Coal mine safety monitoring data self-diagnosis method

Abstract

The invention discloses a coal mine safety monitoring data self-diagnosis method, and relates to the technical field of coal mine safety data monitoring. The coal mine safety monitoring data self-diagnosis method comprises the following steps: S1, collecting coal mine environment data in real time, and transmitting the coal mine environment data to a central control platform for preprocessing; s2, performing burst data detection on the preprocessed coal mine environment data, and performing burst data processing on the coal mine environment data according to a burst data detection result; s3, performing fluctuation anomaly detection on the coal mine environment data subjected to burst data processing, and performing fluctuation correction according to a fluctuation anomaly detection result; s4, in combination with the burst data detection result and the fluctuation anomaly detection result, optimizing alarm triggering conditions, and reducing false alarms and missing alarms; and S5, establishing a fault mode library according to historical data, and automatically identifying potential fault types. The problem of false alarm and missing alarm of the system caused by the fact that the environment in the mine is complex and normal fluctuation and real fault signals are very difficult to distinguish is solved.

Description

Technical Field

[0001] The invention relates to the technical field of coal mine safety data monitoring, and in particular to a coal mine safety monitoring data self-diagnosis method. Background Art

[0002] With the development and expansion of the coal mining industry, coal mine safety monitoring systems are playing an increasingly important role in ensuring mine safety. The environment inside the mine is complex and changeable, and problems such as gas leakage, abnormal temperature, and equipment failure occur from time to time. These failures are often difficult to detect in the first place. Traditional coal mine safety monitoring systems mainly rely on manual detection and regular inspections, and have problems such as slow response, missed detection, and false alarms, and cannot achieve real-time and accurate monitoring and early warning.

[0003] With the advancement of sensor technology, data collection and processing technology, modern coal mine safety monitoring systems have gradually begun to rely on big data analysis and intelligent algorithms to improve monitoring efficiency and safety. Especially when conducting real-time analysis of environmental data in the mine, how to accurately identify possible anomalies, faults or safety hazards has become the key to improving coal mine safety and reducing accidents.

[0004] In the existing coal mine safety monitoring system, although a variety of data analysis methods have been applied, there are still some shortcomings. For example, the changes in coal mine monitoring data are diverse and sudden, and traditional monitoring algorithms are difficult to accurately handle complex nonlinear characteristics, resulting in poor real-time detection capabilities for sudden faults; at the same time, environmental changes such as climate and fluctuations in sensor performance are also prone to false alarms and missed alarms, affecting the reliability and response efficiency of the system.

[0005] Therefore, in response to the above problems, a self-diagnosis method for coal mine safety monitoring data is urgently needed. Summary of the invention

[0006] Technical issues solved In view of the deficiencies in the prior art, the present invention provides a method for self-diagnosis of coal mine safety monitoring data, which solves the problem of system false alarms and missed alarms caused by the difficulty in distinguishing normal fluctuations from real fault signals due to the complex environment in the mine.

[0007] Technical Solution To achieve the above objectives, the present invention is implemented through the following technical solutions: a coal mine safety monitoring data self-diagnosis method, including: S1, real-time collection of coal mine environmental data, and transmission to the central control platform for preprocessing; S2, burst data detection on the preprocessed coal mine environmental data, and burst data processing on the coal mine environmental data according to the burst data detection results; S3, fluctuation anomaly detection on the coal mine environmental data after burst data processing, and fluctuation correction according to the fluctuation anomaly detection results; S4, combining the burst data detection results and the fluctuation anomaly detection results, optimizing the alarm triggering conditions, and reducing false alarms and missed alarms; S5, establishing a fault mode library based on historical data, and automatically identifying potential fault types.

[0008] Furthermore, the real-time collection of coal mine environmental data and transmission to the central control platform for preprocessing specifically includes: real-time collection of coal mine environmental data, the coal mine environmental data including: gas concentration, temperature, pressure and vibration; transmitting the collected coal mine environmental data to the central control platform for preprocessing: denoising the coal mine environmental data, using filters to eliminate interference from the external environment, filling in missing data through interpolation methods, standardizing and normalizing the coal mine environmental data from different sensors; recording the mean of the coal mine environmental data and the standard deviation of the coal mine environmental data, and obtaining the data mean of the sliding window and the data standard deviation of the sliding window.

