VR-based gas safety emergency training system

Abstract

The application relates to the technical field of training simulation, and discloses a VR-based gas safety emergency training system, which comprises a three-dimensional scene construction module, an interactive data acquisition module, an interactive data processing module, an examination data analysis module and an examination evaluation module; the application replaces entity drills, the same scene can be repeatedly trained unlimited times, and training cost is significantly reduced; VR technology provides multi-sensory stimulation, enhances the intuitive cognition of students to dangers, effectively improves safety awareness and emergency response ability, and has no danger; by collecting student operation data, the system automatically calculates key coefficients such as equipment operation precision, process compliance, disposal timeliness and path planning rationality, realizes scientific quantification and accurate scoring of training effect; according to the scoring results, automatic grading evaluation is carried out, and a customized remedial module is triggered for students with insufficient ability, a "training-examination-intervention" closed loop is formed, and the overall training efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of training simulation technology, and more specifically to a VR-based gas safety emergency training system. Background Technology

[0002] Traditional employee safety and emergency training is limited by various objective conditions such as venue, instructors, capabilities, and time. Most of it relies on theoretical knowledge or video observation. To avoid risks during emergency drills, it often only provides simple simulations, which are far removed from reality. Companies spend a significant amount of energy, time, and resources on this training, which is often tedious and ineffective. Specifically, it has the following shortcomings:

[0003] Physical drills need to realistically simulate gas leaks or fire scenarios, which pose risks such as burns to personnel and injuries from explosions, and can also cause equipment damage.

[0004] It requires the construction of a dedicated training ground, which consumes consumables such as gas and fire extinguishers, resulting in high costs per training session and the inability to reuse the equipment.

[0005] Traditional training relies on verbal explanations or two-dimensional diagrams, lacking multi-sensory stimulation. Trainees' understanding of dangerous scenarios remains at the theoretical level, resulting in limited improvement in emergency response capabilities.

[0006] Relying on manual observation and scoring makes it difficult to quantify key indicators such as operational accuracy and timeliness, and is easily affected by the subjective judgment of the evaluator;

[0007] Traditional assessments fail to generate personalized remedial plans, leaving students with weaker abilities without targeted reinforcement training, resulting in low overall training efficiency.

[0008] Therefore, a method that is free from safety hazards, low in cost, multi-dimensional in perception, scientifically evaluated, and allows for closed-loop improvement is needed to solve the above problems. Summary of the Invention

[0009] In order to overcome the above-mentioned defects of the prior art, the present invention provides a VR-based gas safety emergency training system to solve the problems existing in the background art.

[0010] To achieve the above objectives, the present invention provides the following technical solution: a VR-based gas safety emergency training system. This VR-based gas safety emergency training system creates virtual scenarios such as gas leaks and fire escapes through a 3D scene construction module, supporting multi-scenario switching and parameter adjustment; an interactive data acquisition module records trainees' operational behavior in real time; an interactive data processing module analyzes key indicators such as operational accuracy, process compliance, timeliness, and path planning; an assessment data analysis module comprehensively calculates and generates training scores; and an assessment and evaluation module classifies training based on scores and implements intervention measures, forming a closed-loop training system, specifically including:

[0011] 3D scene construction module: used to build virtual 3D scenes of gas leaks and fire escape, supporting intelligent switching between multiple scenes and real-time parameter adjustment;

[0012] Interactive data acquisition module: used to collect multimodal data of trainees' interactive operations in 3D scenes during safety emergency training;

[0013] Interactive data processing module: Cleans and processes the collected interactive data, and calculates the equipment operation accuracy coefficient, process compliance coefficient, handling timeliness coefficient and path planning coefficient.

[0014] The calculation of the equipment operation accuracy coefficient is as follows:

[0015] ;

[0016] Where S is the equipment operation accuracy coefficient, with a value range of 0-1, where 1 indicates perfect compliance with the standard;

[0017] F is the valve operation accuracy coefficient, which is the weighted sum of the rotation angle fraction and the switching timing fraction; where the rotation angle fraction = 1 - |actual number of turns - standard number of turns| / standard number of turns; the switching timing fraction is based on whether the valve closing operation is completed within the specified time, with full marks of 1 point for timely completion and deductions for closing delays.

[0018] M is the fire extinguisher operation accuracy coefficient, which is the weighted sum of the tilt angle score and the spray coverage score. The tilt angle deviation score is judged by whether the deviation angle exceeds the limit. If it is less than or equal to the set deviation value, the full score is 1 point, and if it is greater than the set deviation value, the score is 0. The spray coverage score is judged by the proportion of fire source covered. Coverage x*100% gets x points, and if the coverage is less than 50%, the score is 0.

[0019] G represents the tool operation accuracy, which assesses the standardization of pipeline maintenance operations. It is a weighted sum of torque error score and disassembly sequence score. Torque error score = 1 - |actual torque - standard torque| / standard torque; disassembly sequence score: full marks of 1 point for correct steps, and deductions proportionally based on the number of incorrect steps for errors or omissions.

