The system preprocesses and extracts features from the student's operation video through a camera, and matches the corresponding operation step sequence number according to the student's operation sequence. For example, sequence number 1: cleaning the sensor and dust accumulation in the air chamber, sequence number 2: setting the sensor to enter online calibration mode, etc. If no corresponding step is identified, it means that the student has missed a key step. Time-series data represents the time spent by students in their operations, and is statistically analyzed in seconds. It is used to analyze the time spent by students in their operations and to assess the students' proficiency in their skills. The answer data represents the trainees' performance in different virtual training scenarios; it is used to assess the trainees' mastery of safety standards and regulations. A correct answer earns 1 point, and an incorrect answer earns 0 points. The equipment status data consists of the equipment status parameters when the trainee is operating step t. This includes the equipment status log during the sensor calibration operation, which contains sensor function codes, changes in their own monitored values, calibration start time, and calibration gas flow monitoring values.

[0049] Step 2: Based on the dynamic time warping algorithm, the operation steps of trainees in the process of sensor installation, calibration, and monitoring substation parameter setting are compared with the preset standard operation instruction library. The path similarity is calculated, and the operation sequence is checked to see if it is consistent with the standard sequence. When there is an error in the operation sequence, omission of key steps, or failure to meet the standard of operation, an alarm will be triggered and points will be deducted.

[0050] First, construct the distance matrix, where the standard operation instruction library sequence Q = ( , ,… , ), each It is a multidimensional feature vector, representing the first feature vector in the standard process. The ideal multidimensional state of the steps. The student's operation process sequence C = ( , ,… , ), each It is a multidimensional feature vector, representing the student's position in the th... The multi-dimensional state of actual operation at each point in time. First, construct a The matrix D, each element in the matrix This indicates the first instruction in the standard operation instruction library sequence. Points The first in the sequence of student operation process Points The weighted Euclidean distance between them is calculated using the following formula: =

[0051] in, and They are vectors and In the Eigenvalues ​​in dimension, This represents the weight of the k-th dimension. Higher weights can be assigned to key operational events or state parameters of specific devices, thus enabling more targeted differentiation of operational norms.

[0052] Step 3: Construct the cumulative distance matrix. Create a cumulative distance matrix C, whose dimensions are also... Matrix elements Indicates starting from the path origin arrive The minimum cumulative distance among all paths.

[0053] initialization: = ; First row of the matrix: = ; First column of the matrix: = ; For the other elements of the matrix Its value is determined by the following formula. =

[0054] This recursive relationship means that, upon reaching... The minimum cumulative cost of a point is equal to the local cost of that point plus the cost to its left. The one with the smallest cumulative cost among these three adjacent cells guarantees the continuity and monotonicity of the path, ensuring that each point matches and the path points move forward with each step.

[0055] Finding the optimal regularization path: Optimal regularization path =( , ,… , (From the end point) Back to the beginning A sequence of points, where each point =( The backtracking rule is: start from the endpoint. To begin, each time select one of the three adjacent cells to the left, top, or top left that makes The smallest cell is taken as the previous point in the path, until the starting point (1,1) is reached. Finally, the DTW distance between the two sequences is calculated from the endpoint element of the cumulative distance matrix. This indicates that the smaller the value, the higher the overall similarity between the two sequences.

[0056] Step 4: By measuring the Euclidean distance The scoring is converted to calculate the student's operational score for each step, which is used to assess the student's operational standardization and skill mastery level. The system employs a weighted summation algorithm to automatically evaluate and score trainees' practical exams, generating detailed diagnostic reports with corresponding deductions and improvement measures. The specific calculation formula is as follows:

[0057] The "Score" represents the student's practical exam score in safety monitoring and surveillance operations. Each step of the operation is scored to reflect whether the trainee has followed the standard operating procedures and code of conduct. The total number of steps is used to average the scores of all operation steps.

[0058] Simplified algorithm example: This embodiment uses Figure 3 Taking the three key steps of sensor accuracy calibration as an example: Step A: Open the valve of the standard gas sample bottle and introduce the standard gas sample at a small flow rate. When the sensor display value reaches 1.0, observe the sensor alarm value. Step B: When the sensor display value reaches 1.5, observe whether the actual power failure meets the requirements. Step C: Adjust the standard gas sample flow rate to 200 ml / min, so that its measurement value is displayed stably for more than 90 seconds. Adjust the sensor display value to match the standard gas sample concentration value of 2.0 using the remote control, and press the exit button to save.

[0059] In coal mine safety monitoring and control operation training, the key assessment focuses on whether the trainees' operational sequence and standardization are up to standard. Secondly, the assessment considers the trainees' answers to safety standard and specification questions, as well as the time taken for each operation. Therefore, weights are allocated to these four dimensions. The values ​​are 0.4, 0.1, 0.2, and 0.3 respectively.

