Mine intelligent self-rescuer real-time evaluation system based on multiple modes
By using multimodal data acquisition and an intelligent self-rescue device evaluation system, combined with geographical location and mining environment, dynamic evaluation and predictive scoring of self-rescue devices are achieved, overcoming the limitations of traditional evaluation methods and improving the precision of self-rescue device management in mining environments.
Patent Information
- Application Number
- CN202511569561.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-30
- Publication Date
- 2026-02-03
AI Technical Summary
Existing methods for evaluating mine self-rescue devices lack dynamic assessments that take into account geographical location and mining environment, resulting in evaluation results that are out of touch with reality and fail to reflect the true performance of self-rescue devices in complex environments. Furthermore, the lack of detailed analysis of regional environmental characteristics makes it impossible to quantify the functionality of self-rescue devices and their performance for miners in different regions.
A multimodal data acquisition module is used, combined with physical-chemical-visual three-dimensional data. A multivariate correction model is established through a sensor data calibration module. Real-time evaluation is carried out using an intelligent self-rescue device evaluation module. The environment is further subdivided through a regional joint analysis module to construct a regional functional performance index and a miner wearing performance index.
It enables dynamic and scenario-based evaluation of self-rescue devices, accurately reflects the working status of self-rescue devices in different scenarios, improves the timeliness and scenario adaptability of evaluation, enhances the pertinence of maintenance and optimization, and significantly improves the level of refinement in self-rescue device management.
Smart Images
Figure CN121456528A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of mine safety technology, and more specifically, to a real-time evaluation system for intelligent self-rescue devices in mines based on multimodal operation. Background Technology
[0002] In the field of mine safety, self-rescue devices serve as a "lifeline" for miners when they encounter sudden disasters such as gas explosions, water inrushes, and fires. Their performance evaluation is a key link in ensuring the emergency safety of miners.
[0003] While existing technologies have provided relatively comprehensive assessments of self-rescue devices, they still have shortcomings: Firstly, traditional evaluation methods are mostly static, relying solely on the self-rescue device's own parameters (such as battery and oxygen levels) without considering the actual geographical location and mining environment (such as temperature, gas concentration, and regional characteristics). This leads to a disconnect between the evaluation results and reality, making it difficult to reflect the real-time performance of the self-rescue device in complex environments. Secondly, significant differences exist in mining regional environments (such as large differences in temperature, humidity, and dust concentration in different areas). Analyzing the entire system can be confused by various environmental influencing factors, making it difficult to distinguish and improve the performance of the self-rescue device. Existing technologies lack detailed analysis of regional environmental characteristics, making it impossible to quantify the overall functionality of the self-rescue device and the performance of miners wearing it in different areas. This results in maintenance and optimization measures lacking specificity and failing to adapt to the impact of regional environmental differences on the self-rescue device.
[0004] Therefore, this invention provides a real-time evaluation system for intelligent self-rescue devices in mines based on multimodality, which overcomes the shortcomings of existing technologies. Summary of the Invention
[0005] In view of the shortcomings of existing technologies, the purpose of this invention is to provide a real-time evaluation system for intelligent self-rescue devices in mines based on multimodality.