[0020] The specific calculation of the process compliance coefficient is as follows:

[0021] ;

[0022] Where L is the process compliance coefficient, B is the step accuracy coefficient, the standard process is "valve shut-off - ventilation - detection - alarm - evacuation", and the step accuracy coefficient = Σ(step weight × operation accuracy rate);

[0023] W represents the protective equipment coefficient, which is the wearing integrity coefficient, the tightness compliance rate, and the wearing response time weighted sum. The wearing integrity coefficient is calculated as Σ (equipment weight × correct wearing rate). A respirator leakage rate of less than 5% earns 1 point, 5%-10% earns 0.5 points, and more than 10% earns 0 points. The time from alarm to equipment wearing completion is ≤60 seconds, earning 1 point, and deducting 0.1 points for every 10 seconds exceeding this limit.

[0024] The specific calculation of the disposal timeliness coefficient is as follows:

[0025] ;

[0026] Where C is the processing timeliness coefficient and Y is the response delay compliance coefficient, specifically:

[0027] ;

[0028] Where Y is the response latency compliance coefficient, t s To extend the actual time, t b The standard delay time is used as the benchmark. For every second the actual delay exceeds the standard value, 0.1 points will be deducted, up to a minimum of 0 points.

[0029] D represents the process deviation compliance coefficient, specifically:

[0030] ;

[0031] Where D is the process deviation compliance coefficient, and T s T represents the actual duration, the total time actually consumed by the operation process; b The standard duration is defined as follows: The min function limits the deviation ratio to an upper limit of 1. When the deviation ratio is greater than 1, min outputs 1 and D=0. D=1 indicates that the actual duration is equal to the standard duration, with zero deviation and full compliance. D close to 0 indicates that the deviation is too large and the compliance is low.

[0032] E is the emergency correction factor, which equals the correction quality factor multiplied by the standard correction time, and is calculated as follows: Emergency Correction Factor = Correction Quality Factor * Standard Correction Time / Actual Correction Time.

[0033] H is the interruption recovery capability coefficient. When the actual recovery time is ≤30 seconds, H is 1, which is considered fully compliant. When the actual recovery time is >30 seconds, the interruption recovery capability coefficient = 1 - (actual recovery time - 30) / 60.

[0034] The path planning coefficients are calculated as follows:

[0035] ;

[0036] Where J is the path planning coefficient, and its value ranges from 0 to 1, where 1 represents the optimal value;

[0037] P is the path deviation compliance coefficient, specifically:

[0038] ;

[0039] Where P is the path deviation compliance coefficient, P a (t) represents the student's three-dimensional spatial coordinates at time t, P o (t) represents the three-dimensional coordinates of the optimal escape path pre-generated by the system at time t; T is the total path tracking time, and D... max The maximum allowable deviation threshold is set; if the deviation exceeds the threshold, P is forced to 0, and the operation is directly judged as a failure.

[0040] X is the collaboration accuracy coefficient, specifically:

[0041] ;

[0042] Where X is the cooperation accuracy coefficient, R is the number of correctly executed instructions, Z is the total number of instructions, Δt is the average instruction response latency, and t max For the maximum allowable delay time, 1 - Δt / t max As a delay penalty factor, its value ranges between 0 and 1. If the response is timely, it is close to 1, and if the delay exceeds the limit, it approaches 0.

[0043] Q represents the path security compliance coefficient, specifically:

[0044] ;

[0045] Where Q is the path safety compliance coefficient, d(t) is the real-time three-dimensional spatial distance between the user and the nearest obstacle or hazard at time t; min(d(t)) represents the minimum distance between the user and the obstacle or hazard during the entire escape process. a This represents the preset minimum safe distance threshold. The min function constrains the upper limit of Q to 1, when the minimum distance ≥ d. a When Q=1, it represents the optimal state; otherwise, points are deducted proportionally.

[0046] The assessment data analysis module is used to integrate and calculate the processed interaction coefficients to generate a safety emergency training score.

[0047] The specific scoring criteria for safety and emergency training are as follows:

[0048] ;

[0049] Where A is the comprehensive score for emergency training, S is the equipment operation accuracy coefficient, L is the process compliance coefficient, C is the handling timeliness coefficient, and J is the path planning coefficient;

[0050] Assessment module: Used to grade and evaluate safety emergency training scores, and to take training intervention measures for trainees whose scores are insufficient.

[0051] The technical effects and advantages of this invention are as follows:

[0052] 1. This invention simulates high-risk scenarios such as gas leaks, fires, and explosions in a virtual environment, allowing trainees to conduct emergency response training in a state without actual danger, thus avoiding personal injury and equipment damage caused by real drills.

[0053] 2. This invention replaces the venues, equipment and consumables required for physical training, significantly reducing training costs; the same scenario can be repeated an unlimited number of times, supporting standardized assessment of large-scale personnel.