[0009] Furthermore, the specific process of performing burst data detection on the pre-processed coal mine environmental data and performing burst data processing on the coal mine environmental data according to the burst data detection result is as follows: obtaining the coal mine environmental data, the mean of the coal mine environmental data and the standard deviation of the coal mine environmental data, and obtaining a burst feature evaluation value by comprehensive calculation of the coal mine environmental data, the mean of the coal mine environmental data and the standard deviation of the coal mine environmental data; comparing the burst feature evaluation value with a first threshold in real time, and when the burst feature evaluation value is less than the first threshold, no processing is performed; when the burst feature evaluation value is greater than or equal to the first threshold, increasing the accuracy of data collection, automatically increasing the space for data storage, automatically comparing historical data, predicting the type of fault, and automatically switching related monitoring modes.

[0010] Furthermore, the specific calculation method of the burst characteristic evaluation value is: ; In the formula, represents the burst characteristic evaluation value, represents the coal mine environmental data at time point t, represents the mean value of coal mine environmental data, Represents the standard deviation of coal mine environmental data.

[0011] Furthermore, the specific process of performing fluctuation anomaly detection on the coal mine environmental data after burst data processing and performing fluctuation correction according to the fluctuation anomaly detection result is as follows: obtaining the coal mine environmental data, the data mean of the sliding window and the data standard deviation of the sliding window, and obtaining the fluctuation anomaly detection value by comprehensive calculation of the coal mine environmental data, the data mean of the sliding window and the data standard deviation of the sliding window; comparing the fluctuation anomaly detection value and the second threshold in real time, and when the abnormal fluctuation detection value is less than the second threshold, no processing is performed; when the abnormal fluctuation detection value is greater than or equal to the second threshold, a potential abnormal fluctuation is prompted, and the sampling frequency is automatically increased according to the intensity of the abnormal fluctuation. During the abnormal fluctuation, the data smoothing algorithm is automatically used to correct the data to avoid the abnormal fluctuation affecting the system judgment, and by analyzing the fluctuation trend, future abnormal fluctuations are predicted in advance, and the monitoring strategy is automatically adjusted.

[0012] Furthermore, the specific calculation method of the abnormal fluctuation detection value is: ; In the formula, Indicates the abnormal fluctuation detection value, represents the coal mine environmental data at time point t, Represents a sliding window The mean of the data in Represents a sliding window The standard deviation of the data in .

[0013] Furthermore, the specific process of combining the burst data detection results and the fluctuation anomaly detection results to optimize the alarm trigger conditions and reduce false alarms and missed alarms is as follows: obtaining the burst feature evaluation value and the abnormal fluctuation detection value, obtaining the false alarm and missed alarm optimization value through comprehensive calculation of the burst feature evaluation value and the abnormal fluctuation detection value, and comparing the false alarm and missed alarm optimization value with the third threshold in real time. When the false alarm and missed alarm optimization value is less than the third threshold, no processing is performed; when the false alarm and missed alarm optimization value is greater than or equal to the third threshold, the alarm threshold is automatically adjusted to avoid false alarms and missed alarms.

[0014] Furthermore, the specific process of automatically adjusting the dynamic alarm threshold to avoid false alarms and missed alarms is as follows: automatically adjusting the dynamic alarm threshold according to the false alarm and missed alarm optimization value and the current abnormality detection result; when the coal mine environmental data fluctuates greatly, reducing the dynamic alarm threshold; at the same time, the dynamic alarm threshold will be automatically updated according to the real-time false alarm and missed alarm optimization value.

[0015] Furthermore, the specific calculation method of the false alarm and missed alarm optimization value is: ; In the formula, Indicates the optimization value of false positives and false negatives. represents the burst characteristic evaluation value, represents the weight coefficient of the burst feature evaluation value, Indicates the abnormal fluctuation detection value, Indicates the weight coefficient of the abnormal fluctuation detection value.

[0016] Furthermore, the specific process of establishing a fault mode library based on historical data and automatically identifying potential fault types is as follows: based on historical fault cases, each fault type is classified with a corresponding characteristic pattern to establish a fault mode library; a neural network model is established using historical data and the fault mode library to learn characteristic patterns of different fault types; real-time data and historical data are compared in real time to identify potential fault types through pattern matching; corresponding response measures are automatically initiated based on the identified fault type; as new faults occur, new fault data is automatically added to the fault mode library to improve the accuracy of the model.

[0017] Beneficial Effects The present invention has the following beneficial effects: (1) The present invention ensures the accuracy and integrity of the data by collecting coal mine environmental data in real time and performing preprocessing, thereby realizing comprehensive real-time monitoring of the mine environment by the system, and effectively solving the problems of data delay and incompleteness in traditional monitoring methods.

[0018] (2) The present invention can adjust the monitoring strategy and increase the data sampling frequency in a timely manner by automatically detecting and processing sudden events and fluctuation anomalies, thereby achieving rapid response to sudden events and automatic correction of fluctuations, effectively solving the problems of slow sudden event detection and inability to adjust fluctuation anomalies in a timely manner in the prior art.