[0006] To achieve the above objectives, the present invention provides the following technical solution: A real-time evaluation system for intelligent self-rescue devices in mines based on multimodal data acquisition includes a multimodal data acquisition module, a sensor data calibration module, an intelligent self-rescue device evaluation module, and a regional joint analysis module. Multimodal data acquisition module: This module provides data support for comprehensive real-time evaluation of mine self-rescue devices through real-time linkage of physical, chemical, and visual three-dimensional data. This module includes a physical data acquisition unit, a chemical data acquisition unit, and a visual data acquisition unit. Sensor data calibration module: Analyzes historical data to identify key influencing factors: temperature, humidity, and dust concentration; establishes a multivariate correction model and uses the calculated correction coefficients to correct the original measurement data of the corresponding sensors in real time, thereby improving data accuracy; Intelligent self-rescue device evaluation module: Real-time evaluation of intelligent self-rescue devices in mines includes a self-rescue device physical status evaluation unit and a self-rescue device working performance evaluation unit; The self-rescue device physical condition assessment unit uses data from accelerometers, pressure sensors, and gyroscopes to determine if a collision has occurred, internal pressure anomalies, or abnormal posture. It also combines this data with images captured by a camera to check for shell damage and loose components. A physical condition assessment model for the self-rescue device is established, and the weights of each assessment indicator are determined using the analytic hierarchy process (AHP). A comprehensive assessment score is calculated based on the actual data and weights of each indicator. (out of 100) Self-rescue device performance evaluation unit: Taking the evaluation of the performance of a compressed oxygen self-rescue device as an example; 1. Location- and Environment-Based Self-Rescue Device Evaluation: Based on the self-rescue device's performance indicators (oxygen output concentration stability) Oxygen supply pressure fluctuation range Oxygen flow rate compliance rate (etc.) calculate the current location- and environment-based comprehensive performance score. Combined with the self-rescue device's built-in positioning module (BeiDou positioning) to obtain the specific location of the mine construction. At the same time, it correlates with the environmental parameters of that location, including temperature. ,humidity air pressure gas concentration Dust content etc.; Construct a location-environment-performance correlation model, and calculate the predicted score of the self-rescue device's comprehensive performance based on location and environment by using predicted environmental parameters and the coordinates of the self-rescue device. ; 2. User-based self-rescue device evaluation: Collect basic user information, including usage habits (such as the dominant hand used to operate the self-rescue device, preferred operating speed, etc.), work experience (such as years of experience working in mining, number of times the self-rescue device has been used, etc.), and body size (such as height, head circumference, face shape, etc.). Evaluations are conducted at different stages of the user's self-rescue device use, with a maximum score of 100 points.
[0007] Regional Joint Analysis Module: The regional joint analysis module includes a regional division unit, a functional performance analysis unit, and a miner wearing performance analysis unit; 1. Regional Division Units: Using various parameters of the mining environment as feature vectors, the K-means clustering algorithm is applied to divide the mine into regions. Based on feature similarity, the mine is divided into... ,in Indicates the number of sub-regions; 2. Functional Capability Analysis Unit: Using the evaluation results of intelligent self-rescue devices as the core indicator, and combining the regional characteristics output by the regional division unit, a regional functional capability index is constructed. Quantify the overall functionality of self-rescue devices within the sub-region; according to The calculation results range can be divided into regions of high performance, good performance, and poor performance. For regions with poor performance, performance can be improved by optimizing regional attributes or increasing the evaluation score. 3. Miner Wearing Performance Analysis Unit: Combining user-based comprehensive performance evaluation of self-rescue devices. The average score of the physical status of all self-rescue devices in the sub-region Construct a regional miner wearing performance index The overall performance of miners wearing the device within the quantified sub-region; according to The calculation results can be categorized into high-performance, good-performance, and poor-performance regions. For regions with poor performance, timely measures can be taken to improve performance, such as replacing old equipment. or The evaluation score; Furthermore, the multivariate correction model established by the sensor data calibration module is as follows: ; in, The relative error coefficient, This is the difference between the actual temperature and the calibration temperature. Relative humidity (%) Dust concentration ( ); The rated operating relative humidity and dust concentration for the sensor; The error sensitivity is obtained by fitting historical data collected by the sensor; These represent the maximum error percentages caused by temperature, humidity, and dust concentration, respectively. The actual settings are obtained by considering the sensor type, sensor model, and historical data collection errors.
[0008] Furthermore, a comprehensive performance score based on location and environment. The specific calculation method is as follows: The self-rescue device's performance indicators are calculated based on real-time collected data, including the stability of oxygen output concentration. Oxygen supply pressure fluctuation range Oxygen flow rate compliance rate According to the formula Calculate a comprehensive performance score based on location and environment; among which, Let be the weighting coefficients for each indicator, and It can be adjusted according to the importance of different environments.
[0009] Furthermore, the performance indicators of the self-rescue device are calculated using the following specific methods: During the current period, multiple times ( Collect self-rescue device operating data, including oxygen output concentration, oxygen supply pressure, oxygen flow rate, and oxygen output concentration stability. The calculation formula is as follows: ; in, For the first The oxygen concentration measured in this test. , for The average oxygen concentration from each collection. This refers to the number of data collection sessions. Oxygen supply pressure fluctuation range The calculation formula is as follows: ; in, This is the maximum pressure value. This is the minimum pressure value. This is the standard pressure value; Oxygen flow rate compliance rate The calculation formula is as follows: ; in, The number of times the traffic target is met. This represents the total number of tests.