[0054] 3. The VR technology of this invention provides multi-sensory stimulation of vision, hearing and touch, enhances trainees' intuitive understanding of danger, and effectively improves safety awareness and emergency response capabilities.

[0055] 4. This invention collects trainee operation data, and the system automatically calculates key coefficients such as equipment operation accuracy, process compliance, handling timeliness, and path planning rationality, thereby achieving scientific quantification and accurate scoring of training effectiveness.

[0056] 5. This invention automatically grades and evaluates trainees based on their scores, and triggers customized remedial modules for trainees with insufficient abilities, forming a closed loop of "training-assessment-intervention" to improve overall training efficiency. Attached Figure Description

[0057] Figure 1 This is a structural block diagram of the present invention.

[0058] Figure 2 This is a flowchart of the present invention. Detailed Implementation

[0059] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. In addition, the forms of the various structures described in the following embodiments are merely illustrative. The automatic unloading device for rotary kiln with self-cooling function involved in the present invention is not limited to the structures described in the following embodiments. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] Reference Figure 1 This invention provides a VR-based gas safety emergency training system, including a three-dimensional scene construction module, an interactive data acquisition module, an interactive data processing module, an assessment data analysis module, and an assessment and evaluation module.

[0061] Reference Figure 2 The specific implementation steps of the present invention include the following steps:

[0062] S1. Construct virtual 3D scenarios for gas leaks and fire escape, supporting intelligent switching between multiple scenarios and real-time parameter adjustment.

[0063] It should be specifically noted that the 3D scene construction module uses existing 3D modeling tools to create detailed models of virtual environments for gas leaks and fire escape, simulating the dynamic effects of smoke diffusion and fire spread. It also supports intelligent switching between multiple scenes and real-time adjustment of wind direction and temperature parameters, providing comprehensive digital virtual simulation training for the gas industry in areas such as risk identification, hazard investigation, hazardous operation management, and emergency rescue practices.

[0064] S2. Collect multimodal data of trainees' interactions with 3D scenes during safety emergency training.

[0065] It should be specifically noted that, based on the VR headset and touch interaction system, the interactive data acquisition module captures real-time multimodal operation data of the learner in the virtual scene, including device operation, protective gear wearing, decision-making paths, action sequences, and position coordinates, providing raw input data for subsequent analysis.

[0066] S3. Clean and process the collected interactive data, and calculate the equipment operation accuracy coefficient, process compliance coefficient, disposal timeliness coefficient, and path planning coefficient.

[0067] It should be noted that the calculation of the equipment operation accuracy coefficient is as follows:

[0068] ;

[0069] Where S is the equipment operation accuracy coefficient, which quantitatively evaluates the accuracy and standardization of trainees' operation of key equipment valves, fire extinguishers and pipeline tools in the virtual environment. The value ranges from 0 to 1, where 1 indicates perfect compliance with the standard.

[0070] F is the valve operation accuracy coefficient, which evaluates the standardization of valve closing operation. It is a weighted sum of the rotation angle fraction and the opening / closing sequence fraction.

[0071] The rotation angle fraction = 1 - |actual number of rotations - standard number of rotations| / standard number of rotations represents the matching degree between the actual number of rotations and the standard number of rotations. It should be noted that the number of rotations here can be a decimal. The rotation angle error directly affects the valve's sealing performance. Angle deviation leads to an increase in the gas leakage risk index, so it is given a higher weight, ranging from 0.55 to 0.6.

[0072] The timing score is based on whether the valve closing operation is completed within the specified time. Full marks of 1 point are awarded for timely completion, and points are deducted proportionally for closing delays. Timing errors can be remedied through subsequent emergency steps, with a weighting range between 0.4 and 0.45.

[0073] Valve operation errors are the main cause of gas leaks and the primary cause of household gas accidents. Gas valves must strictly follow specifications such as "rotate 3 turns to confirm locking," otherwise the system cannot trigger safety confirmation. Therefore, it is given the highest weight, and the value of the weight ωf is between 0.45 and 0.5.

[0074] M is the fire extinguisher operation accuracy coefficient, which assesses the standardization of fire extinguisher use. It is a weighted sum of the tilt angle score and the spray coverage score.

[0075] The criteria for judging the tilt angle deviation score is to see if the deviation angle exceeds the limit. If it is less than or equal to the set deviation value, the full score is 1 point. If it is greater than the set deviation value, the score is 0 points. That is, if the deviation exceeds the limit, it is directly judged as an error. The spray coverage score is judged based on the proportion of fire source covered. Coverage x*100% gets x points. It should be noted that if the coverage is less than 50%, the score is 0 points. The impact of angle error and insufficient coverage on the fire extinguishing result is equally weighted, so equal weighting is used.

[0076] The operation of fire extinguishers directly affects the initial fire control effect. Gas fires spread rapidly, and operational errors can lead to the fire getting out of control. The value of the weight ωm ranges from 0.25 to 0.3.