[0019] (3) The present invention optimizes the alarm triggering conditions and reduces false alarms and missed alarms by combining the burst data detection results and the fluctuation anomaly detection results, thereby realizing intelligent adjustment and optimization of the alarm system and effectively solving the problem of frequent false alarms and missed alarms in the traditional alarm system.

[0020] (4) The present invention establishes a fault mode library based on historical data and realizes automatic fault type identification, thereby realizing automatic early warning and response to potential faults, effectively solving the problem in the prior art that potential faults in a mine environment cannot be identified and handled in advance.

[0021] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 The present invention is a flow chart of a method for self-diagnosis of coal mine safety monitoring data. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] See also Figure 1 The embodiment of the present invention provides a technical solution: a method for self-diagnosis of coal mine safety monitoring data, including S1, real-time collection of coal mine environmental data, and transmission to a central control platform for preprocessing; S2, burst data detection on the preprocessed coal mine environmental data, and burst data processing on the coal mine environmental data according to the burst data detection result; S3, fluctuation anomaly detection on the coal mine environmental data after burst data processing, and fluctuation correction according to the fluctuation anomaly detection result; S4, combining the burst data detection result and the fluctuation anomaly detection result, optimizing the alarm triggering condition, and reducing false alarms and missed alarms; S5, establishing a fault mode library according to historical data, and automatically identifying potential fault types.

[0025] Specifically, the coal mine environmental data is collected in real time and transmitted to the central control platform for preprocessing. The specific process is as follows: First, a variety of high-precision sensors (such as gas concentration sensors, temperature and humidity sensors, pressure sensors, and vibration sensors) are installed inside the coal mine to monitor the environmental changes in the mine in real time. Coal mine environmental data includes key parameters such as gas concentration, temperature, pressure, and vibration, which can reflect the safety status and equipment operation status in the mine. After data collection, the sensor will transmit the real-time collected environmental data to the central control platform. Next, the central control platform preprocesses the collected raw data to ensure the accuracy and reliability of the data.

[0026] During the preprocessing process, the coal mine environmental data is first denoised, and filter technology (such as low-pass filter) is used to eliminate interference signals from the external environment or equipment, so as to ensure that the values ​​collected by the sensor are more accurate and avoid the influence of noise on the analysis results. At the same time, for data loss that may occur during the transmission process, the system uses interpolation methods (such as linear interpolation or spline interpolation) to fill in the missing data, maintain the continuity and integrity of the data, and ensure that the data set can provide complete input for subsequent analysis.

[0027] In addition, in order to make different types of sensor data have the same metric standard, the system will standardize and normalize the sensor data. Standardization eliminates the scale differences of different sensor output data so that different types of data can be compared and analyzed under the same benchmark. Normalization adjusts all sensor data to the same numerical range to facilitate subsequent algorithm processing.

[0028] On this basis, the system will record the mean and standard deviation of the coal mine environmental data in real time to reflect the concentration trend and volatility of the data. These statistics help the system evaluate the changing trend of the mine environment in real time. At the same time, in order to capture more detailed changes, the system will use the sliding window method to calculate the mean and standard deviation of the data in the sliding window in real time. The size of the sliding window can be adjusted according to specific needs. The common window size is the environmental data of the past 24 hours. This helps the system monitor short-term fluctuations and provide more accurate anomaly detection and response capabilities. Through these preprocessing steps, the coal mine environmental data can maintain high quality and high accuracy before entering further analysis and model processing, providing a solid data foundation for subsequent fault diagnosis and safety monitoring.

[0029] In this implementation scheme, the accuracy and completeness of the data are ensured by real-time collection and preprocessing of coal mine environmental data, thus providing a reliable basis for subsequent monitoring and analysis. De-noising effectively eliminates external interference and noise to ensure the purity of the data; the interpolation method fills in the missing parts of the data to avoid analysis deviations caused by data loss; standardization and normalization processing enable data from different types of sensors to be compared and analyzed under the same benchmark, improving data consistency. By recording the mean and standard deviation of coal mine environmental data, as well as data statistics within the sliding window, the system can monitor and evaluate changes in the mine environment in real time and detect abnormal fluctuations in a timely manner. This series of data preprocessing measures enables the system to accurately capture subtle changes in the mine environment, improve the efficiency of subsequent fault diagnosis and safety warnings, and ensure the safe operation of the mine.

[0030] Specifically, the coal mine environmental data after preprocessing is subjected to burst data detection, and the coal mine environmental data is subjected to burst data processing according to the burst data detection results. The specific process is as follows: first, the coal mine environmental data is obtained, including key parameters such as gas concentration, temperature, pressure, vibration, etc. These data can reflect the real-time operating status and environmental changes in the mine. Then, the system calculates the mean and standard deviation of the coal mine environmental data in real time as a benchmark for measuring data fluctuations. The mean reflects the central trend of the data, while the standard deviation reveals the fluctuation amplitude of the data, which is an important basis for judging whether there is a sudden change.

