A control method and system for preventing false alarms in a coal mine system

Abstract

The application provides a control method and system for preventing false alarms of a coal mine system, the method comprising: obtaining detection data of a target sensor and auxiliary data detected by an auxiliary device related to the detection data of the target sensor; determining whether the detection data is greater than a preset threshold, and determining that an alarm event is generated when the detection data is greater than the preset threshold; calculating a consistency index of the detection data and the auxiliary data; determining a confidence score of the alarm event based on the consistency index, the detection data and a machine learning model; determining whether the alarm event is a real alarm based on the confidence score, and issuing a highest-level audible and visual alarm if the alarm event is a real alarm. The application can quantify the credibility of the alarm event, avoid false alarms caused by changes in sensor values due to external interference, realize zero delay or extremely short delay triggering for high-confidence alarms, and the discrimination process can usually be completed within milliseconds, thereby winning valuable time for emergency disposal of real danger.

Description

Technical Field

[0001] This invention relates to the field of coal mine alarm technology, and in particular to a control method and system for preventing false alarms in coal mine systems. Background Technology

[0002] Existing coal mine sensor alarm technologies typically include fixed threshold comparison and fixed delay alarm methods. The fixed threshold comparison method sets a fixed safety threshold (e.g., 1.0% methane concentration) for each monitored parameter. The system continuously compares sensor readings to this threshold, and triggers an alarm immediately if the reading exceeds the threshold. The fixed delay alarm method, on the other hand, adds a delay judgment to filter out brief, transient interference. Only after the sensor reading has consistently exceeded the threshold for a certain period (e.g., 30 seconds) is it considered a valid alarm. However, with the fixed threshold comparison method, the system will alarm unconditionally regardless of whether it's a genuine methane leak or a reading exceeding the limit due to sensor movement or transient interference, resulting in a high probability of false alarms. With the fixed delay alarm method, valuable response time is wasted on real and rapidly developing hazards (e.g., rapid accumulation before a methane explosion), leading to significant safety risks. Summary of the Invention

[0003] In view of this, the purpose of this invention is to provide a control method and system for preventing false alarms in coal mine systems. This method and system can quantify the credibility of the alarm event, avoid false alarms caused by changes in sensor values ​​due to external interference, and achieve zero-delay or extremely short-delay triggering of high-confidence alarms. The intelligent judgment process can usually be completed within milliseconds without waiting, thus winning valuable time for emergency response to real dangers.

[0004] To achieve the above objectives, the technical solutions adopted in the embodiments of the present invention are as follows: In a first aspect, embodiments of the present invention provide a control method for preventing false alarms in coal mine systems, comprising: Acquire the detection data from the target sensor and the auxiliary data detected by the auxiliary device related to the detection data from the target sensor; Determine whether the detected data is greater than a preset threshold. When the detected data is greater than the preset threshold, determine that an alarm event is generated and calculate the consistency index between the detected data and the auxiliary data. The confidence score of the alarm event is determined based on the consistency index, the detection data, and the machine learning model. Based on the confidence score, it is determined whether the alarm event is a real alarm. If the alarm event is a real alarm, the highest level of audible and visual alarm is issued.

[0005] Furthermore, this embodiment of the invention provides a first possible implementation of the first aspect, wherein the step of determining whether the alarm event is a real alarm based on the confidence score includes: When the confidence score is greater than the first score threshold, the alarm event is determined to be a real alarm. When the confidence score is greater than or equal to the second score threshold and less than or equal to the first score threshold, the alarm event is determined to be an alarm pending confirmation. When the confidence score is less than the second score threshold, the alarm event is determined to be a false alarm.

[0006] Furthermore, embodiments of the present invention provide a second possible implementation of the first aspect, which further includes: If the alarm event is an alarm to be confirmed, the alarm event is marked as pending confirmation, a pending confirmation reminder is issued to remind process personnel to check, and the monitoring video corresponding to the alarm event is displayed.

[0007] Furthermore, the embodiments of the present invention provide a third possible implementation of the first aspect, which further includes: If the alarm event is a false alarm, the alarm event is marked as a false alarm state, and the alarm event of the false alarm state, the corresponding consistency index, and the detection data are recorded.