[0060] The sequence is as follows = ; = ; Calculate the weighted distance matrix: =

[0061] By constructing the cumulative distance matrix and backtracking the path, the cumulative distance matrix C is obtained as follows: =

[0062] Calculated Distance is The value is 3.16. The smaller this value, the greater the similarity between the trainee's operation process and the standard operating procedure, and the higher the degree of standardization. Backtracking yields the optimal regularization path as follows: to to This indicates that the trainee's operating steps are consistent with the standard procedure.

[0063] Finally, by analyzing the Euclidean distances of the three steps in the embodiment... Perform rating conversion: ; The student's performance score for each step is calculated, and then a weighted sum is applied to obtain the total score. point.

[0064] The first step Step 2 All operations received a perfect score of 100. The third step scored only 24 points, indicating improper operation. The system intelligently outputs a detailed diagnostic report, providing corresponding deductions and improvement measures. The deductions are as follows: Step 3 Issue: Accuracy calibration failed, and ventilation stabilization time was insufficient. The standard operating procedure requires a ventilation stabilization time of 90 seconds for the standard instrument, but the trainee's actual operation time was 80 seconds, which does not meet the operating specifications. II. Problem in Step 3: The sensor reading exceeds the basic error, and the trainee did not immediately replace the sensor as required by the specifications. The sensor reading is 1.85, exceeding the basic error of 6% of the sensor's actual measurement value, which does not meet the operating specifications. The sensor should be replaced immediately, and recalibrated after a warm-up period of no less than 15 minutes.

[0065] Example 3: Typical Training Process This embodiment uses two typical scenarios, "safety calibration of a mine methane sensor" and "simulation troubleshooting of a safety monitoring system," to illustrate the application of this system in detail.

[0066] (1) Practical training on safety calibration of mine methane sensors, the process is as follows: Figure 3 As shown: Preparation before calibration: The virtual scene generates a safety monitoring worker's pre-calibration report to the dispatch room in sequence according to the process, and an animation of checking and cleaning the appearance of the sensor. At this time, the trainee confirms the working environment and working conditions in sequence according to the operation animation, and sets the sensor to enter the online calibration mode through the remote control. Sensor zero-point calibration: Following the instructions in the virtual scenario, trainees sequentially open the air cylinder valve, connect the standard air sample to the sensor chamber using a rubber hose, adjust the air flow rate to 200 ml / min, and wait for the sensor data to stabilize for more than 90 seconds. Then, use the remote control to adjust the sensor display value to zero. At this point, the zero-point calibration is complete, and the air cylinder valve is closed. During the trainee's operation, the real-time data, operating status, and real-time gas flow rate of the methane sensor are transmitted to the physical operation training module in real time, and the module judges whether the operation procedure is qualified based on the judgment model. Sensor accuracy calibration: Following the instructions in the virtual scenario, trainees sequentially open the valve of a standard gas sample bottle with a concentration of 2%, initially introducing the standard gas sample at a small flow rate. They observe the alarm and power-off values ​​as the displayed value slowly rises, simulating the continuous increase of methane gas in a coal mine. When the methane sensor reading rises above 1.0%, the sensor triggers an audible and visual alarm according to a pre-set value. When the methane sensor reading rises above 1.5%, the lighting devices and coal mining machines in the simulated underground working face in the virtual scenario trigger a power-off due to excessive methane levels. Trainees must observe the sensor alarm values, power-off values, and the actual power-off situation. Check if the requirements are met and confirm; the trainee adjusts the standard gas sample flow rate to 200ml / min, and after the measured value is displayed stably for more than 90s, observe whether the sensor display value is within the basic error range. If it is within the basic error range, adjust the sensor display value to be consistent with the standard gas sample concentration value through the remote control, press the exit button to save, close the valve of the standard gas sample bottle, and wait for the sensor detection data to drop back to 1.0%. All equipment in the underground working face in the virtual scene will automatically restore power and start working. The trainee needs to simulate and confirm the power restoration function. After confirmation, exit the calibration and use the remote control to cancel the sensor online calibration status.

[0067] Work Completion: The virtual scenario sequentially generates animations of retracting the calibration nozzle, hanging the sensor, and filling in the on-site work record, following the established process. Once the trainees have confirmed their completion, the training on the safety calibration of mine methane sensors concludes. The virtual training software's intelligent evaluation module analyzes the compliance of the calibration process based on indicators such as methane concentration, duration of methane stabilization, gas flow rate, and basic errors collected during the calibration process. It provides scores for each operational step and corresponding improvement measures, determining whether the trainees have passed the assessment. This allows trainees to master the standard operating procedures and behavioral norms for methane sensor calibration, reducing inaccurate monitoring data and false alarms caused by improper calibration.