[0031] Based on the coal mine environmental data, mean and standard deviation, the system obtains the sudden feature evaluation value through comprehensive calculation, which can help the system identify sudden events or drastic environmental fluctuations. This feature evaluation value is compared with historical data to monitor in real time whether there are abnormal fluctuations or sudden conditions. By comparing the sudden feature evaluation value with the set first threshold, the system can make corresponding judgments. If the sudden feature evaluation value is less than the first threshold, it means that the data fluctuation is within the normal range, and the system will not process it and continue to monitor; however, if the sudden feature evaluation value is greater than or equal to the first threshold, the system will automatically take a series of optimization measures.

[0032] When sudden data is detected, the system will automatically increase the accuracy of data collection and the sampling frequency to better capture subsequent data changes and ensure that abnormal fluctuations are not ignored. At the same time, the system will automatically increase the space for data storage to ensure that all relevant data can be recorded and saved during emergencies to avoid loss due to a surge in data volume. At the same time, the system will automatically compare historical data, use historical failure modes and currently collected data to predict possible failure types, and start predictive models for failure analysis. In order to further improve monitoring efficiency, the system will also automatically switch related monitoring modes based on the prediction results to adapt to emergencies and respond quickly to ensure that the mine environment can be monitored and processed in a timely manner.

[0033] In this implementation scheme, through burst data detection and processing, abnormal fluctuations and emergencies in the coal mine environment can be effectively identified, ensuring that the system can respond quickly when an abnormal situation occurs in the mine. When the burst feature evaluation value exceeds the set threshold, the system will automatically improve the data collection accuracy and expand the storage space to fully capture abnormal data and prevent data loss. At the same time, through historical data comparison and fault type prediction, the system can identify potential faults in advance and automatically adjust the monitoring mode to adapt to emergencies. This series of automated processing ensures real-time monitoring and early warning of the coal mine environment, effectively improves mine safety, and avoids delays and omissions that may be caused by manual intervention.

[0034] Specifically, the burst characteristic evaluation value is calculated as follows: ; In the formula, It represents the sudden characteristic evaluation value, reflecting the abnormal fluctuation degree of the coal mine environmental data at that time point. Represents the coal mine environmental data at time point t, including but not limited to gas concentration, temperature, humidity, pressure, vibration, etc. These data are collected in real time by sensors and transmitted to the central control platform for processing and analysis. It represents the mean value of coal mine environmental data. It is the average value of data within a period of time (such as the past 24 hours). It is usually updated by the sliding window method to ensure that it reflects the normal fluctuation level of the coal mine environment. It represents the standard deviation of coal mine environmental data, reflects the amplitude and volatility of data fluctuations, and is used to measure the degree of dispersion of data.

[0035] In this implementation scheme, through the real-time calculation of the burst feature evaluation value, the system can accurately identify the violent fluctuations and emergencies in the coal mine environment. When the coal mine environmental data mutates, the burst feature evaluation value will quickly reflect this change and be compared with the preset threshold. If the evaluation value exceeds the threshold, the system can automatically identify abnormal fluctuations and respond, such as increasing data collection accuracy, expanding storage space, and starting a fault prediction mechanism. This automated processing can effectively reduce manual intervention and improve the response speed and accuracy of mine monitoring, thereby ensuring real-time monitoring and early warning of mine safety conditions, timely discovering and responding to potential risks, and greatly improving the safety assurance capabilities of coal mines.

[0036] Specifically, the coal mine environmental data after burst data processing is subjected to fluctuation anomaly detection, and fluctuation correction is performed according to the fluctuation anomaly detection results. The specific process is as follows: First, it is necessary to obtain coal mine environmental data, including data collected in real time by sensors such as gas concentration, temperature, humidity, pressure and vibration. These data can reflect various changes in the mine environment. Subsequently, the system processes these data through the sliding window technology and calculates the mean and standard deviation within the sliding window. The data mean of the sliding window reflects the average level of the coal mine environment over the past period of time, while the data standard deviation of the sliding window reveals the fluctuation amplitude and changes of the data during this period of time. Through the calculation of these two indicators, the system can comprehensively evaluate the stability and fluctuation of coal mine environmental data.