[0010] Furthermore, the self-rescue device's predicted score is based on a comprehensive performance evaluation of its location and environment. The specific calculation method is as follows: A multi-layer neural network architecture is used to predict geographic locations. The parameters serve as direct input layer data for the model, while also incorporating environmental parameters. These intermediate variables are used as intermediate variables. They are fused with the input layer data through the embedding layer to uncover the potential correlation between environmental parameters and input layer data, mapping them to the same feature space and achieving organic feature fusion. The hidden layer of the model adopts a structure combining a long short-term memory network and a convolutional neural network. The former is used to process environmental data with time-series characteristics, while the latter is used to extract features in the spatial dimension. Finally, a fully connected layer outputs a comprehensive performance score of the self-rescuer based on location and environment, ranging from 0 to 100 points. By deriving environmental prediction parameters from historical data, the device can output a predicted score based on a comprehensive evaluation of its operational performance in terms of location and environment, using real-time acquired location and environmental prediction parameters. .
[0011] Furthermore, evaluations were conducted at different stages of the user's use of the self-rescue device, as detailed below: Start-up Phase: This phase includes actions such as opening the outer casing, clipping the nose with the nose clip, biting the mouth onto the breathing mask, and opening the oxygen cylinder valve. The user's operational process and completion status are monitored through a visual data acquisition unit. The startup phase score (normalized to a percentage) is calculated using the following formula: ; in, These are the weighting coefficients, and ; Standard operating time; The actual operation time of the user; For the first The standardization score of each operation (0-1 point); Number of items to be operated; The user adaptability adjustment coefficient is calculated based on user habits and body size data: usage habits are quantified by the user's operation speed in each stage of the self-rescue device operation and the ranking score in the collected miner dataset; body size is evaluated based on the fit between body parameters such as height and head circumference and the self-rescue device's size, with the coefficient value ranged from 0.8 to 1.2. Oxygen Supply Phase: This phase primarily focuses on the decompression process and oxygen delivery. The formula for calculating the oxygen supply phase score (normalized to a percentage) is as follows: ; in, These are the weighting coefficients, and ; This refers to the pressure fluctuation value during the decompression process; This represents the actual oxygen concentration. Standard oxygen concentration; Adjust the coefficient based on user experience, with a value ranging from 0.9 to 1.1; Based on the user's comprehensive score of the self-rescue device's working performance .
[0012] Furthermore, construct a regional functional capacity index. The details are as follows: ; in, As the weight of the evaluation results, and This reflects the contribution of each indicator to the function; It is a non-linear adjustment index.
[0013] Furthermore, construct a regional miner wearing performance index. The details are as follows: ; in, The index is adjusted to reflect the impact of device status on wearability.
[0014] Compared with the prior art, the present invention has the following beneficial effects: 1. Breaking through the traditional static evaluation model, achieving dynamic and scenario-based evaluation: By using geographical location as direct input and introducing environmental parameters as intermediate variables, a multi-layer neural network architecture (integrating long short-term memory networks and convolutional neural networks) is used to build a predictive model. This model can not only calculate the current working performance score based on location and environment, but also output a predicted score. This breaks the limitation of traditional static evaluation focusing only on the equipment's own parameters, deeply correlates the evaluation results with the specific location and real-time environment of the mine, accurately reflects the real working status of the self-rescue device in different scenarios, and improves the timeliness and scenario adaptability of the evaluation. 2. Precise assessment based on regional environmental segmentation enhances the targeted nature of maintenance and optimization: Through a regional joint analysis module, the mine is divided into sub-regions with similar environmental characteristics using the K-means clustering algorithm. Then, regional functional performance indices and miner wearing performance indices are constructed separately. Addressing the issue of significant differences in mine regional environments, this upgrade from single-equipment evaluation to overall regional assessment helps maintenance personnel accurately identify environmental influencing factors in different sub-regions, thereby enabling targeted optimization of the regional environment or adjustment of equipment maintenance strategies, significantly improving the precision of self-rescue device management. Attached Figure Description