[0068] (2) The process of simulating fault diagnosis in the security monitoring system is as follows: Figure 4 As shown: In the simulated graphical interface, trainees build a typical coal mine safety monitoring system according to the exam requirements. The connection relationships of the equipment in the system, the status of the equipment in the system, and the sensor information can be displayed on the host computer software interface of the coal mine safety monitoring system, providing trainees with user interaction between the simulation system and the host computer software.

[0069] Trainees can randomly select fault scenarios in the simulation's graphical interface. After selecting a scenario, the host computer software configures and modifies the fault phenomena of the coal mine safety monitoring system according to the fault setting description. This includes the monitoring values ​​and status of each device at each location in the simulation system. The monitoring system's measurement point definition page provides user interaction for browsing and modifying device configurations. The modified device configurations are sent to the data exchange area by the host computer software of the coal mine safety monitoring system. The data exchange area notifies the simulation equipment, and the simulation equipment displays different operating states based on the modified configuration, simulating the phenomena after a fault occurs.

[0070] Based on the description of the fault symptoms, trainees perform troubleshooting operations on the simulated equipment in the simulation system. The simulation system supports various troubleshooting methods, including remote control, wiring, and hardware module replacement, simulating the trainee's troubleshooting process. After the trainee completes the operation, the real-time data and status of the simulated equipment change immediately and are sent to the data exchange area. The data exchange area then notifies the coal mine safety monitoring system. Trainees can compare the results with the monitoring system's host computer software display to help determine whether the current status of the simulated equipment is correct and whether the trainee has followed the correct troubleshooting measures. This can assist safety monitoring workers in quickly mastering the troubleshooting solutions and correct handling measures for common faults in coal mine safety monitoring systems.

[0071] The simulation system and the coal mine safety monitoring system use the MQTT protocol for data exchange. Real-time equipment data is published by the simulation system and subscribed to by the monitoring system. Equipment configuration information is published by the coal mine safety monitoring system, and modifications to the equipment configuration are subscribed to by the simulation system.

[0072] The design of the simulation test questions is shown in Table 1.

[0073] Table 1. Fault Simulation Troubleshooting Test Design

[0074] This invention fully integrates with the realities of coal mine production, employing physical simulation and electromechanical interaction technologies to simulate actual production scenarios, standard operating procedures, common fault phenomena, and handling solutions for coal mine safety monitoring and control operations. Through highly realistic, immersive, and experiential training, it achieves intelligent teaching of operational procedures and operational compliance assessments. This equips trainees with comprehensive and solid practical skills in safety monitoring tools, enabling them to accurately master the entire process of coal mine safety monitoring and control operations, from equipment installation and commissioning to daily operation and maintenance. It significantly improves their fault diagnosis and emergency response capabilities in complex environments, enhancing practical skills and contributing to the improvement of coal mining enterprises' safety monitoring system maintenance capabilities, thus ensuring coal mine safety. This invention constructs a multi-dimensional data acquisition and weighted fusion intelligent scoring algorithm model. By collecting and processing data from the entire process of trainees' actual operational behavior, it compares the recorded operational behavior with standard operating procedures, transforming subjective and vague operational evaluations into objective and accurate data scores. This comprehensive and objective assessment avoids interference from human factors. It not only provides the final practical results, but also accurately identifies the specific problems students encounter in terms of operational procedures, standardization, efficiency, and decision-making logic, providing a precise and in-depth diagnosis to promote learning through testing.

[0075] Based on the above-described preferred embodiments of the present invention, and through the foregoing description, those skilled in the art can make various changes and modifications without departing from the inventive concept. The technical scope of this invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.

Claims

1. A virtual training system for coal mine safety monitoring and control operations, characterized in that, The device comprises a physical operation training device and a computer; The physical operation training device is used for simulating a coal mine safety monitoring device; the computer is installed with virtual training software; the physical operation training device and the virtual training software in the computer interact with each other through a communication protocol; The physical operation training device comprises a monitoring and control physical operation device and a mine sensor calibration practical training device; the monitoring and control physical operation device collects real-time monitoring data and operating parameters through a plurality of sensors and outputs them; the mine sensor calibration practical training device provides standard gas samples required for calibration for mine gas sensors and monitors gas flow during the calibration process; The virtual training software comprises a virtual training scene construction module, a monitoring and control data collection module and an intelligent evaluation module; the real-time monitoring data and operating parameters collected by the physical operation training device are synchronously output to the virtual training software through the monitoring and control data collection module to train and examine trainees; The virtual training scene construction module is used for constructing a three-dimensional virtual environment in a coal mine; the monitoring and control data collection module is used for collecting multi-dimensional real-time operation data of trainees from the physical operation training device to drive the three-dimensional virtual environment to dynamically change with the collected multi-dimensional real-time operation data; the intelligent evaluation module comprises a sequence alignment unit and a score calculation unit; the sequence alignment unit is configured to use a dynamic time warping algorithm to align a data time sequence of trainees in a safety monitoring and control operation with a time sequence of a standard operation process of a safety monitoring and control operation; the score calculation unit is configured to output an operation evaluation score based on an alignment result of the sequence alignment unit by using a weighted scoring model.