[0077] G represents the tool operation accuracy, which assesses the standardization of pipeline maintenance operations, and is a weighted sum of the torque error score and the disassembly sequence score.

[0078] The torque error fraction = 1 - |actual torque - standard torque| / standard torque represents the degree of matching between the actual torque and the standard value. The torque accuracy of pipeline maintenance tools directly affects the sealing performance. Insufficient torque will lead to gas leakage, while excessive torque will damage the equipment threads or sealing surfaces, inducing the risk of explosion. Therefore, it has a high weight and the value range is between 0.65 and 0.7.

[0079] The order of disassembly is scored, with a full score of 1 point for correct steps and a deduction of points proportional to the number of incorrect or omitted steps. In certain scenarios, incorrect order may trigger a chain of risks, but such errors can usually be corrected in time through system prompts, with a weighting range between 0.3 and 0.35.

[0080] Pipeline maintenance requires precise torque control to avoid equipment damage or seal failure. Incorrect disassembly sequence can lead to secondary leakage risks. The weight ωg ranges from 0.2 to 0.25.

[0081] It should be noted that the calculation of the process compliance coefficient is as follows:

[0082] ;

[0083] Where L is the process compliance coefficient, which reflects the trainee's mastery of the process by verifying the trainee's operation steps and the wearing of protective equipment.

[0084] B represents the step accuracy coefficient. The standard procedure is "valve shut-off - ventilation - detection - alarm - evacuation". The step accuracy coefficient = Σ (step weight × operation accuracy rate), and the weights are allocated according to the risk level.

[0085] The primary action is to shut off the gas source to prevent continuous leakage from causing combustion / explosion. Failure to do so significantly increases the risk of subsequent operational failures. The probability of explosion is higher if valves are not closed in the early stages of a leak; therefore, valve closure has the highest weight, ranging from 0.3 to 0.35. Next, reduce the gas concentration to below the lower explosive limit. Regulations require that the ventilation system be turned on within 15 seconds of fire zone confirmation. Delayed ventilation will cause flammable gas to accumulate, increasing the risk of reaching the explosive limit concentration. Ventilation has a weight between 0.25 and 0.3. Verifying environmental safety determines subsequent actions. Methane and oxygen concentrations must be tested. Operations other than evacuation are prohibited if these levels are not met. Incorrect test data can cause secondary accidents. This has a weight between 0.15 and 0.2. Regulations require a two-level alarm mechanism. Delayed alarms will delay external rescue and expand the scope of the accident's impact. This has a weight between 0.1 and 0.15. When the detected concentration exceeds the limit or the leak cannot be controlled, evacuation must be completed within 30 seconds. This has a low weight because it is a passive avoidance action, but failure to do so will directly lead to casualties. This has a weight between 0.1 and 0.15.

[0086] Emergency procedures directly impact the speed of accident spread. Statistics show that standardized procedures can reduce accident losses, with the weight ωb of the accuracy coefficient for each step ranging from 0.6 to 0.7.

[0087] W represents the protective equipment coefficient, which is the wearing integrity coefficient, the fit compliance rate, and the weighted sum of wearing response time.

[0088] The incompleteness coefficient is calculated as Σ(equipment weight × correct wearing rate). The equipment weight is determined by the importance of the equipment. The weight of respiratory protective equipment is between 0.25 and 0.35; the weight of head protective equipment is between 0.15 and 0.25; the weight of torso protective equipment is between 0.2 and 0.25; the weight of hand and foot protective equipment is between 0.15 and 0.2; and the weight of eye and face protective equipment is between 0.05 and 0.15. The weight is dynamically adjusted according to different scenarios. Incompleteness of wearing is the main factor, with a weight of 0.5 to 0.6.

[0089] A respirator leakage rate of less than 5% earns 1 point, 5%-10% earns 0.5 points, and more than 10% earns 0 points; the weighting is between 0.2 and 0.25.

[0090] 1 point is awarded for a time of ≤60 seconds from alarm activation to equipment donning, and 0.1 points are deducted for every 10 seconds exceeding this time; the weighting is between 0.2 and 0.25.

[0091] Equipment integrity is a prerequisite for entering hazardous areas, but it is a basic capability with a weight ωw between 0.3 and 0.4. The weight of ωw can be increased during leak escalation.

[0092] It should be noted that the calculation of the disposal timeliness coefficient is as follows:

[0093] ;

[0094] Where C is the timeliness coefficient, which quantifies the effectiveness of the response action within the time window.

[0095] Y is the response latency compliance coefficient, specifically:

[0096] ;

[0097] Where Y is the response delay compliance coefficient, which assesses whether the time from the triggering of a system alarm to the user's first action complies with security regulations.

[0098] ts represents the actual delay time, and tb represents the standard delay time. The standard delay time is as follows: ≤10 seconds for gas leak scenarios, which is the response time limit after the combustible gas alarm is triggered; ≤3 seconds for emergency shut-off operations, such as valve closure, which must be activated within 3 seconds after the alarm; other scenarios are dynamically set according to the equipment type; 0.1 points are deducted for every second the actual delay exceeds the standard value, up to a minimum of 0 points.