[0008] Furthermore, this embodiment of the invention provides a fourth possible implementation of the first aspect, wherein the step of determining the confidence score of the alarm event based on the consistency index, the detection data, and the machine learning model includes: The consistency index, the detection data, and the health status of the target sensor are input into a pre-trained neural network model to obtain the confidence score of the alarm event output by the neural network model; wherein, the neural network model is trained based on a data sample set labeled with confidence scores, and the data sample set includes the consistency index and the health status of the corresponding sensor for each detection data sample.

[0009] Furthermore, this embodiment of the invention provides a fifth possible implementation of the first aspect, wherein the confidence score ranges from 0 to 1.

[0010] Secondly, embodiments of the present invention also provide a control system for preventing false alarms in coal mine systems, including: a data acquisition module, a multi-source consistency analysis module, a dynamic confidence assessment module, and a graded alarm execution module; The data acquisition module is used to acquire the detection data of the target sensor and the auxiliary data detected by the auxiliary device related to the detection data of the target sensor; The multi-source consistency analysis module is used to determine whether the detection data is greater than a preset threshold. When the detection data is greater than the preset threshold, an alarm event is generated, and the consistency index between the detection data and the auxiliary data is calculated. The dynamic confidence assessment module is used to determine the confidence score of the alarm event based on the consistency index, the detection data, and the machine learning model. The graded alarm execution module is used to determine whether the alarm event is a real alarm based on the confidence score. If the alarm event is a real alarm, the highest level audible and visual alarm is issued.

[0011] Furthermore, the graded alarm execution module is also used to determine that the alarm event is a real alarm when the confidence score is greater than the first score threshold; When the confidence score is greater than or equal to the second score threshold and less than or equal to the first score threshold, the alarm event is determined to be an alarm pending confirmation. When the confidence score is less than the second score threshold, the alarm event is determined to be a false alarm.

[0012] Furthermore, the hierarchical alarm execution module is also used to mark the alarm event as pending confirmation when the alarm event is the pending confirmation alarm, issue a pending confirmation reminder to remind process personnel to check, and display the monitoring video screen corresponding to the alarm event; The hierarchical alarm execution module is further configured to mark the alarm event as a false alarm state when the alarm event is a false alarm, and record the alarm event of the false alarm state, the corresponding consistency index, and the detection data.

[0013] Furthermore, the dynamic confidence assessment module is also used to input the consistency index, the detection data, and the health status of the target sensor into a pre-trained neural network model to obtain the confidence score of the alarm event output by the neural network model; wherein, the neural network model is trained based on a data sample set labeled with confidence scores, and the data sample set includes the consistency index and the health status of the corresponding sensor for each detection data sample.

[0014] This invention provides a control method and system for preventing false alarms in coal mine systems. The method includes: acquiring detection data from a target sensor and auxiliary data detected by auxiliary equipment related to the target sensor's detection data; determining whether the detection data exceeds a preset threshold; when the detection data exceeds the preset threshold, determining that an alarm event has occurred; calculating a consistency index between the detection data and the auxiliary data; determining a confidence score for the alarm event based on the consistency index, the detection data, and a machine learning model; and determining whether the alarm event is a real alarm based on the confidence score. If the alarm event is a real alarm, issuing the highest-level audible and visual alarm. This invention quantifies the credibility of the alarm event by calculating the consistency index between the target sensor's detection data and the related auxiliary data when the sensor's detection data exceeds a threshold, and by calculating the confidence score of the alarm event. This avoids false alarms caused by changes in sensor values ​​due to external interference. Furthermore, by immediately issuing the highest-level audible and visual alarm when the confidence score of the alarm event is high, it achieves zero-delay or extremely short-delay triggering of high-confidence alarms. The intelligent judgment process can typically be completed within milliseconds, eliminating waiting time and saving valuable time for emergency response to real dangers.

[0015] Other features and advantages of the embodiments of the present invention will be set forth in the following description, or some features and advantages may be inferred from the description or determined without doubt, or may be learned by practicing the techniques described above in the embodiments of the present invention.

