Mine safety risk prediction system and method
By collecting underground data in real time and performing comprehensive diagnosis and prediction, combined with video monitoring and equipment linkage control, the problem of accurate prediction of mine safety risks has been solved, and the safety of underground mine operations has been improved.
Patent Information
- Application Number
- CN202511564183.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-30
- Publication Date
- 2026-02-13
AI Technical Summary
How to accurately predict mine safety risks to ensure the safety of underground operations, especially in complex situations such as equipment malfunctions, gas leaks, and personnel operations.
The sensing module collects real-time location information of personnel underground, environmental information, and equipment operating parameters. Combined with video monitoring, it performs abnormal personnel behavior diagnosis and equipment status diagnosis. The safety risk prediction module is used to predict risks, and the emergency dispatch module sends early warning notifications and equipment linkage control.
It enables accurate prediction and timely response to mine safety risks, reduces the occurrence of safety accidents, and improves the safety of underground mine operations.
Smart Images
Figure CN121526293A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of mine safety monitoring, in particular to a mine safety risk prediction system and method. BACKGROUND
[0002] A mine refers to a mine exploited underground, and sometimes the inclined shaft, vertical shaft, adit, etc. in the underground development of a mine are also called mines. The mine development has a significant and far-reaching impact on the overall production and construction of a metal mine or a coal mine. The situation underground is complex, and the abnormality of equipment, gas leakage, and personnel operation can all have a significant impact on the safety of the mine. In order to ensure the safety of the underground operation, it is particularly important to accurately predict the safety risk of the mine. SUMMARY
[0003] The present application aims to provide a mine safety risk prediction system and method, which combines diagnosis and prediction to predict risks based on the monitoring data of man-machine-environment in multiple aspects, so as to timely respond to accidents that have not occurred and further ensure the safety of the underground operation of the mine.
[0004] In a first aspect, the present application provides a mine safety risk prediction system, comprising: a perception module configured to collect positioning information of underground personnel, environmental information, equipment operation parameter information, and video monitoring information underground in real time; a personnel behavior anomaly diagnosis module configured to perform anomaly behavior diagnosis based on the video monitoring information and the positioning information at the current time, and obtain a personnel behavior anomaly diagnosis result at the current time; an equipment state diagnosis module configured to perform state diagnosis based on the equipment operation parameter information at the current time, and obtain an equipment state diagnosis result at the current time; a safety risk prediction module configured to perform risk prediction based on the personnel behavior anomaly diagnosis result, the equipment state diagnosis result, the environmental information, and the equipment operation parameter information at each time within a current time period, and obtain a safety risk prediction result at the current time; the current time period comprises the current time and a plurality of continuous times before the current time; an emergency dispatch module configured to send an early warning notice and / or equipment linkage control based on the personnel anomaly diagnosis result, the equipment state diagnosis result, and the safety risk prediction result at the current time.
[0005] In some embodiments, the perception module comprises a positioning unit, an environmental monitoring unit, an equipment monitoring unit, and a video monitoring unit; wherein, the positioning unit is configured to obtain the positioning information of the underground personnel through a fixed base station deployed in a tunnel and a wearable device worn by the underground personnel; The environment monitoring unit is configured to collect the environment information through a sensor deployed underground. The device monitoring unit is configured to collect the device operation parameter information. The video monitoring unit is configured to collect the video monitoring information based on an explosion-proof camera deployed underground.
[0006] As an example, the environment information includes at least one of temperature, smoke concentration, gas concentration, dust concentration, oxygen concentration, humidity, and air pressure.
[0007] In some embodiments, the personnel behavior anomaly diagnosis module includes a target detection unit, a posture extraction unit, a behavior detection unit, a matching unit, and a generation unit, wherein, The target detection unit is configured to perform target detection on personnel and objects in the video monitoring information at the current time by using a YOLOv5 algorithm to obtain a target recognition result. The target recognition result includes a pixel region where the personnel are located, a pixel region where the objects are located, and an object category. The posture extraction unit is configured to perform key point feature extraction on the pixel region where the personnel are located to obtain key point feature information. The behavior detection unit is configured to perform behavior detection based on the key point feature information, the pixel region where the objects are located, and the object category to obtain a target personnel with an abnormal behavior and abnormal behavior content of the target personnel. The matching unit is configured to match the coordinate information determined based on the pixel region where the target personnel are located with the positioning information to determine identity information of the target personnel. The generation unit is configured to generate the personnel behavior anomaly diagnosis result based on the identity information of the target personnel and the abnormal behavior content of the target personnel.