2. The virtual training system for coal mine safety monitoring and control operation of claim 1, wherein: The physical operation training device and the virtual training software interact with each other through Modbus and MQTT communication protocols.

3. The virtual training system for coal mine safety monitoring and control operation of claim 1, wherein: The monitoring and control physical operation device comprises a monitoring substation, a ring network access device, a mine methane sensor, a remote control switch, a mine intrinsic safety circuit junction box and an infrared remote controller for system operation, collects real-time monitoring data and operating parameters of each sensor and outputs them to the monitoring and control data collection module.

4. The coal mine safety monitoring and control operation virtual training system of claim 3, wherein: The mine sensor calibration practical training device comprises a gas cylinder assembly, a gas flow sensor and a methane sensor and is used for trainees to perform safety calibration training of methane sensors; The gas cylinder assembly comprises a standard methane gas cylinder, an air gas cylinder and a pressure reduction control valve connected with the standard methane gas cylinder and the air gas cylinder; the pressure reduction control valve is used for controlling the standard methane gas cylinder and the air gas cylinder to output gas with different flow rates to provide standard methane gas samples and air gas samples for methane sensor calibration training; the gas flow sensor is used for monitoring real-time values of gas flow during sensor calibration to determine whether trainees operate according to standard required gas sample flow.

5. The virtual training system for coal mine safety monitoring and control operation of claim 1, wherein: The virtual training software further comprises a theoretical training module, which comprises a three-dimensional equipment simulation model, model interactive animation and multimedia teaching materials, and is used for theoretical knowledge training of the overall composition of the coal mine safety monitoring system, equipment operation mode and fault diagnosis.

6. The virtual training system for coal mine safety monitoring and control operation of claim 1, wherein: The virtual training software further comprises a physical operation training module, which is used for guiding trainees to interactively learn practical operation skills based on the standard operation process logic of the coal mine safety monitoring system and common system troubleshooting schemes.

7. The virtual training system for coal mine safety monitoring and control operation of claim 1, wherein: The virtual training scene construction module constructs a three-dimensional virtual environment in a coal mine, which is designed according to the training needs of the coal mine safety monitoring worker and comprises a three-dimensional regional scene model of a coal mining face, a tunneling face and a return airway, a two-dimensional system topology graph model of the coal mine safety monitoring system and a three-dimensional equipment model of safety monitoring and control equipment in the system.

8. A method for evaluating the actual operation of the coal mine safety monitoring and control operation virtual training system according to any one of claims 1-7, characterized in that, The steps include collecting monitoring and control data and analyzing trainee practical operation process data: The step of collecting monitoring and control data: through the physical operation examination and training device and the virtual training software, multi-dimensional real-time operation data of the trainee in the practical operation training process are collected, the multi-dimensional real-time operation data including at least two of the following four types: operation event data, time sequence data, answer data and equipment state data; The step of analyzing trainee practical operation process data includes a sequence comparison step and an evaluation scoring step; The sequence comparison step: based on a dynamic time warping algorithm, the sequence of the trainee's operation process in the safety monitoring and control operation is compared with the standard operation instruction sequence of the safety monitoring and control operation, and the similarity between the sequences is calculated; The evaluation scoring step: according to the similarity, an operation evaluation score of the trainee is calculated through a weighted scoring model, the weighted scoring model converts the similarity into operation step scores, and the final operation evaluation score is generated after weighted summation.

9. The method of claim 8, wherein the method further comprises: determining the virtual training score based on the virtual training score and the virtual training score of the virtual training simulation; and determining the virtual training score based on the virtual training score and the virtual training score of the virtual training simulation. In the sequence comparison step, the step of collecting monitoring and control data includes: integrating the collected multi-dimensional real-time operation data into a multi-dimensional feature vector; The sequence alignment step includes constructing a distance matrix between the standard operation instruction sequence and the student operation procedure sequence, and constructing a cumulative distance matrix and finding an optimal regularized path; elements in the distance matrix are weighted Euclidean distance Computing.

10. The method of claim 9, wherein the method further comprises: determining the virtual training score based on the virtual training score and the real operation score. In the evaluation scoring step, the Euclidean distance is calculated to obtain a score item of each operation step through a scoring conversion function; the weighted scoring model adopts a weighted summation algorithm to evaluate and score the trainee's actual operation examination, output a diagnosis report, and the calculation formula is: ； ； Wherein Score represents the score of the student safety monitoring and control operation test; is the score item for the i-th operation step; is the total number of steps, used for averaging the scores of all operation steps.

Images