[0015] Figure 1 The structural block diagram of a multimodal-based real-time evaluation system for intelligent self-rescue devices in mines; Figure 2 This is a structural diagram of the multimodal data acquisition module of the present invention; Figure 3 This is a structural diagram of the intelligent self-rescue device evaluation module of the present invention; Figure 4 This is a structural diagram of the regional joint analysis module of the present invention. Detailed Implementation
[0016] Example, refer to Figure 1 The real-time evaluation system for intelligent self-rescue devices in mines based on multimodal operation in this embodiment includes a multimodal data acquisition module, a sensor data calibration module, an intelligent self-rescue device evaluation module, and a regional joint analysis module. Multimodal data acquisition module: such as Figure 2As shown, this module provides data support for the comprehensive real-time evaluation of mine self-rescue devices through real-time linkage of physical, chemical, and visual three-dimensional data. This module includes a physical data acquisition unit, a chemical data acquisition unit, and a visual data acquisition unit. Physical data acquisition unit: Miniature sensors such as accelerometers, pressure sensors, and gyroscopes are embedded inside the shell and key components of the intelligent self-rescue device. The accelerometer monitors changes in acceleration during use; a change exceeding 5g within 0.1 seconds is considered a severe impact, indicating potential damage to the internal structure. The pressure sensor monitors the oxygen supply system pressure; the normal pressure for the compressed oxygen self-rescue device is 15-20MPa, and a deviation triggers an alert. The gyroscope senses the attitude, providing real-time feedback on abnormal tilting or flipping. Chemical data acquisition unit: High-precision gas sensors are deployed at the inlet and outlet of the self-rescue device to detect harmful gases such as methane in underground mines. A catalytic combustion sensor is used, leveraging the linear relationship between the electrical signal and methane concentration within the 0-4% concentration range to achieve accurate detection. Visual data acquisition unit: Small cameras are installed on the front and back of the self-rescue device, using image recognition technology to collect and analyze images of the device's appearance and user operation. On one hand, it detects physical damage such as shell breakage and loose parts; on the other hand, it monitors the user's wearing and operational compliance, such as the mask fit and the oxygen supply switch status. For different models, the system pre-builds dedicated image recognition models, analyzes user operation behavior patterns, compares them with standard procedures to determine operational correctness, and identifies design flaws.
[0017] Sensor data calibration module: Analyzes historical data (sensor error records of the same model self-rescuer in different downhole environments over the past three months) to determine the core influencing factor: temperature (the gas sensor error increases by ±T℃). ),humidity( Pressure sensor accuracy decreases at relative humidity. ), dust concentration ( At that time, the transmittance of the optical sensor decreases. ); Establish a multivariate correction model: ;in, The relative error coefficient, This is the difference between the actual temperature and the calibration temperature. Relative humidity (%) Dust concentration ( ); The rated operating relative humidity and dust concentration for the sensor; The error sensitivity is obtained based on the analysis results of historical data collected by the sensor. In this embodiment... ; These represent the maximum error percentages caused by temperature, humidity, and dust concentration, respectively. In actual implementation, these values are determined by considering sensor type, sensor model, and historical data collection errors. (In this embodiment...) =30%; among which, error sensitivity The fitting method is as follows: Focusing on self-rescue devices of the same model over the past three months, environmental parameters (temperature, humidity, dust concentration) and sensor error data (deviation compared with high-precision standard values) were extracted from downhole monitoring systems or sensor logs. For gas sensors, screening Based on the data, focusing on the influence of the single variable of temperature, the error sensitivity of the gas sensor was fitted using the least squares method. For pressure sensors, screening Based on the data, focusing on the influence of the single variable of humidity, the error sensitivity of the pressure sensor was fitted using the least squares method. For optical sensors, screening Based on the data, focusing on the influence of the single variable of dust concentration, the error sensitivity of the optical sensor was fitted using the least squares method. ; The environmental parameters collected in real time are input into the model for calculation. ,use The raw measurement data from the corresponding sensors are corrected in real time to improve data accuracy.