[0008] In some embodiments, the safety risk prediction module includes a data preprocessing unit, a feature extraction unit, and a risk prediction unit, wherein, The data preprocessing unit is configured to perform time and space alignment processing on the personnel behavior anomaly diagnosis results, the device state diagnosis results, the environment information, and the device operation parameter information at each time in the current period. The feature extraction unit is configured to perform time-frequency domain feature extraction on the data processed by the data preprocessing unit to obtain extracted feature information. The risk prediction unit is configured to input the feature information into a time series risk prediction model to obtain a safety risk prediction result output by the time series risk prediction model.
[0009] In some embodiments, the device state diagnosis module includes a first diagnosis unit, a second diagnosis unit, and a diagnosis result output unit, wherein, The first diagnosis unit is configured to compare the device operation parameter information at the current time with corresponding threshold values, and obtain a first diagnosis result. The second diagnosis unit is configured to extract device-related video monitoring data from the video monitoring data, and perform abnormality recognition on the device-related video monitoring data, and obtain a second diagnosis result. The diagnosis result output unit is configured to output the device state diagnosis result based on the first diagnosis result and the second diagnosis result.
[0010] As a possible implementation, the emergency dispatch module is configured to send the early warning notification to a wearable device of a relevant person.
[0011] As an example, the early warning notification can be in any of the following forms: alarm sound, text, light, and voice broadcast.
[0012] In a second aspect, the present application provides a mine safety risk prediction method applied to the mine safety risk prediction system of the first aspect, comprising: The perception module is configured to collect positioning information, environmental information, device operation parameter information, and underground video monitoring information of underground personnel in real time; the positioning information comprises a correspondence between an identity of the underground personnel and positioning information of the underground personnel; The personnel behavior abnormality diagnosis module is configured to perform abnormality behavior diagnosis based on the video monitoring information and the positioning information at the current time, and obtain a personnel behavior abnormality diagnosis result at the current time; The device state diagnosis module is configured to perform state diagnosis based on the device operation parameter information at the current time, and obtain a device state diagnosis result at the current time; The safety risk prediction module is configured to perform risk prediction based on the personnel behavior abnormality diagnosis result, the device state diagnosis result, the environmental information, and the device operation parameter information at each time within a current time period, and obtain a safety risk prediction result at the current time; the current time period comprises the current time and a plurality of continuous times before the current time; The emergency dispatch module is configured to send an early warning notification and / or device linkage control based on the personnel abnormality diagnosis result, the device state diagnosis result, and the safety risk prediction result at the current time.
[0013] The environmental information comprises at least one of the following: temperature, smoke concentration, gas concentration, dust concentration, oxygen concentration, humidity, and air pressure.
[0014] The mine safety risk prediction system and method provided by the present application have the following beneficial effects: The application acquires the positioning information, environment information, device running parameter information and video monitoring information of the downhole personnel in real time through the perception module; the personnel behavior abnormality diagnosis module performs abnormal behavior diagnosis based on the video monitoring information and positioning information at the current time to obtain the personnel behavior abnormality diagnosis result at the current time; the device state diagnosis module performs state diagnosis based on the device running parameter information at the current time to obtain the device state diagnosis result at the current time; the safety risk prediction module performs risk prediction based on the personnel behavior abnormality diagnosis result, device state diagnosis result, environment information and device running parameter information at each time in the current period to obtain the safety risk prediction result at the current time; the emergency dispatch module sends the early warning notice and / or device linkage control based on the personnel abnormality diagnosis result, device state diagnosis result and safety risk prediction result at the current time. The application combines diagnosis and prediction, predicts the risk by monitoring data of man-machine-environment, responds to the accident in time, and further guarantees the safety of the downhole operation of the mine. BRIEF DESCRIPTION OF DRAWINGS
[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only constitute a part of the embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor based on the provided drawings.
[0016] Figure 1 A structural schematic diagram of a mine safety risk prediction system provided for the embodiment of the present application; Figure 2 A flow schematic diagram of a mine safety risk prediction method provided for the embodiment of the present application. DETAILED DESCRIPTION
[0017] The technical solutions of the present application will be described clearly and completely in combination with the drawings. Obviously, the described embodiments are a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.
[0018] The present application provides a mine safety risk prediction system and method, which realizes the prediction of the safety risk of the mine, reduces the probability of the safety accident in the mine, and further guarantees the safety of the downhole operation of the mine.