[0099] Response delay compliance directly affects the initial risk control capability. A leak response delay of 10 seconds can double the risk of explosion, accounting for the highest weight. The weight ωy ranges from 0.35 to 0.4.

[0100] D represents the process deviation compliance coefficient, specifically:

[0101] ;

[0102] Where D is the process deviation compliance coefficient, which quantifies the degree of deviation between the total process operation time and the standard time.

[0103] Ts represents the actual duration, the total time actually consumed in the operation; Tb represents the standard duration, the preset standard time, dynamically set based on safety regulations or industry benchmarks; the standard duration is set as follows: Leakage handling process ≤ 180 seconds, including valve closure, ventilation, and detection steps; fire extinguishing operation ≤ 60 seconds, the time requirement for fire extinguisher spray coverage.

[0104] |Actual Duration − Standard Duration| represents the absolute value of the duration error, ignoring the directionality of being ahead or behind; Deviation Ratio converts the absolute deviation into a ratio relative to the standard duration, facilitating cross-scenario comparison; The min function limits the upper limit of the deviation ratio to 1, ensuring that D≥0. When the deviation ratio>1, min outputs 1 and D=0.

[0105] D=1 indicates that the actual duration is equal to the standard duration, with zero deviation and full compliance; D close to 0 indicates that the deviation is too large, the actual duration far exceeds the standard, and the compliance is low.

[0106] End-to-end timeouts will amplify the impact of the incident; a balance must be struck between efficiency and operational integrity. The weight ωd should range from 0.25 to 0.3.

[0107] E is the emergency correction coefficient, which assesses the speed and effectiveness of correcting errors in emergency response. It quantifies the efficiency from error identification to full restoration of safety, and the higher the value, the stronger the correction capability.

[0108] Emergency correction coefficient = correction quality coefficient * standard correction time / actual correction time. The standard correction time is the ideal time threshold preset based on the scenario. The efficiency decreases when the threshold is exceeded. The actual correction time is the actual time from when the user discovers the error to when the correction is completed. The correction quality coefficient is assigned according to the degree of risk residue. It is 1.0 for completely eliminating the risk and 0.8 for partial control, to ensure that the result includes a safety integrity assessment.

[0109] The correction speed determines the probability of secondary accidents; standard correction timeouts significantly increase cascading risks. The weight ωe ranges from 0.2 to 0.25.

[0110] H represents the interruption recovery capability coefficient, which quantifies the timeliness of recovery operations after a sudden interruption; the interruption scenario includes equipment failure and new risk emergencies; the recovery actions include resetting the safety status and restarting the operation process.

[0111] When the actual recovery time is ≤30 seconds, H is 1. The standard response time requirement for the gas emergency shut-off device is ≤30 seconds. If the operation can be restored within this time limit after the interruption, it is judged to be fully compliant. When the actual recovery time is >30 seconds, the interruption recovery capability coefficient = 1 - (actual recovery time - 30) / 60. The deduction slope after the timeout is 1 / 60 points for every second overdue. This is derived from the tolerance ratio of the standard total operation time of the whole process ≤180 seconds, that is, the interruption recovery time accounts for no more than 1 / 3 of the total process.

[0112] Interruption recovery is a low-frequency event, but recovery timeouts can exacerbate secondary disasters. The weight ωh ranges from 0.1 to 0.15.

[0113] It should be noted that the path planning coefficients are calculated as follows:

[0114] ;

[0115] J is the path planning coefficient, which comprehensively evaluates the rationality of the escape route selection and the accuracy of the collaborative task execution. Its value ranges from 0 to 1, with 1 indicating the optimal value.

[0116] P is the path deviation compliance coefficient, specifically:

[0117] ;

[0118] Where P is the path deviation compliance coefficient, which quantifies the spatial deviation between the user's actual path and the optimal path.

[0119] Pa(t) represents the three-dimensional spatial coordinates of the trainee at time t. The actual path trajectory sequence is formed by collecting user location information in real time through VR positioning devices. Po(t) is the three-dimensional coordinates of the optimal escape path pre-generated by the system at time t. It should be noted that the optimal escape path here is based on the algorithm in the existing technology, taking into account path length, avoidance of dangerous sources such as gas leaks, and complex terrain factors. The numerator in the formula is the Euclidean norm, which calculates the straight-line distance between two points and quantifies the real-time spatial deviation between the user's position and the optimal path.

[0120] T represents the total time for path tracking, ensuring that deviation calculations cover the escape process and preventing partial path deviations from affecting the overall assessment; Dmax represents the maximum allowable deviation threshold, which serves as a standardized deviation value. When the threshold is exceeded, P is forced to 0, and the operation is directly judged as a failure.

[0121] The deviation of the escape route directly determines the emergency response efficiency. Three-dimensional spatial coordinate tracking shows that deviation from the optimal route exceeding the threshold will significantly increase escape time and risk. High weight reflects its fundamental impact on overall safety, and the weight ωp ranges from 0.45 to 0.5.