[0037] The fluctuation anomaly detection value is obtained by comprehensively calculating the mean and standard deviation of the coal mine environmental data and the sliding window. The fluctuation anomaly detection value can reflect the degree of deviation between the current data and the historical data, and is used to identify whether there is an abnormal fluctuation. The system compares the fluctuation anomaly detection value with the set second threshold in real time. When the detection value is less than the second threshold, it means that the current data fluctuation is within the normal range, and the system does not perform further processing and continues normal monitoring; if the detection value is greater than or equal to the second threshold, it means that there is a potential abnormal fluctuation, and the system will automatically enter the emergency response mode.

[0038] When abnormal fluctuations occur, the system will automatically increase the data sampling frequency according to the intensity of the fluctuations in order to more accurately capture the changes in fluctuations. At the same time, the system will automatically use the data smoothing algorithm to correct the data, eliminate the impact of abnormal fluctuations on data analysis, and avoid misjudgments caused by abnormal fluctuations. Through these automated processes, the system ensures the stability and accuracy of coal mine environmental data. Through continuous analysis of fluctuation trends, the system can predict abnormal fluctuations that may occur in the future in advance and automatically adjust the monitoring strategy to ensure that when abnormal fluctuations occur, the system can respond in time and take corresponding adjustment measures. All operations in this process are automatically completed by the system, which greatly improves the real-time and accuracy of coal mine safety monitoring, reduces the need for manual intervention, and ensures the safe operation of the mine.

[0039] In this implementation scheme, through fluctuation anomaly detection, the system can accurately monitor fluctuation changes in the coal mine environment and automatically identify potential abnormal fluctuations. When the abnormal fluctuation detection value exceeds the preset threshold, the system can promptly identify the fluctuation intensity and automatically adjust the response measures. For example, the data collection accuracy is improved by increasing the data sampling frequency, and the data smoothing algorithm is used to correct the data during the fluctuation period to ensure that the fluctuation does not affect the system's judgment. At the same time, the system can also predict abnormal fluctuations that may occur in the future based on the analyzed fluctuation trends, and automatically adjust the monitoring strategy to better deal with potential safety hazards. This automated processing process greatly improves the safety monitoring capabilities of the mine, ensures timely and accurate abnormal responses and more efficient risk prediction, reduces the need for manual intervention, and improves the overall safety of the mine.

[0040] Specifically, the specific calculation method of the abnormal fluctuation detection value is: ; In the formula, It represents the abnormal fluctuation detection value, which reflects the deviation degree between the current coal mine environmental data and the historical data in the sliding window. Represents the coal mine environmental data at time point t, including key parameters such as gas concentration, temperature, humidity, pressure and vibration. These data are collected in real time by sensors in the mine and transmitted to the central control platform for processing and analysis. Represents a sliding window The mean value of the data in the sliding window is usually calculated by averaging the environmental data over a period of time (such as the past 24 hours or a set time window). By calculating the mean value of the data in the sliding window, the normal fluctuation level of the mine in the time period can be reflected. The size of the sliding window It can be set according to actual needs. The longer the window size, the more the average can smooth short-term fluctuations, which is suitable for long-term trend monitoring. When the window size is shorter, it can more sensitively capture short-term fluctuations. Represents a sliding window The standard deviation of the data in reflects the volatility of the data during that time period.

[0041] In this implementation scheme, the coal mine environmental data is accurately monitored through abnormal fluctuation detection values, and abnormal fluctuations and sudden changes in the data can be identified in real time. When the fluctuation between the current data and the historical data exceeds the normal range, the abnormal fluctuation detection value will exceed the preset threshold, thereby triggering the system's automatic response mechanism. The system can identify abnormal fluctuations in a timely manner, automatically adjust the data sampling frequency to more accurately capture subsequent fluctuation changes, and correct possible noise through data smoothing algorithms to avoid the impact of abnormal fluctuations on subsequent analysis. In addition, the system can also automatically adjust the monitoring strategy according to the predicted fluctuation trend to ensure early warning of potential risks. Through these automated processes, the coal mine safety monitoring system can detect and respond to environmental changes in real time and accurately, reduce manual intervention, and improve the safety and reliability of the system.

[0042] Specifically, combining the results of burst data detection and abnormal fluctuation detection, the alarm triggering conditions are optimized to reduce false alarms and missed alarms. The specific process is as follows: First, the burst feature evaluation value and abnormal fluctuation detection value are obtained. These two indicators reflect the abnormal conditions of sudden fluctuations and normal fluctuations in the coal mine environment respectively. The burst feature evaluation value identifies sudden changes by comparing the fluctuations of current data and historical data, while the abnormal fluctuation detection value is calculated through a sliding window to evaluate the degree of deviation of the current data relative to normal fluctuations. Through the comprehensive calculation of the two, the optimization value of false alarms and missed alarms can be obtained, which reflects the comprehensive impact of the two abnormal indicators on the system alarm.