[0019] Figure 1 A structural schematic diagram of a mine safety risk prediction system provided for the embodiment of the present application. As shown in Figure 1As shown, the system includes a perception module 110, a personnel behavior anomaly diagnosis module 120, a device state diagnosis module 130, a safety risk prediction module 140, and an emergency dispatch module 150.
[0020] In some embodiments, the perception module 110 is configured to collect positioning information of personnel underground, environmental information, device operation parameter information, and video monitoring information underground in real time; the personnel behavior anomaly diagnosis module 120 is configured to perform anomaly behavior diagnosis based on the video monitoring information and the positioning information at the current time, and obtain a personnel behavior anomaly diagnosis result at the current time; the device state diagnosis module 130 is configured to perform state diagnosis based on the device operation parameter information at the current time, and obtain a device state diagnosis result at the current time; the safety risk prediction module 140 is configured to perform risk prediction based on the personnel behavior anomaly diagnosis result, the device state diagnosis result, the environmental information, and the device operation parameter information at each time within a current time period, and obtain a safety risk prediction result at the current time; the current time period includes the current time and a plurality of continuous times before the current time; and the emergency dispatch module 150 is configured to send an early warning notification and / or device linkage control based on the personnel anomaly diagnosis result, the device state diagnosis result, and the safety risk prediction result at the current time.
[0021] As an example, the environmental information can include at least one of the following: temperature, smoke concentration, gas concentration, dust concentration, oxygen concentration, humidity, air pressure, and the like, which are collected by corresponding sensors.
[0022] As an example, the positioning information of the personnel underground can be collected by the GPS signal of a handheld terminal equipped by each personnel underground.
[0023] In some embodiments, the devices underground include ventilation devices, electromechanical devices, fluid devices, transportation devices, and the like. For the electromechanical devices, the operation parameter information of the devices can include motor current, voltage, power, temperature, vibration frequency and amplitude, and the like; for the fluid devices, the operation parameter information of the devices can include pump outlet pressure, flow rate, bearing temperature, and the like; for the ventilation devices, the operation parameter information of the devices can include total air volume, air pressure, fan speed, efficiency, and the like; and for the transportation devices, the operation parameter of the devices can include belt conveyor speed, load current, tension, and the like.
[0024] In some embodiments, the operation parameter information of the devices can be obtained by a PLC (Programmable Controllers) control system of the devices, or by corresponding sensors of the devices.
[0025] As a possible implementation manner, the personnel behavior anomaly diagnosis result can include a personnel identity and an abnormal behavior content, where the personnel identity is obtained through matching with the positioning information. The perception module can collect positioning information of each personnel, such as the positioning information can be in a corresponding relationship between the personnel identity and the positioning information. The abnormal behavior content refers to a description of a specific abnormal behavior.
[0026] In some embodiments, a personnel abnormal behavior database including abnormal behavior features and corresponding abnormal behavior contents can be pre-set, and the behavior feature information of the personnel in the video monitoring is matched with the abnormal behavior features in the personnel abnormal behavior database to determine the abnormal behavior content of the personnel.
[0027] In some embodiments, the safety risk prediction module 140 is equivalent to associating and analyzing the diagnosis results of the equipment state diagnosis module 130 and the personnel behavior anomaly diagnosis module 120 in a continuous period of time, and the environment information and the equipment operation parameter information collected by the perception module 110, to obtain a safety risk that can occur in the future. The safety risk prediction module 140 can implement risk prediction based on a deep learning neural network model that has been trained. As an example, the safety risk prediction module 140 can implement risk prediction based on an LSTM (Long Short Term Memory) model.
[0028] In some embodiments, the perception module 110 includes a positioning unit 111, an environment monitoring unit 112, an equipment monitoring unit 113, and a video monitoring unit 114. The positioning unit 111 is configured to obtain positioning information of personnel in the mine through fixed base stations deployed in the roadway and wearable devices worn by the personnel in the mine. The fixed base stations can be integrated into roadway substation equipment and mine lamps, and the wearable devices of the personnel in the mine include a communication unit that communicates with the fixed base stations, thereby realizing the positioning of the personnel in the mine. The wearable device can be a smart helmet, a tool, a watch, or the like. The environment monitoring unit 112 is configured to collect environment information through sensors deployed in the mine. The equipment monitoring unit 113 is configured to collect equipment operation parameter information. The video monitoring unit 114 is configured to collect video monitoring information based on an explosion-proof camera deployed in the mine.