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0017] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 A flowchart of a control method for preventing false alarms in a coal mine system, provided by an embodiment of the present invention, is shown. Figure 2 The diagram shows a control system structure for preventing false alarms in a coal mine system, as provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be described below in conjunction with the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0020] Currently, existing coal mine sensor alarm technologies still have the following drawbacks: 1. Most existing coal mine systems rely solely on isolated data streams from individual sensors for judgment. The decision-making process is completely detached from the specific environment in which the sensor is located. The system cannot distinguish whether a reading is caused by a real disaster or by non-hazardous factors such as equipment start-up / shutdown, personnel operation, or environmental disturbances. For example, if a gas sensor reading rises, the system cannot autonomously determine whether a coal mining machine is cutting coal nearby (potentially causing a gas outburst) or whether it is simply a movement of the sensor itself. This "data silo" phenomenon is one of the root causes of false alarms.

[0021] 2. The decision-making mechanism is rigid and unable to cope with complex dynamic environments. The decision-making logic of existing coal mine sensor alarm technology is based on preset, static rules, such as fixed thresholds, fixed delays, and fixed voting numbers. The set rules cannot be dynamically adjusted according to the actual situation during operation.

[0022] 3. Alarm results are limited and lack tiered early warning and precise handling guidance. The output of existing coal mine sensor alarm technology is usually binary, that is, "alarm" or "no alarm", which only indicates "there is an anomaly", but cannot provide further judgment information on the nature of the anomaly, the level of confidence, and the possible root causes.

[0023] 4. Poor system interoperability and lack of a closed-loop safety system. Safety monitoring, equipment monitoring, and personnel positioning systems operate independently. Even if data is physically interconnected, there is a lack of deep integration and intelligent linkage at the logical level. When a gas sensor alarms, the system cannot automatically query and correlate video surveillance footage for visual confirmation, nor can it automatically execute preliminary closed-loop control commands such as "pause nearby equipment operation" or "start emergency ventilation." The alarm handling results (whether confirmed as a real accident or a false alarm) cannot be effectively fed back to the system as knowledge accumulation for future optimization.

[0024] To address the aforementioned problems, this invention provides a control method and system for preventing false alarms in coal mine systems. The following provides a detailed description of the embodiments of this invention.

[0025] This embodiment provides a control method for preventing false alarms in a coal mine system, applied to a control system (or device for preventing false alarms in a coal mine system). See [link to relevant documentation]. Figure 1The flowchart shown illustrates a control method for preventing false alarms in coal mine systems. This method mainly includes the following steps: Step S102: Acquire the detection data of the target sensor and the auxiliary data detected by the auxiliary device related to the detection data of the target sensor; Continuously and synchronously acquire target sensor data and its associated multi-source auxiliary data. Deploying a multi-source heterogeneous sensor network in underground coal mines not only includes traditional target sensors (such as sensors that detect important parameters like gas, carbon monoxide, or wind speed sensors), but can also add auxiliary sensors, such as environmental sensors, temperature sensors, and dust concentration sensors.

[0026] The aforementioned multi-source heterogeneous sensors also include equipment status sensors for coal mining equipment, which may include coal mining machines, tunneling machines, and conveyor belts. These equipment status sensors may include sensors that measure equipment status such as voltage, current, vibration, and temperature. The aforementioned multi-source heterogeneous sensors also include position and activity sensors, such as personnel positioning systems and video surveillance systems. The aforementioned multi-source heterogeneous sensors also include sensor self-diagnostic units, installed in each sensor, used to monitor the sensor's own health status (such as zero-point drift and signal strength).

[0027] The control system for preventing false alarms in coal mine systems provided in this embodiment adopts a three-layer architecture of "cloud-edge-device" collaboration. The device side (sensing layer) consists of a multi-source heterogeneous sensor network deployed underground in the coal mine (i.e., data acquisition module).

[0028] The edge computing gateway (decision execution layer) is deployed in key downhole areas. Embedded within this gateway is the core of this invention: a false alarm intelligent identification engine. This engine includes: a data preprocessing module, a multi-source consistency analysis module, a dynamic confidence assessment module, and a tiered alarm execution module. The data preprocessing module filters, denoises, and performs spatiotemporal alignment on the raw data. The multi-source consistency analysis module calculates the consistency index between the target sensor data and auxiliary data locally in real time. The dynamic confidence assessment module runs a lightweight machine learning model and outputs the confidence score of the current alarm event. The tiered alarm execution module executes different alarm strategies based on the confidence score.