[0122] X is the collaboration accuracy coefficient, specifically:

[0123] ;

[0124] Where X is the collaboration accuracy coefficient, which quantifies the consistency and accuracy of instruction execution in multi-user collaboration, ensuring that team operations conform to preset specifications.

[0125] R represents the number of correctly executed instructions, indicating the number of instructions correctly executed by the user team in the virtual environment. For example, after the master trainee issues the instruction to "close the valve," other trainees accurately complete the operation. Z represents the total number of instructions issued by the system in the training scenario, including both correct and incorrect executions. The ratio represents the accuracy of instruction execution, directly reflecting the team's ability to coordinate key operations. The higher the ratio, the better the team's consistency in collaboration.

[0126] △t is the average response delay time of the instruction, that is, the average time difference from the issuance of the instruction to the user's first operation; tmax is the maximum allowable delay time, representing the maximum allowable delay time threshold, that is, the response time limit set by the system; the ratio is the timeliness dynamic correction, which identifies potential risks caused by communication delays and quantifies the severity of the delay, with a value range between 0 and 1; 1-△t / tmax is used as a delay penalty factor, with a value range between 0 and 1, approaching 1 if the response is timely, and approaching 0 if the delay exceeds the limit.

[0127] In multi-user scenarios, the accuracy of instruction transmission affects the consistency of team response, and erroneous execution leads to a chain of risks. The weight is secondary because it depends on team collaboration rather than individual ability. The value of the weight ωx ranges from 0.3 to 0.35.

[0128] Q represents the path security compliance coefficient, specifically:

[0129] ;

[0130] Q represents the path safety compliance coefficient, which refers to ensuring that the trainee's movement trajectory conforms to the preset safety specifications during path planning. The safety level is quantified by real-time monitoring of the distance between the trainee and hazardous sources such as obstacles and leaks.

[0131] d(t) is the real-time three-dimensional spatial distance between the user and the nearest obstacle or hazard at time t; min(d(t)) represents the minimum distance between the user and the obstacle or hazard during the entire escape process, identifying the critical point closest to the hazard on the path and reflecting the overall safety of the path; da represents the preset minimum safe distance threshold, setting a safety boundary based on scenario risk; the min function constrains the upper limit of Q to 1 to avoid inflated scores due to distances far exceeding the threshold. When the minimum distance ≥ da, Q = 1 represents the optimal state; otherwise, points are deducted proportionally.

[0132] While the safe distance between the path and the hazard is important, it can be predicted and avoided using a VR physics engine, offering a large dynamic adjustment space. Therefore, it has the lowest weight, with the weight ωq ranging from 0.2 to 0.25.

[0133] S4. Perform fusion calculation on the processed interaction coefficients to generate a comprehensive score for safety emergency training.

[0134] The specific scoring criteria for safety and emergency training are as follows:

[0135] ;

[0136] A represents the comprehensive score for emergency training, which assesses the emergency response capabilities of trainees through a multi-dimensional coefficient evaluation.

[0137] S is the equipment operation accuracy coefficient, which directly determines the effectiveness of various equipment controls. If errors occur, they can lead to escalation of accidents. Most accidents are directly caused by improper equipment operation, so it has the highest weight. The weight α ranges from 0.35 to 0.4.

[0138] L represents the process compliance coefficient. Omission of key steps or incorrect step sequence can lead to systemic risks. Standardizing processes can reduce the probability of secondary accidents. Process errors are prone to triggering chain failures, so they have a secondary weight. The weight β ranges from 0.25 to 0.3.

[0139] C represents the response time coefficient. Response time directly affects the golden rescue window. In leakage accidents, time delays increase the risk of deflagration. Timeliness and risk are exponentially related, but timeliness is not a direct cause. Therefore, the weight γ ranges from 0.15 to 0.2.

[0140] J is the path planning coefficient. If the path deviates from the optimal route or the safe distance is insufficient, the risk of casualties will be significantly increased. A reasonable path can improve evacuation efficiency. The weight focuses on optimizing the survival rate, and the value of the weight δ is between 0.15 and 0.2.

[0141] S5. Conduct a graded assessment of the comprehensive score of safety emergency training, and take training intervention measures for trainees whose scores are insufficient.

[0142] The comprehensive evaluation of safety and emergency training is conducted in a tiered manner as follows:

[0143] When a1 < A ≤ 100, it is considered excellent.

[0144] When a2 < A ≤ a1, it is a good grade;

[0145] When a3 < A ≤ a2, it is of medium grade;

[0146] When a4 < A ≤ a3, it is a passing grade;

[0147] When 0 < A ≤ a4, it is a failing grade;

[0148] The values ​​of a1 are between 89 and 90, a2 are between 79 and 80, a3 are between 69 and 70, and a4 are between 59 and 60.