[0029] In some embodiments, the personnel behavior anomaly diagnosis module 120 includes a target detection unit 121, a posture extraction unit 122, a behavior detection unit 123, a matching unit 124, and a generation unit 125.
[0030] The target detection unit 121 is configured to perform target detection on the personnel and objects in the video monitoring information at the current moment by using a YOLOv5 algorithm to obtain a target recognition result. The target recognition result includes a pixel region where the personnel are located, a pixel region where the objects are located, and an object category. The posture extraction unit 122 is configured to perform key point feature extraction on the pixel region where the personnel are located to obtain key point feature information. The behavior detection unit 123 is configured to perform behavior detection based on the key point feature information, the pixel region where the objects are located, and the object category to obtain a target personnel with an abnormal behavior and abnormal behavior content of the target personnel. The matching unit 124 is configured to match the coordinate information determined based on the pixel region where the target personnel are located with the positioning information to determine identity information of the target personnel. The generation unit 125 is configured to generate a personnel behavior anomaly diagnosis result based on the identity information of the target personnel and the abnormal behavior content of the target personnel.
[0031] As an example, the posture extraction unit 122 can perform skeleton key point feature extraction on the pixel region where the personnel are located, such as feature extraction on 17 key points including the head, shoulders, elbows, knees, etc., to form a human skeleton feature map, i.e., key point feature information.
[0032] Since the target detection module can only identify people and objects but cannot confirm the identity of the personnel, the identity information corresponding to the personnel identified by the target detection module can be determined by matching the positioning information of each personnel obtained by the perception module 110. As a possible implementation, the matching unit 124 can determine the relative coordinate information of the target personnel in the video based on the pixel region where the target personnel are located. Based on the global coordinate information of the camera, the relative coordinate information of the target personnel in the video can be converted into global coordinate information. The global coordinate information of the target personnel is matched with the positioning information of each personnel, and the identity information of the personnel corresponding to the positioning information that matches successfully is determined as the identity information of the target personnel.
[0033] As an example, when performing coordinate matching, the matching unit 124 can determine the personnel corresponding to the positioning information with the smallest difference from the global coordinate information as the target personnel, and determine the identity corresponding thereto as the identity information of the target personnel. The identity information can be an identity identifier of each personnel, such as a name, a work number, etc.
[0034] In some embodiments, the safety risk prediction module 140 comprises a data preprocessing unit 141, a feature extraction unit 142, and a risk prediction unit 143. Among them, the data preprocessing unit 141 is configured to perform time and space alignment processing on the personnel behavior anomaly diagnosis result, the equipment state diagnosis result, the environmental information and the equipment operation parameter information at each time point in the current period; the feature extraction unit 142 is configured to perform time-frequency domain feature extraction on the data processed by the data preprocessing unit, and obtain the extracted feature information; and the risk prediction unit 143 is configured to input the feature information into a time sequence risk prediction model to obtain a safety risk prediction result output by the time sequence risk prediction model.
[0035] As an example, the feature extraction unit 142 can extract features such as the change trend and change amplitude of the equipment operation parameters, the change trend and change amplitude of the environmental information, the number of times of the same person appearing abnormal behavior, the frequency of the same abnormal behavior, the frequency of the equipment state anomaly, and the number of times of the same equipment anomaly.
[0036] As a possible implementation manner, the risk prediction unit 143 is configured to predict the future safety risk that will occur based on the feature information and the correlation between the features. As an example, the risk prediction unit 143 can realize safety risk prediction based on a pre-constructed knowledge graph, and the knowledge graph contains the direct correlation between all feature information and safety risk.
[0037] In order to ensure the accuracy of the equipment state diagnosis and avoid omission of abnormal states, as shown in Figure 1 The equipment state diagnosis module 130 can comprise a first diagnosis unit 131, a second diagnosis unit 132, and a diagnosis result output unit 133. Among them, the first diagnosis unit 131 is configured to compare the equipment operation parameter information at the current time point with the corresponding threshold value to obtain a first diagnosis result; the second diagnosis unit 132 is configured to extract equipment-related video monitoring data from the video monitoring data, and perform abnormal identification on the equipment-related video monitoring data to obtain a second diagnosis result; and the diagnosis result output unit 133 is configured to output the equipment state diagnosis result based on the first diagnosis result and the second diagnosis result. As an example, the first diagnosis result and the second diagnosis result are taken as a union set.