[0029] The cloud-side (learning and optimization layer) is located in the ground monitoring center and includes a data warehouse and a case library. It is used to store all historical data, alarm events and final processing results. The dynamic learning and optimization engine uses big data and machine learning algorithms to mine false alarm patterns from the case library, regularly generate update packages, and send them to the edge gateway to optimize its local model.

[0030] Step S104: Determine whether the detection data is greater than a preset threshold. When the detection data is greater than the preset threshold, determine that an alarm event has been generated and calculate the consistency index between the detection data and the auxiliary data. Data is cleaned at the edge gateway, and key features (such as data mutation rate, trend, and correlation coefficient with auxiliary data) are extracted. When the detection data of the target sensor exceeds a preset threshold, an alarm is not immediately triggered; instead, an intelligent verification process is initiated: a physical model or statistical model is invoked to analyze whether the current data pattern conforms to the characteristics of a real disaster. The statistical model can be used to calculate the correlation coefficient between the data mutation rate or trend of the target sensor's detection data and that of the auxiliary data. A higher consistency index indicates a closer similarity between the two data mutation rates or trends; conversely, a lower consistency index indicates a larger difference between the two. For example, a real gas outburst is usually accompanied by coal cutting by a mining machine (increased equipment current) or roof pressure (micro-vibration event). If only gas readings change in isolation, the consistency index will be low.

[0031] Through multi-source information fusion and consistency analysis, the system can effectively distinguish between real disaster signals and false signals caused by sensor malfunctions or environmental interference (such as humidity and dust). The multi-source consistency analysis module can also identify potential risk patterns from massive amounts of data. For example, by analyzing the weak correlation between equipment vibration data and gas concentration, early warnings such as "sensor readings are at risk of drifting" or "risk level in a certain area is rising" can be issued before sensor readings officially exceed limits, thus shifting from post-event alarms to pre-event warnings.

[0032] Step S106: Determine the confidence score of the alarm event based on the consistency index, detection data, and machine learning model; The consistency index, data characteristics, and sensor health status are input into a lightweight machine learning model (such as a neural network model) to calculate a confidence score between 0 and 1. Each alarm event is accompanied by a confidence score, providing dispatchers with important decision-making references. The system actively suppresses low-confidence alarms, fundamentally improving the proportion (accuracy) of alarm information that ultimately reaches the dispatcher as having a true threat.

[0033] Step S108: Determine whether the alarm event is a real alarm based on the confidence score. If the alarm event is a real alarm, issue the highest level audible and visual alarm.

[0034] When the confidence score is greater than the first score threshold, the alarm event with a high confidence score is determined to be a real alarm, and the highest level of audible and visual alarm is immediately triggered, and the emergency system is activated.

[0035] The processing results of all alarm events (whether automatically determined by the system or manually confirmed) are recorded and fed back to the cloud case library. The cloud-based dynamic learning engine regularly retrains the neural network model with new case data, and sends the optimized neural network model parameters and rules to the edge gateway, enabling it to continuously evolve.

[0036] Unlike the fixed delay method, this embodiment uses zero-delay or extremely short-delay triggering for high-confidence alarms. Because the intelligent judgment process is completed at the edge in milliseconds, there is no waiting time, thus gaining valuable time for emergency response to real dangers.

[0037] By deeply integrating safety monitoring, equipment monitoring, and personnel positioning systems at both the data and decision-making levels, information silos have been broken down, forming a unified mine-wide safety situational awareness. This lays a key technological foundation for building a "smart mine brain," creating a closed-loop management system that achieves a complete cycle from "perception-analysis-decision-execution-feedback-learning." This transforms the safety management process from an open-loop, passive response to a closed-loop, proactive, and continuously optimized intelligent control.

[0038] The control method for preventing false alarms in coal mine systems provided in this embodiment calculates the consistency index between the detection data of the target sensor and the relevant auxiliary data when the sensor's detection data exceeds a threshold, and calculates the confidence score of the alarm event to quantify the credibility of the alarm event. This avoids false alarms caused by changes in sensor values ​​due to external interference. At the same time, by immediately issuing the highest level audible and visual alarm when the confidence score of the alarm event is high, it achieves zero-delay or extremely short-delay triggering of high-confidence alarms. The intelligent judgment process can usually be completed within milliseconds without waiting, winning valuable time for emergency response to real dangers.