[0043] The system compares the false alarm and missed alarm optimization value with the set third threshold in real time to determine whether an alarm needs to be triggered. If the false alarm and missed alarm optimization value is less than the third threshold, it means that the current fluctuations and emergencies are within an acceptable range, and the system does not take further action and continues to monitor normally; when the false alarm and missed alarm optimization value is greater than or equal to the third threshold, the system automatically adjusts the alarm threshold to further optimize the alarm condition. At this time, the system adjusts the alarm threshold to avoid false alarms caused by environmental fluctuations or sensor noise, and increases sensitivity to actual faults, effectively reducing the possibility of missed alarms.

[0044] In this implementation scheme, by combining the results of burst data detection and fluctuation anomaly detection, the system can automatically optimize the alarm triggering conditions and effectively reduce false alarms and missed alarms. When the system detects abnormal fluctuations or emergencies, it generates false alarm and missed alarm optimization values ​​by comprehensively calculating the burst feature evaluation value and the abnormal fluctuation detection value, and compares them with the set threshold in real time. If the detection result exceeds the threshold, the system will automatically adjust the alarm standard to ensure that the alarm is more accurate and sensitive. This mechanism greatly improves the accuracy of the coal mine safety monitoring system, can avoid unnecessary false alarms and missed alarms, improves the response speed and reliability of the alarm system, and effectively ensures the safety of the mine environment.

[0045] Specifically, the dynamic alarm threshold is automatically adjusted to avoid false alarms and missed alarms. The specific process is: by analyzing the false alarm and missed alarm optimization value and the current abnormal detection results in real time, the sensitivity and accuracy of the alarm system are ensured. When the coal mine environmental data fluctuates greatly, the system will automatically reduce the dynamic alarm threshold according to the false alarm and missed alarm optimization value, so that the system can respond more sensitively to potential abnormal changes. This dynamic adjustment mechanism can quickly improve the sensitivity of the alarm when the data fluctuates abnormally, identify potential faults or safety hazards in time, and avoid missed alarms due to excessively high thresholds. At the same time, the dynamic alarm threshold will be automatically updated according to the real-time false alarm and missed alarm optimization value to ensure that the system can be flexibly adjusted according to the actual fluctuations of the current environment. For example, during normal fluctuations, the alarm threshold will be higher to reduce the frequency of false alarms; when abnormal fluctuations occur, the system will automatically lower the threshold to increase the response speed to faults or abnormal fluctuations. Through this automated adjustment process, the system can adapt to changes in the mine environment in real time, effectively reduce false alarms and missed alarms, and improve the accuracy and efficiency of mine safety monitoring.

[0046] In this implementation scheme, by automatically adjusting the dynamic alarm threshold, the system can flexibly respond to changes in the coal mine environment and effectively avoid false alarms and missed alarms. When the coal mine environmental data fluctuates greatly, the system will automatically lower the alarm threshold and increase the sensitivity to potential anomalies, so as to capture abnormal situations in a timely manner. At the same time, the system will automatically update the threshold according to the real-time false alarm and missed alarm optimization value to ensure that the alarm sensitivity always matches the current environment. This dynamic adjustment mechanism makes the alarm system more accurate, which can avoid missed alarms caused by too high thresholds and false alarms caused by too low thresholds, significantly improving the reliability and real-time response capabilities of mine safety monitoring.

[0047] Specifically, the specific calculation method of the false positive and false negative optimization value is: ; In the formula, It represents the optimization value of false alarm and missed alarm, which reflects the sensitivity of the system to adjust the alarm under the combined effect of the sudden feature evaluation value and the abnormal fluctuation detection value. It represents the sudden characteristic evaluation value, which is used to measure the intensity of sudden changes in coal mine environmental data. It is usually calculated by comparing the fluctuation degree of current data with historical data. represents the weight coefficient of the burst feature evaluation value, It represents the abnormal fluctuation detection value, which is used to measure the degree of abnormality of the data relative to the normal fluctuation. The calculation method is based on the mean and standard deviation of the sliding window. Represents the weight coefficient of the abnormal fluctuation detection value, and It is the weight coefficient of the sudden characteristic evaluation value and the abnormal fluctuation detection value. The value range of the weight coefficient is usually 0 to 1, where and The value of determines the influence of these two factors on the optimization value of false alarm and missed alarm. In practical applications, these weight coefficients can be dynamically adjusted according to the analysis of historical data and the needs of real-time monitoring. For example, if the sudden event feature evaluation value is more important to the system response, then The value of may be larger; when the abnormal fluctuation detection value has a greater impact on the alarm, The weight coefficient is obtained from the fault mode library to achieve the optimal response to different types of abnormalities.