[0149] The specific training intervention measures for trainees with insufficient qualifications are as follows:

[0150] Outstanding trainees will be given priority in participating in high-risk emergency response missions, granted on-site command authority, and recommended to participate in national-level emergency rescue worker qualification certification.

[0151] Award bonuses or training resources, such as advanced skills training opportunities, and include them in the core candidates for annual performance evaluation;

[0152] Served as an internal training instructor, recording standardized operation videos for the team to learn from.

[0153] For high-achieving trainees, targeted reinforcement will be provided; for those with weaknesses, such as insufficient precision in equipment operation, VR simulation training will be conducted, with retraining once a month until the standards are met.

[0154] Participate in real-world scenarios and be assigned to lead secondary tasks, such as material allocation, to accumulate experience in complex situations;

[0155] Second-class bonuses will be awarded, and priority will be given to recommending social emergency response forces for qualification assessment.

[0156] For intermediate trainees, mandatory retraining will be conducted, requiring them to participate in full-process closed training away from their posts, focusing on improving equipment operation accuracy and process compliance. Those who fail the assessment will be downgraded.

[0157] Reduce privileges and suspend high-risk operation qualifications, such as gas valve operation, allowing only auxiliary tasks;

[0158] Monthly performance bonus deductions will result in disqualification from annual performance evaluations.

[0159] For trainees who pass the test, they will undergo special training and participate in "one-on-one" corrective training. They will also have to practice for an additional 2 hours every day and submit an operation log, which will be signed and confirmed by the instructor.

[0160] Arrange emergency psychological stress resistance courses to eliminate hesitation or fear in operation;

[0161] A warning will be given, and if the score is still ≤a3 after reassessment, the individual will be transferred from the emergency response post.

[0162] For those who fail the qualification test, their positions will be frozen immediately, all emergency-related duties will be suspended, and they will be required to retake the pre-job qualification training.

[0163] Individuals sign an improvement commitment letter and submit weekly training reports; team instructors bear joint responsibility, and departmental contingency plans are re-filed.

[0164] Employees who fail n consecutive evaluations will have their employment contracts terminated or be reassigned in accordance with the law.

[0165] Through the above description of the embodiments, those skilled in the art can clearly understand that the various embodiments of this application can be implemented by means of software or software combined with necessary general-purpose hardware platforms, and of course, they can also be implemented by hardware functions. Based on this understanding, the technical solution of this application, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. The software product is stored in a storage medium and includes several instructions to cause a computer device, such as including but not limited to a personal computer, server, or network device, to execute all or part of the steps of the method described in any embodiment of this application.

[0166] The foregoing describes exemplary embodiments of this application. It should be understood that the above exemplary embodiments are not restrictive but illustrative, and the scope of protection of this application is not limited thereto. It should be understood that those skilled in the art can make modifications and variations to the embodiments of this application without departing from the spirit and scope of this application, and such modifications and variations should be within the scope of protection of this application.