[0039] In one implementation, the confidence score ranges from 0 to 1. This embodiment provides a specific implementation method for determining whether an alarm event is a real alarm based on the confidence score: When the confidence score is greater than the first score threshold (the first score threshold can be in the range of 0.7 to 0.9, with 0.8 preferred), the alarm event is determined to be a real alarm. When the confidence score is greater than or equal to the second score threshold (the value of the first score threshold can be, for example, 0.3~0.5, with 0.4 preferred) and less than or equal to the first score threshold, the alarm event is determined to be an alarm pending confirmation; the medium confidence score is determined to be a suspected alarm, triggering the "pending confirmation" state, notifying nearby personnel to carry a portable device for on-site verification, and automatically popping up the relevant video screen.

[0040] In one implementation, if the alarm event is an alarm pending confirmation, the alarm event is marked as pending confirmation, a pending confirmation reminder is issued to remind process personnel to check, and the monitoring video corresponding to the alarm event is displayed.

[0041] When the confidence score is less than the second threshold, the alarm event is determined to be a false alarm. Low confidence scores are also classified as false alarms; the system automatically logs the event but does not trigger audible or visual alarms to avoid disrupting production.

[0042] In one implementation, if the alarm event is a false alarm, no alarm information is issued, the alarm event is marked as a false alarm, and the alarm event in the false alarm state, along with the corresponding consistency index and detection data, are recorded.

[0043] By determining the alarm status based on the confidence score of alarm events and executing corresponding strategies, the system's decision-making no longer relies entirely on fixed rules preset by engineers. Instead, it makes judgments autonomously through data-driven approaches, reducing the skill requirements for manual debugging and maintenance. This allows the system to be quickly deployed and maintain high performance in different mines and under different conditions. It also reduces ineffective emergency responses by significantly decreasing false alarms, avoiding unplanned production shutdowns and unnecessary rescue deployments caused by false alarms, directly saving substantial production time and economic costs.

[0044] In one implementation, this embodiment provides a specific implementation method for determining the confidence score of an alarm event based on a consistency index, detection data, and a machine learning model: The consistency index, detection data, and the health status of the target sensor are input into a pre-trained neural network model to obtain the confidence score of the alarm event output by the neural network model. The neural network model is trained based on a data sample set labeled with confidence scores. The data sample set includes the consistency index (i.e., the consistency index between the detection data sample and the auxiliary data collected at the same time) and the health status of the corresponding sensor (i.e., the sensor that collected the detection data sample) for each detection data sample.

[0045] The control method for preventing false alarms in coal mine systems provided in this embodiment fundamentally solves the problem of false alarms from coal mine sensors by constructing an intelligent closed-loop system of "perception-fusion-decision-feedback-learning". By adopting a perception-based multi-source information fusion false alarm identification technology, the limitations of traditional single sensor data criteria have been overcome. It creatively introduces data such as equipment operating status, environmental parameters, and personnel activity information into the false alarm identification process. Consistency verification is carried out by combining physical models with data-driven methods, which fundamentally solves the problem of false alarms that cannot be identified due to "information silos" and realizes the leap from "single-point judgment" to "global correlation analysis". By adopting a cloud-edge collaborative dynamic learning and evolution mechanism, a collaborative architecture was designed that places a lightweight evaluation model on the edge side to ensure real-time performance, and places complex model training and optimization on the cloud to ensure the algorithm's advancement. The system can automatically learn new false alarm patterns from human feedback and historical data, and dynamically update the edge-side model to achieve "the more it is used, the more accurate it becomes." This solves the problem that traditional static systems cannot adapt to the dynamic changes in the downhole environment and rely on frequent manual debugging, enabling the system to have the ability to self-optimize and continuously evolve. By adopting a multi-level alarm and handling strategy based on confidence scores, a hierarchical and differentiated response mechanism was proposed to transform the single binary alarm (yes / no) into a hierarchical and differentiated response mechanism based on continuous confidence scores. Alarm events are divided into three levels: "confirmed alarm", "alarm pending verification" and "false alarm silent log", and each level corresponds to a different handling process. This effectively solves the problem of production interruption caused by false alarms and improves the accuracy of dispatching and command and the utilization efficiency of emergency resources. By adopting a lightweight and interpretable confidence assessment model for the mining environment, and considering the limited resources of underground edge computing equipment, a lightweight machine learning model was specifically designed or optimized for confidence assessment. At the same time, the model has a certain degree of interpretability, which can inform the dispatcher of the main basis for the judgment (such as "due to no change in wind speed, it is judged as low confidence"). While ensuring the performance of the algorithm, it meets the strict requirements of real-time performance, reliability and interpretability in the mining industrial scenario.