[0048] In this implementation scheme, the alarm triggering conditions are automatically optimized by comprehensively calculating the sudden feature evaluation value and the abnormal fluctuation detection value (VADV), and adjusting the false alarm and missed alarm optimization values ​​according to their respective weight coefficients 𝛼 and 𝛽. When abnormal fluctuations or sudden changes occur in environmental data, the system can flexibly adjust the alarm threshold to ensure timely response to real faults or safety hazards and avoid false alarms or missed alarms caused by environmental fluctuations. By dynamically adjusting the weight coefficients, the system can optimize the alarm strategy according to different abnormal types and monitoring requirements, improving the intelligence and response accuracy of coal mine safety monitoring. This automated false alarm and missed alarm optimization mechanism significantly improves the accuracy and reliability of the system, ensuring the real-time and effectiveness of mine safety monitoring.

[0049] Specifically, a fault mode library is established based on historical data to automatically identify potential fault types. The specific process is as follows: first, the characteristic patterns of each fault type are extracted from a large number of historical fault cases and classified to establish a detailed fault mode library. Each fault type is closely associated with the corresponding characteristic patterns, which include characteristic values ​​of sensor data, such as fluctuation patterns and abnormal values ​​of gas concentration, temperature, vibration and other data. Through in-depth analysis of these historical data, the typical manifestations of different faults in coal mine environments can be identified, forming multi-dimensional fault modes.

[0050] Next, the system uses these historical data and the fault pattern library to build a neural network model. The neural network can effectively recognize new data by learning the characteristic patterns of different fault types. Through training, the model can extract the potential laws of fault patterns and accurately identify the differences between real-time data and historical data. When comparing real-time monitoring data with historical data in real time, the model automatically identifies potential fault types through pattern matching technology.

[0051] Once a potential fault type is identified, the system will automatically initiate corresponding response measures according to the fault type, such as launching an emergency plan, adjusting equipment parameters, or sounding an alarm, to ensure that the mine can take preventive and emergency measures in a timely manner to prevent the fault from further expanding or causing a safety accident. As new faults occur, the system will automatically add new fault data to the fault pattern library for pattern updates and retraining. This process not only improves the accuracy of fault detection, but also enables the system to gradually adapt to changes in the mine environment, improve the real-time adaptability and accuracy of the model, and thus ensure the continuous optimization of mine safety monitoring.

[0052] In this implementation scheme, by establishing a fault pattern library based on historical data and combining it with the learning of a neural network model, the system can automatically identify potential fault types. By analyzing historical fault cases, the system can extract the characteristic patterns of each fault type and perform pattern matching in real-time monitoring to achieve accurate identification of potential faults in the current mine environment. When a fault is identified, the system will automatically initiate the corresponding response measures to ensure timely response and prevent the fault from expanding. At the same time, the system can continuously learn and update the fault pattern library, add new fault data to it, and improve the accuracy of fault identification and the adaptability of the system. This intelligent fault identification and response mechanism significantly improves the real-time, accuracy and reliability of the mine safety monitoring system, enhances the adaptability to changes in the mine environment, and ensures the safe operation of the mine.

[0053] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0054] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for self-diagnosis of coal mine safety monitoring data, characterized in that: include: S1, collects coal mine environmental data in real time and transmits it to the central control platform for preprocessing; S2, performing burst data detection on the pre-processed coal mine environment data, and performing burst data processing on the coal mine environment data according to the burst data detection result; S3, performing fluctuation anomaly detection on the coal mine environmental data after burst data processing, and performing fluctuation correction according to the fluctuation anomaly detection result; S4, combining the burst data detection results and fluctuation anomaly detection results, optimizes the alarm triggering conditions and reduces false alarms and missed alarms; S5, establishes a failure mode library based on historical data and automatically identifies potential failure types.

2. A method for self-diagnosis of coal mine safety monitoring data according to claim 1, characterized in that: The specific process of real-time collection of coal mine environmental data and transmission to the central control platform for preprocessing is as follows: Real-time collection of coal mine environmental data, including gas concentration, temperature, pressure and vibration; The collected coal mine environmental data is transmitted to the central control platform for pre-processing: De-noising the coal mine environmental data, using filters to eliminate interference from the external environment, filling in missing data through interpolation methods, and standardizing and normalizing the coal mine environmental data from different sensors; The mean value and standard deviation of the coal mine environmental data are recorded, and the mean value and standard deviation of the data in the sliding window are obtained at the same time.