Claims

1. A VR-based gas safety emergency training system, characterized in that, Specifically, it includes: 3D scene construction module: used to build virtual 3D scenes of gas leaks and fire escape, supporting intelligent switching between multiple scenes and real-time parameter adjustment; Interactive data acquisition module: used to collect multimodal data of trainees' interactive operations in 3D scenes during safety emergency training; Interactive data processing module: Cleans and processes the collected interactive data, and calculates the equipment operation accuracy coefficient, process compliance coefficient, handling timeliness coefficient and path planning coefficient. The calculation of the equipment operation accuracy coefficient is as follows: ； Wherein, weight ω f The value ranges from 0.45 to 0.5, and the weight ω m The value ranges from 0.25 to 0.3, and the weight ω g The value ranges from 0.2 to 0.25, and S is the equipment operation accuracy coefficient, which ranges from 0 to 1, where 1 indicates perfect compliance with the standard. F is the valve operation accuracy coefficient, which is the weighted sum of the rotation angle fraction and the switching timing fraction; where the rotation angle fraction = 1 - |actual number of turns - standard number of turns| / standard number of turns; the switching timing fraction is based on whether the valve closing operation is completed within the specified time, with full marks of 1 point for timely completion and deductions for closing delays. M is the fire extinguisher operation accuracy coefficient, which is the weighted sum of the tilt angle score and the spray coverage score. The tilt angle deviation score is judged by whether the deviation angle exceeds the limit. If it is less than or equal to the set deviation value, the full score is 1 point, and if it is greater than the set deviation value, the score is 0. The spray coverage score is judged by the proportion of fire source covered. Coverage x*100% gets x points, and if the coverage is less than 50%, the score is 0. G represents the tool operation accuracy, which assesses the standardization of pipeline maintenance operations. It is a weighted sum of torque error score and disassembly sequence score. Torque error score = 1 - |actual torque - standard torque| / standard torque; disassembly sequence score: full marks of 1 point for correct steps, and deductions proportionally based on the number of incorrect steps for errors or omissions. The specific calculation of the process compliance coefficient is as follows: ； Wherein, the weights ωb and ωw are between 0.6 and 0.7, respectively, L is the process compliance coefficient, and B is the step accuracy coefficient. The standard process is "valve shut-off - ventilation - detection - alarm - evacuation". The step accuracy coefficient = Σ(step weight × operation accuracy rate). W represents the protective equipment coefficient, which is the wearing integrity coefficient, the tightness compliance rate, and the wearing response time weighted sum. The wearing integrity coefficient is calculated as Σ (equipment weight × correct wearing rate). A respirator leakage rate of less than 5% earns 1 point, 5%-10% earns 0.5 points, and more than 10% earns 0 points. The time from alarm to equipment wearing completion is ≤60 seconds, earning 1 point, and deducting 0.1 points for every 10 seconds exceeding this limit. The specific calculation of the disposal timeliness coefficient is as follows: ； Wherein, weight ω y The value ranges from 0.35 to 0.4, and the weight ω d The value ranges from 0.25 to 0.3, and the weight ω e The value ranges from 0.2 to 0.25, and the weight ω h The value ranges from 0.1 to 0.15, where C is the processing timeliness coefficient and Y is the response delay compliance coefficient, specifically: ； Where Y is the response latency compliance coefficient, t s To extend the actual time, t b The standard delay time is used as the benchmark. For every second the actual delay exceeds the standard value, 0.1 points will be deducted, up to a minimum of 0 points. D represents the process deviation compliance coefficient, specifically: ； Where D is the process deviation compliance coefficient, and T s T represents the actual duration, the total time actually consumed by the operation process; b The standard duration is defined as follows: The min function limits the deviation ratio to an upper limit of 1. When the deviation ratio is greater than 1, min outputs 1 and D=0. D=1 indicates that the actual duration is equal to the standard duration, with zero deviation and full compliance. D close to 0 indicates that the deviation is too large and the compliance is low. E is the emergency correction factor, which equals the correction quality factor multiplied by the standard correction time, and is calculated as follows: Emergency Correction Factor = Correction Quality Factor * Standard Correction Time / Actual Correction Time. H is the interruption recovery capability coefficient. When the actual recovery time is ≤30 seconds, H is 1, which is considered fully compliant. When the actual recovery time is >30 seconds, the interruption recovery capability coefficient = 1 - (actual recovery time - 30) / 60. The path planning coefficients are calculated as follows: ； Wherein, weight ω p The value ranges from 0.45 to 0.5, and the weight ω x The value ranges from 0.3 to 0.35, and the weight ω q The value of is between 0.2 and 0.25, and J is the path planning coefficient, which is between 0 and 1, where 1 represents the optimal value. P is the path deviation compliance coefficient, specifically: ； Where P is the path deviation compliance coefficient, P a (t) represents the student's three-dimensional spatial coordinates at time t, P o (t) represents the three-dimensional coordinates of the optimal escape path pre-generated by the system at time t; T represents the total time for path tracking; Dmax represents the maximum allowable deviation threshold. When the threshold is exceeded, P is forced to 0, and the operation is directly judged as a failure. X is the collaboration accuracy coefficient, specifically: ； Where X is the cooperation accuracy coefficient, R is the number of correctly executed instructions, Z is the total number of instructions, Δt is the average instruction response latency, and t max For the maximum allowable delay time, 1 - Δt / t max As a delay penalty factor, its value ranges between 0 and 1. If the response is timely, it is close to 1, and if the delay exceeds the limit, it approaches 0. Q represents the path security compliance coefficient, specifically: ； Where Q is the path safety compliance coefficient, d(t) is the real-time three-dimensional spatial distance between the user and the nearest obstacle or hazard at time t; min(d(t)) represents the minimum distance between the user and the obstacle or hazard during the entire escape process. a This represents the preset minimum safe distance threshold. The min function constrains the upper limit of Q to 1, when the minimum distance ≥ d. a When Q=1, it represents the optimal state; otherwise, points are deducted proportionally. The assessment data analysis module is used to integrate and calculate the processed interaction coefficients to generate a safety emergency training score. The specific scoring criteria for safety and emergency training are as follows: ； Among them, the weight β ranges from 0.25 to 0.3, the weight γ ranges from 0.15 to 0.2, the weight δ ranges from 0.15 to 0.2, A is the comprehensive score of emergency training, S is the equipment operation accuracy coefficient, L is the process compliance coefficient, C is the handling timeliness coefficient, and J is the path planning coefficient. Assessment module: Used to grade and evaluate safety emergency training scores, and to take training intervention measures for trainees whose scores are insufficient.

2. The VR-based gas safety emergency training system according to claim 1, characterized in that: The comprehensive evaluation of safety and emergency training is conducted in a tiered manner as follows: When a1 < A ≤ 100, it is considered excellent. When a2 < A ≤ a1, it is a good grade; When a3 < A ≤ a2, it is of medium grade; When a4 < A ≤ a3, it is a passing grade; When 0 < A ≤ a4, it is a failing grade.

Images