[0046] Corresponding to the control method for preventing false alarms in coal mine systems provided in the above embodiments, this invention provides a control system for preventing false alarms in coal mine systems, see [link to relevant documentation]. Figure 2 The diagram shows a control system structure for preventing false alarms in a coal mine system. The control system includes: a data acquisition module 21, a multi-source consistency analysis module 22, a dynamic confidence assessment module 23, and a hierarchical alarm execution module 24. The data acquisition module 21 is used to acquire the detection data of the target sensor and the auxiliary data detected by the auxiliary equipment related to the detection data of the target sensor; The multi-source consistency analysis module 22 is used to determine whether the detection data is greater than a preset threshold. When the detection data is greater than the preset threshold, an alarm event is generated, and the consistency index between the detection data and the auxiliary data is calculated. The dynamic confidence assessment module 23 is used to determine the confidence score of alarm events based on the consistency index, detection data, and machine learning model; The graded alarm execution module 24 is used to determine whether an alarm event is a real alarm based on the confidence score. If the alarm event is a real alarm, the highest level of audible and visual alarm is issued.

[0047] The control system for preventing false alarms in coal mine systems provided in this embodiment calculates the consistency index between the detection data of the target sensor and the relevant auxiliary data when the sensor's detection data exceeds a threshold, and calculates the confidence score of the alarm event to quantify the credibility of the alarm event. This avoids false alarms caused by changes in sensor values ​​due to external interference. At the same time, by immediately issuing the highest level audible and visual alarm when the confidence score of the alarm event is high, it achieves zero-delay or extremely short-delay triggering of high-confidence alarms. The intelligent judgment process can usually be completed within milliseconds without waiting, winning valuable time for emergency response to real dangers.

[0048] In one implementation, the graded alarm execution module provided in this embodiment is further used to determine that the alarm event is a real alarm when the confidence score is greater than the first score threshold. When the confidence score is greater than or equal to the second score threshold and less than or equal to the first score threshold, the alarm event is determined to be an alarm pending confirmation. When the confidence score is less than the second score threshold, the alarm event is determined to be a false alarm.

[0049] In one embodiment, the hierarchical alarm execution module provided in this embodiment is further used to mark the alarm event as pending confirmation when the alarm event is a pending confirmation alarm, issue a pending confirmation reminder to remind process personnel to check, and display the monitoring video screen corresponding to the alarm event; In one embodiment, the hierarchical alarm execution module provided in this embodiment is further used to mark the alarm event as a false alarm state when the alarm event is a false alarm, and record the alarm event of the false alarm state and the corresponding consistency index and detection data.

[0050] In one embodiment, the dynamic confidence assessment module provided in this embodiment is further used to input the consistency index, detection data, and the health status of the target sensor into a pre-trained neural network model to obtain the confidence score of the alarm event output by the neural network model; wherein, the neural network model is trained based on a data sample set labeled with confidence scores, and the data sample set includes the consistency index and the health status of the corresponding sensor for each detection data sample.

[0051] The system provided in this embodiment has the same implementation principle and technical effects as the aforementioned embodiments. For the sake of brevity, any parts not mentioned in the system embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0052] This invention provides an electronic device, which includes a processor and a memory. The memory stores a computer program that can run on the processor. When the processor executes the computer program, it implements the steps of the method provided in the above embodiments.

[0053] This invention provides a computer-readable medium storing computer-executable instructions. When these computer-executable instructions are invoked and executed by a processor, they cause the processor to implement the methods described in the above embodiments.

[0054] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system described above can be referred to the corresponding process in the foregoing embodiments, and will not be repeated here.

[0055] The computer program product of the control method and system for preventing false alarms in coal mine systems provided by the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the preceding method embodiments. For specific implementation, please refer to the method embodiments, which will not be repeated here.