[0038] In some embodiments, the emergency dispatch module 150 is further configured to send the early warning notification to the wearable device of the relevant personnel, and timely notify the personnel of evacuation information and the like.
[0039] In some embodiments, the early warning notification can be in any of the following forms: alarm sound form, text form, light form, and voice broadcast form.
[0040] In some embodiments, the emergency dispatch module 150 will promptly issue early warning notifications and / or implement equipment linkage control upon receiving personnel anomaly diagnosis results, equipment status diagnosis results, and safety risk prediction results, in order to handle emergencies in a timely manner. In other words, the emergency dispatch module 150 not only processes the current diagnosis results but also promptly notifies users of predicted risks, thereby maximizing the prevention of accidents.
[0041] As an example, if the personnel behavior anomaly diagnosis module outputs a diagnosis result that personnel A has entered the danger zone, the emergency dispatch module 150 can directly issue a voice warning, "Personnel A, please leave the danger zone in time." If the equipment status diagnosis module 140 outputs at 10:00:30 that the motor current of water pump No. 1 fluctuates, and the seepage rate of the tunnel top increases between 10:00:00 and 10:05:00, and the safety risk prediction module 150 outputs that there is a risk of water inrush in the area surrounding drainage point No. 1, the emergency dispatch module 150 will issue a warning notification to the corresponding personnel so that drainage point No. 1 can be inspected in time and the backup pump can be started.
[0042] As an example, if the equipment status diagnosis result at a certain moment is that the vibration frequency of the ventilation fan has increased, and the gas concentration in the corresponding time period shows an upward trend, the safety risk prediction module 150 can determine through correlation analysis that the ventilation fan failure has led to an increase in gas concentration, and predict the risk of gas accumulation. Thus, the correlation analysis can be used to make accurate predictions of future risks to a large extent, thereby improving the safety of mine operations.
[0043] As an example, if personnel operate the equipment improperly, directly affecting their safety, the emergency dispatch module can directly control the equipment to shut down after receiving the abnormal personnel diagnosis results, in order to avoid an accident.
[0044] The mine safety risk prediction system according to embodiments of the present invention, through the combination of diagnosis and prediction, and the correlation analysis of monitoring data from multiple aspects such as humans, machines, and environment, can accurately predict future mine safety risks, promptly respond to and handle accidents that have not yet occurred, and further ensure the safety of underground mine operations.
[0045] This invention also provides a method for predicting mine safety risks.
[0046] Figure 2 This is a flowchart illustrating a mine safety risk prediction method provided in an embodiment of the present invention. The mine safety risk prediction method of this embodiment is applied to the mine safety risk prediction system described in the above embodiment. Figure 2 As shown, the method includes the following steps: Step 201, the positioning information, environment information, device running parameter information and video monitoring information of the downhole personnel are collected in real time by the perception module.
[0047] Step 202, the personnel behavior anomaly diagnosis module performs anomaly behavior diagnosis based on the video monitoring information and the positioning information at the current time, and obtains the personnel behavior anomaly diagnosis result at the current time.
[0048] Step 203, the device state diagnosis module performs state diagnosis based on the device running parameter information at the current time, and obtains the device state diagnosis result at the current time.
[0049] Step 204, the safety risk prediction module performs risk prediction based on the personnel behavior anomaly diagnosis result, the device state diagnosis result, the environment information and the device running parameter information at each time in the current period, and obtains the safety risk prediction result at the current time; the current period includes the current time and a plurality of continuous times before the current time.
[0050] Step 205, the emergency dispatch module sends an early warning notice and / or device linkage control based on the personnel anomaly diagnosis result, the device state diagnosis result and the safety risk prediction result at the current time.
[0051] In some embodiments, the implementation process of step 201 can include: based on the positioning unit of the perception module, obtaining the positioning information of the downhole personnel by the fixed base station deployed in the tunnel and the wearable device worn by the downhole personnel; based on the environment monitoring unit, collecting the environment information by the sensor deployed underground; based on the device monitoring unit, collecting the device running parameter information; based on the video monitoring unit, collecting the video monitoring information based on the explosion-proof camera deployed underground.
[0052] In some embodiments, the environment information includes at least one of the following: temperature, smoke concentration, gas concentration, dust concentration, oxygen concentration, humidity, air pressure.