3. A method for self-diagnosis of coal mine safety monitoring data according to claim 1, characterized in that: The specific process of performing burst data detection on the pre-processed coal mine environment data and performing burst data processing on the coal mine environment data according to the burst data detection result is as follows: Obtaining coal mine environmental data, the mean of the coal mine environmental data and the standard deviation of the coal mine environmental data, and obtaining a sudden characteristic assessment value through comprehensive calculation of the coal mine environmental data, the mean of the coal mine environmental data and the standard deviation of the coal mine environmental data; Compare the burst feature evaluation value with the first threshold in real time, and do not process when the burst feature evaluation value is less than the first threshold; When the burst feature evaluation value is greater than or equal to the first threshold, the accuracy of data collection is increased, the space for data storage is automatically increased, historical data is automatically compared, the fault type is predicted, and the relevant monitoring mode is automatically switched.

4. A method for self-diagnosis of coal mine safety monitoring data according to claim 3, characterized in that: The specific calculation method of the burst feature evaluation value is: ； In the formula, represents the burst characteristic evaluation value, represents the coal mine environmental data at time point t, represents the mean value of coal mine environmental data, Represents the standard deviation of coal mine environmental data.

5. A method for self-diagnosis of coal mine safety monitoring data according to claim 1, characterized in that: The specific process of performing fluctuation anomaly detection on the coal mine environmental data after burst data processing and performing fluctuation correction according to the fluctuation anomaly detection result is as follows: Obtain coal mine environmental data, data mean of the sliding window, and data standard deviation of the sliding window, and obtain a fluctuation anomaly detection value by comprehensive calculation of the coal mine environmental data, data mean of the sliding window, and data standard deviation of the sliding window; Compare the abnormal fluctuation detection value with the second threshold in real time. When the abnormal fluctuation detection value is less than the second threshold, no processing is performed. When the abnormal fluctuation detection value is greater than or equal to the second threshold, a potential abnormal fluctuation is prompted. The sampling frequency is automatically increased according to the intensity of the abnormal fluctuation. During the abnormal fluctuation, the data is automatically corrected using the data smoothing algorithm to avoid the abnormal fluctuation from affecting the system judgment. By analyzing the fluctuation trend, future abnormal fluctuations are predicted in advance and the monitoring strategy is automatically adjusted.

6. A method for self-diagnosis of coal mine safety monitoring data according to claim 5, characterized in that: The specific calculation method of the abnormal fluctuation detection value is: ； In the formula, Indicates the abnormal fluctuation detection value, represents the coal mine environmental data at time point t, Represents a sliding window The mean of the data in Represents a sliding window The standard deviation of the data in .

7. A method for self-diagnosis of coal mine safety monitoring data according to claim 1, characterized in that: The specific process of combining the burst data detection results and the fluctuation anomaly detection results to optimize the alarm triggering conditions and reduce false alarms and missed alarms is as follows: Obtain the burst feature evaluation value and the abnormal fluctuation detection value, and obtain the false alarm and missed alarm optimization value through comprehensive calculation of the burst feature evaluation value and the abnormal fluctuation detection value. Compare the false positive and false negative optimization value with the third threshold in real time. When the false positive and false negative optimization value is less than the third threshold, no processing is performed. When the false alarm and missed alarm optimization value is greater than or equal to the third threshold, the alarm threshold is automatically adjusted to avoid false alarms and missed alarms.

8. A method for self-diagnosis of coal mine safety monitoring data according to claim 7, characterized in that: The specific process of automatically adjusting the dynamic alarm threshold to avoid false alarms and missed alarms is as follows: The dynamic alarm threshold is automatically adjusted according to the false alarm and missed alarm optimization value and the current anomaly detection result. When the coal mine environmental data fluctuates greatly, the dynamic alarm threshold is reduced. At the same time, the dynamic alarm threshold will be automatically updated according to the real-time false alarm and missed alarm optimization value.

9. A method for self-diagnosis of coal mine safety monitoring data according to claim 7, characterized in that: The specific calculation method of the false positive and false negative optimization value is: ； In the formula, Indicates the optimization value of false positives and false negatives. represents the burst characteristic evaluation value, represents the weight coefficient of the burst feature evaluation value, Indicates the abnormal fluctuation detection value, Indicates the weight coefficient of the abnormal fluctuation detection value.

10. A method for self-diagnosis of coal mine safety monitoring data according to claim 1, characterized in that: The specific process of establishing a fault mode library based on historical data and automatically identifying potential fault types is as follows: According to historical failure cases, each failure type is classified with the corresponding characteristic mode, and a failure mode library is established; Use historical data and fault pattern libraries to build a neural network model to learn the characteristic patterns of different fault types; Compare real-time data with historical data in real time and identify potential fault types through pattern matching; Automatically initiate appropriate response measures based on the identified fault type; As new faults occur, new fault data is automatically added to the fault mode library to improve the accuracy of the model.

Images