[0056] Furthermore, in the description of the embodiments of the present invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in the present invention based on the specific circumstances.

[0057] 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 invention, essentially, 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 invention. 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.

[0058] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0059] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of protection of the claims.

Claims

1. A control method for preventing false alarms in coal mine systems, characterized in that, include: Acquire the detection data from the target sensor and the auxiliary data detected by the auxiliary device related to the detection data from the target sensor; Determine whether the detected data is greater than a preset threshold. When the detected data is greater than the preset threshold, determine that an alarm event is generated and calculate the consistency index between the detected data and the auxiliary data. The confidence score of the alarm event is determined based on the consistency index, the detection data, and the machine learning model. Based on the confidence score, it is determined whether the alarm event is a real alarm. If the alarm event is a real alarm, the highest level of audible and visual alarm is issued.

2. The method according to claim 1, characterized in that, The step of determining whether the alarm event is a real alarm based on the confidence score includes: When the confidence score is greater than the first score threshold, the alarm event is determined to be a real alarm. When the confidence score is greater than or equal to the second score threshold and less than or equal to the first score threshold, the alarm event is determined to be an alarm pending confirmation. When the confidence score is less than the second score threshold, the alarm event is determined to be a false alarm.

3. The method according to claim 2, characterized in that, Also includes: If the alarm event is an alarm to be confirmed, the alarm event is marked as pending confirmation, a pending confirmation reminder is issued to remind process personnel to check, and the monitoring video corresponding to the alarm event is displayed.

4. The method according to claim 2, characterized in that, Also includes: If the alarm event is a false alarm, the alarm event is marked as a false alarm state, and the alarm event of the false alarm state, the corresponding consistency index, and the detection data are recorded.

5. The method according to claim 1, characterized in that, The step of determining the confidence score of the alarm event based on the consistency index, the detection data, and the machine learning model includes: The consistency index, the detection data, and the health status of the target sensor are input into a pre-trained neural network model to obtain the confidence score of the alarm event output by the neural network model; wherein, the neural network model is trained based on a data sample set labeled with confidence scores, and the data sample set includes the consistency index and the health status of the corresponding sensor for each detection data sample.

6. The method according to claim 1, characterized in that, The confidence score ranges from 0 to 1.

7. A control system for preventing false alarms in coal mine systems, characterized in that, include: The system includes a data acquisition module, a multi-source consistency analysis module, a dynamic confidence assessment module, and a tiered alarm execution module. The data acquisition module is used to acquire the detection data of the target sensor and the auxiliary data detected by the auxiliary device related to the detection data of the target sensor; The multi-source consistency analysis module is used to determine whether the detection data is greater than a preset threshold. When the detection data is greater than the preset threshold, an alarm event is generated, and the consistency index between the detection data and the auxiliary data is calculated. The dynamic confidence assessment module is used to determine the confidence score of the alarm event based on the consistency index, the detection data, and the machine learning model. The graded alarm execution module is used to determine whether the alarm event is a real alarm based on the confidence score. If the alarm event is a real alarm, the highest level audible and visual alarm is issued.

8. The control system according to claim 7, characterized in that, The graded alarm execution module is also used to determine that the alarm event is a real alarm when the confidence score is greater than the first score threshold. When the confidence score is greater than or equal to the second score threshold and less than or equal to the first score threshold, the alarm event is determined to be an alarm pending confirmation. When the confidence score is less than the second score threshold, the alarm event is determined to be a false alarm.

9. The control system according to claim 8, characterized in that, The hierarchical alarm execution module is also used to mark the alarm event as pending confirmation when the alarm event is the pending confirmation alarm, issue a pending confirmation reminder to remind process personnel to check, and display the monitoring video screen corresponding to the alarm event; The hierarchical alarm execution module is further configured to mark the alarm event as a false alarm state when the alarm event is a false alarm, and record the alarm event of the false alarm state, the corresponding consistency index, and the detection data.

10. The control system according to claim 7, characterized in that, The dynamic confidence assessment module is further used to input the consistency index, the detection data, and the health status of the target sensor into a pre-trained neural network model to obtain the confidence score of the alarm event output by the neural network model; wherein, the neural network model is trained based on a data sample set labeled with confidence scores, and the data sample set includes the consistency index and the health status of the corresponding sensor for each detection data sample.

Images