[0018] Intelligent self-rescue device evaluation module: like Figure 3 As shown, real-time evaluation of intelligent self-rescue devices in mines is carried out, including a self-rescue device physical status evaluation unit and a self-rescue device working performance evaluation unit. The self-rescue device physical state assessment unit integrates data from the physical data acquisition unit and the visual data acquisition unit to evaluate the physical state of the intelligent self-rescue device. Data from accelerometers, pressure sensors, and gyroscopes is used to determine if a collision has occurred, internal pressure anomalies, or abnormal posture. Simultaneously, images captured by a camera are used to check for shell damage and loose components. A physical state assessment model for the self-rescue device is established, and the analytic hierarchy process (AHP) is used to determine the weights of each assessment indicator. For example, the weight for collision impact is set to 0.4, pressure anomalies to 0.3, shell damage to 0.2, and component looseness to 0.1. A comprehensive assessment score is calculated based on the actual data and weights of each indicator. (Score based on a percentage system) If the score is below a set threshold (e.g., 60 points), the self-rescue device is deemed to be in an abnormal physical condition. For different models of self-rescue devices, the weights and thresholds of the evaluation indicators are adjusted according to their structural characteristics and vulnerable parts. Self-rescue device performance evaluation unit: Taking the evaluation of the performance of a compressed oxygen self-rescue device as an example; 1. Location- and Environment-Based Self-Rescue Device Evaluation: This evaluation combines the self-rescue device's built-in positioning module (BeiDou positioning) with the specific location of the mine construction site. At the same time, it correlates with the environmental parameters of that location, including temperature. ,humidity air pressure gas concentration Dust content A location-environment-performance correlation model is constructed, using location and depth as input variables, environmental parameters as intermediate variables, and the self-rescue device's performance indicators (oxygen output concentration stability) as the intermediate variables. Oxygen supply pressure fluctuation range Oxygen flow rate compliance rate (etc.) are used as output variables; the comprehensive performance score based on location and environment (based on the scores of all self-rescue devices in the mining area, normalized to a percentage) is as follows: ; in, Let be the weighting coefficients for each indicator, and It can be adjusted according to the importance of different environments; in this embodiment ; in Based on the current period of time, multiple ( The results of data collection are used for calculation: Oxygen output concentration stability The calculation formula is as follows: ; in, For the first The oxygen concentration measured in this test. , for The average oxygen concentration from each collection. This refers to the number of data collection sessions. Oxygen supply pressure fluctuation range The calculation formula is as follows: ; in, This is the maximum pressure value. This is the minimum pressure value. This is the standard pressure value; Oxygen flow rate compliance rate The calculation formula is as follows: ; in, The number of times the traffic target is met. This represents the total number of data collections within the specified time period. A multi-layer neural network architecture is used to predict geographic locations. The parameters serve as direct input layer data for the model, while also incorporating environmental parameters. These intermediate variables are used as intermediate variables. They are fused with the input layer data through the embedding layer to explore the potential correlation between environmental parameters and input layer data, mapping them to the same feature space and achieving organic feature fusion. The hidden layer of the model adopts a structure combining long short-term memory network and convolutional neural network. The former is used to process environmental data with time series characteristics, and the latter is used to extract features in the spatial dimension. Finally, the self-rescue device's performance score of 0-100 is output through the fully connected layer. By deriving environmental prediction parameters from historical data, the device can output a predicted score based on a comprehensive evaluation of its operational performance in terms of location and environment, using real-time acquired location, depth, and environmental prediction parameters. If the score is lower than the set threshold, it indicates that the employee's work performance is not up to par, triggering an early warning mechanism. 2. User-based self-rescue device evaluation: Collect basic user information, including usage habits (such as dominant hand for operating the self-rescue device, preferred operating speed, etc.), work experience (such as years of experience in mining, number of times the self-rescue device has been used, etc.), and body size (such as height, head circumference, face shape, etc.). Evaluations are conducted at different stages of the user's self-rescue device use, with a maximum score of 100 points, as detailed below: 2.1 Start-up Phase: This phase includes operations such as opening the outer casing, clipping the nose with the nose clip, biting the mouth onto the breathing mask, and opening the oxygen cylinder valve. The user's operational process and completion status are monitored through a visual data acquisition unit. The startup phase score (normalized to a percentage) is calculated using the following formula: ; in, These are the weighting coefficients, and In this embodiment ; Standard operating time; The actual operation time of the user; For the first The standardization score of each operation (0-1 point); Number of items to be operated; The user adaptability adjustment coefficient is calculated based on user habits and body size data: usage habits are quantified by the user's operation speed in each stage of the self-rescue device operation and the ranking score in the collected miner dataset; body size is evaluated based on the fit between body parameters such as height and head circumference and the self-rescue device's size, with the coefficient value ranged from 0.8 to 1.2. like A score of 1 indicates that the user's startup operation was excellent; if If the score is 1, it indicates that the startup operation is good; if A score of 0 indicates that the startup operation is unqualified; 2.2 Oxygen Supply Stage: This stage mainly focuses on the decompression process and oxygen delivery. The formula for calculating the oxygen supply stage score (normalized to a percentage) is as follows: ; in, These are the weighting coefficients, and In this embodiment ; This refers to the pressure fluctuation value during the decompression process; This represents the actual oxygen concentration. Standard oxygen concentration; The user experience adjustment factor is determined based on work experience and ranges from 0.9 to 1.1. like A score of 10 indicates excellent performance during the oxygen supply phase; if... A score of 1 indicates good performance; if A score of 0 indicates unsatisfactory performance; Furthermore, based on the comprehensive performance rating of the user's self-rescue device... .
[0019] Regional Joint Analysis Module: such as Figure 4 As shown, the regional joint analysis module includes a regional division unit, a functional performance analysis unit, and a miner wearing performance analysis unit. 1. Regional Division Units: Based on mine design drawings and real-time environmental monitoring data, a multi-dimensional feature space is constructed, using various parameters of the mine environment, such as temperature, humidity, air pressure, gas concentration, and dust content, as feature vectors. The K-means clustering algorithm is used to divide the mine into regions. By calculating the similarity between different feature vectors, interference from cross-regional environmental differences is eliminated. The mine is divided into regions based on feature similarity. ,in Indicates the number of sub-regions; 2. Functional Capability Analysis Unit: Using the evaluation results of intelligent self-rescue devices as the core indicator, and combining the regional characteristics output by the regional division unit, a regional functional capability index is constructed. Quantify the overall functionality of self-rescue devices in this area: ; in, As the weight of the evaluation results, and This reflects the contribution of each indicator to the function. In this embodiment... ; The nonlinear adjustment index is used in this embodiment. ; according to The calculation results range can be divided into regions of high performance, good performance, and poor performance. For regions with poor performance, performance can be improved by optimizing regional attributes or increasing the evaluation score. 3. Miner Wearing Performance Analysis Unit: Combining user-based comprehensive performance evaluation of self-rescue devices. The average score of the physical status of all self-rescue devices in the sub-region Construct a regional miner wearing performance index Overall performance of miners wearing the device within the quantified sub-region: ; in, To adjust the index and reflect the impact of device status on wearability; according to The calculation results can be categorized into high-performance, good-performance, and poor-performance regions. For regions with poor performance, timely measures can be taken to improve performance, such as replacing old equipment. or The evaluation score; Through the detailed description of the above embodiments, the multimodal real-time evaluation system for intelligent self-rescue devices in mines of the present invention innovatively combines geographical location and environmental parameters to analyze the working performance of self-rescue devices, thereby achieving dynamic evaluation and predictive evaluation of self-rescue devices. At the same time, by using a clustering algorithm to divide the mine into sub-regions with similar environmental characteristics, and then constructing a regional functional performance index and a miner wearing performance index respectively, the working performance of self-rescue devices in sub-regions is quantified. This realizes an upgrade from single-device evaluation to overall regional assessment, helping maintenance personnel to accurately identify environmental influencing factors in different sub-regions, thereby optimizing the regional environment or adjusting equipment maintenance strategies in a targeted manner, significantly improving the level of precision in self-rescue device management.
[0020] The above formulas are all dimensionless calculations, and the preset parameters in the formulas should be set by those skilled in the art according to the actual situation.
[0021] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.
[0022] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.
[0023] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0024] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.
[0025] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.
[0026] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0027] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
Claims
1. A real-time evaluation system for intelligent self-rescue devices in mines based on multimodal operation, characterized in that: It includes a multimodal data acquisition module, a sensor data calibration module, an intelligent self-rescue device evaluation module, and a regional joint analysis module; Multimodal data acquisition module: This module provides multimodal data support for the mine intelligent self-rescue device evaluation module through real-time linkage of physical, chemical and visual three-dimensional data; Sensor data calibration module: Analyzes historical data, establishes a multivariate correction model, and corrects the raw data collected by the sensor; Intelligent self-rescue device evaluation module: Real-time evaluation of intelligent self-rescue devices in mines, including a self-rescue device physical status evaluation unit and a self-rescue device working performance evaluation unit; The self-rescue device physical condition assessment unit establishes a self-rescue device physical condition evaluation model and outputs a comprehensive evaluation score. The self-rescue device performance evaluation unit is divided into location- and environment-based self-rescue device evaluation and user-based self-rescue device evaluation. The location- and environment-based self-rescue device evaluation includes calculating the current comprehensive performance score based on the location and environment and the predicted score based on the comprehensive performance score based on the location and environment. The user-based self-rescue device evaluation evaluates different stages in the user's use of the self-rescue device. The regional joint analysis module includes a regional division unit, a functional performance analysis unit, and a miner wearing performance analysis unit. The regional division unit is responsible for dividing the mine into sub-regions with similar characteristics. The functional performance analysis unit constructs a regional functional performance index to quantify the overall functional performance of self-rescue devices within the sub-regions. The miner wearing performance analysis unit constructs a regional miner wearing performance index to quantify the overall performance of miners wearing self-rescue devices within the sub-regions.
2. The real-time evaluation system for intelligent self-rescue devices in mines based on multimodal operation as described in claim 1, characterized in that, The sensor data calibration module establishes a multivariate correction model as follows: Use formula Calculate the relative error coefficient ,in, This is the difference between the actual temperature and the calibration temperature. To generate The relative error corresponds to the temperature change. Relative humidity, Dust concentration; The rated operating relative humidity and dust concentration for the sensor; For error sensitivity; These represent the maximum errors caused by temperature, humidity, and dust concentration, respectively.
3. The real-time evaluation system for intelligent self-rescue devices in mines based on multimodal operation according to claim 1, characterized in that, The current performance evaluation based on location and environment is calculated using the following method: The performance indicators of the self-rescue device are calculated, including the stability of oxygen output concentration, the range of oxygen supply pressure fluctuation, and the oxygen flow rate compliance rate. The overall performance score based on location and environment is calculated by weighted summation of the performance indicators.
4. The real-time evaluation system for intelligent self-rescue devices in mines based on multimodal operation according to claim 3, characterized in that, The specific method for calculating the performance indicators of self-rescue devices is as follows: Parameters of the self-rescue device's operation were collected multiple times, including oxygen output concentration, oxygen supply pressure, and oxygen flow rate. The stability of oxygen output concentration is obtained by calculating the oxygen output concentration. The range of oxygen supply pressure fluctuation is calculated from the oxygen supply pressure. The oxygen flow rate compliance rate is calculated using the oxygen flow rate.
5. The real-time evaluation system for intelligent self-rescue devices in mines based on multimodal operation according to claim 1, characterized in that, The predicted score based on a comprehensive performance evaluation considering location and environment is calculated as follows: A multi-layer neural network architecture is used to predict the model. Geographic location is used as the direct input layer data, and environmental parameters are introduced as intermediate variables. The intermediate variables are fused with the input layer data through the embedding layer. The hidden layer of the model adopts a structure that combines long short-term memory network and convolutional neural network. Finally, the self-rescuer's performance score based on location and environment is output through a fully connected layer. By deriving environmental prediction parameters from historical data, inputting them into a trained prediction model, and outputting a predicted score for the self-rescue device's overall performance based on location and environment.
6. The real-time evaluation system for intelligent self-rescue devices in mines based on multimodal operation according to claim 1, characterized in that, The evaluation was conducted at different stages of the user's use of the self-rescue device, as detailed below: Start-up phase: This phase includes opening the outer shell, clipping the nose with the nose clip, biting the mouth onto the breathing mask, and opening the oxygen cylinder valve. The user's operation time is compared with the standard operation time, and the degree of operation standardization is scored. The start-up phase score is calculated using the start-up phase scoring formula. Oxygen supply stage: This stage is divided into decompression process and oxygen delivery process. The pressure fluctuation value during decompression process is analyzed, the actual oxygen supply concentration is compared with the standard oxygen concentration, and the oxygen supply stage score is calculated using the oxygen supply stage scoring formula. Finally, the scores for the startup phase and the oxygen supply phase are weighted and summed to calculate a comprehensive score for the self-rescue device's performance based on the user.
7. The real-time evaluation system for intelligent self-rescue devices in mines based on multimodal operation according to claim 1, characterized in that, The regional functional capability index is constructed as follows: Using formula Calculate the regional functional capacity index; in, As the weight of the evaluation results, and ; It is a non-linear adjustment index. The score is a comprehensive evaluation of the physical condition of the self-rescue device. A comprehensive performance score based on current location and environment; This is a predicted score based on a comprehensive performance evaluation of work based on location and environment.
8. The real-time evaluation system for intelligent self-rescue devices in mines based on multimodal operation according to claim 1, characterized in that, The regional miner wearing performance index is constructed as follows: Using formula Calculate the performance index of miners wearing protective gear in the region; among which, The self-rescue device performance is rated based on the user's overall performance. The average score of the physical states of all self-rescue devices within the divided sub-regions. To adjust the index.