[0053] In some embodiments, the implementation process of step 202 can include: based on the target detection unit, using the YOLOv5 algorithm to detect the personnel and objects in the video monitoring information at the current time, obtaining a target recognition result; the target recognition result includes the pixel area where the detected personnel is located, the pixel area where the object is located, and the object category; based on the posture extraction unit, key point feature information of the pixel area where the personnel is located is extracted, and the key point feature information is obtained; through the behavior detection unit, based on the key point feature information, the pixel area where the object is located, and the object category, the target personnel with abnormal behavior and the abnormal behavior content thereof are detected, and the target personnel with abnormal behavior and the abnormal behavior content thereof are obtained; through the matching unit, the coordinate information determined based on the pixel area where the target personnel is located is matched with the positioning information to determine the identity information of the target personnel; through the generation unit, the personnel behavior anomaly diagnosis result is generated based on the identity information of the target personnel and the abnormal behavior content of the target personnel.
[0054] In some embodiments, the implementation process of step 204 can include: based on the data preprocessing unit, the personnel behavior anomaly diagnosis results, the equipment state diagnosis results, the environmental information and the equipment operation parameter information at each time in the current period are aligned in time and space; based on the feature extraction unit, the data processed by the data preprocessing unit is subjected to time-frequency domain feature extraction, and the extracted feature information is obtained; based on the risk prediction unit, the feature information is input into a time sequence risk prediction model, and a safety risk prediction result output by the time sequence risk prediction model is obtained.
[0055] In some embodiments, the implementation process of step 203 can include: based on the first diagnosis unit, the equipment operation parameter information at the current time is compared with the corresponding threshold value, and a first diagnosis result is obtained; based on the second diagnosis unit, the equipment-related video monitoring data is extracted from the video monitoring data, and the equipment-related video monitoring data is subjected to abnormal identification, and a second diagnosis result is obtained; through the diagnosis result output unit, based on the first diagnosis result and the second diagnosis result, the equipment state diagnosis result is output.
[0056] As an example, the early warning notification can be sent to the wearable device of the relevant personnel.
[0057] As an example, the early warning notification can be in any of the following forms: alarm form, text form, light form, and voice broadcast form.
[0058] According to the mine safety risk prediction method, through the combination of diagnosis and prediction, the correlation analysis of the monitoring data of man-machine-ring is cooperated, the safety risk of the future mine can be accurately predicted, the accidents that have not occurred can be timely handled, and the safety of the underground operation of the mine is further ensured.
[0059] The device embodiments described above are merely illustrative, wherein the units described as separate components can or can not be physically separated, and the components displayed as units can or can not be physical units, i.e., can be located in one place, or can be distributed to multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment scheme according to actual needs. Those skilled in the art can understand and implement without creative labor.
[0060] Through the description of the above embodiments, those skilled in the art can clearly understand that the embodiments can be realized by means of software and the necessary general hardware platform, and of course can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments.
[0061] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.
Claims
1. A mine safety risk prediction system, characterized in that, include: The sensing module is used to collect real-time location information of personnel underground, environmental information, equipment operating parameter information, and video monitoring information underground; The personnel behavior anomaly diagnosis module is used to diagnose abnormal behavior based on the video surveillance information and location information at the current moment, and to obtain the personnel behavior anomaly diagnosis results at the current moment; The equipment status diagnosis module is used to perform status diagnosis based on the equipment operating parameter information at the current moment and obtain the equipment status diagnosis result at the current moment. The safety risk prediction module is used to predict risks based on the abnormal personnel behavior diagnosis results, equipment status diagnosis results, environmental information and equipment operating parameter information at various times within the current time period, and to obtain the safety risk prediction result for the current time period; the current time period includes the current time and multiple consecutive times preceding the current time. The emergency dispatch module is used to send early warning notifications and / or coordinate equipment control based on the current personnel anomaly diagnosis results, equipment status diagnosis results, and safety risk prediction results.
2. The system according to claim 1, characterized in that, The sensing module includes a positioning unit, an environmental monitoring unit, an equipment monitoring unit, and a video monitoring unit; wherein, The positioning unit is used to obtain the positioning information of underground personnel through a fixed base station deployed in the tunnel and wearable devices worn by underground personnel. The environmental monitoring unit is used to collect environmental information through sensors deployed underground; The equipment monitoring unit is used to collect the equipment operating parameter information; The video monitoring unit is used to collect the video monitoring information based on the explosion-proof camera deployed underground.
3. The system according to claim 1, characterized in that, The environmental information includes at least one of the following: temperature, smoke concentration, gas concentration, dust concentration, oxygen concentration, humidity, and air pressure.
4. The system according to claim 1, characterized in that, The abnormal personnel behavior diagnosis module includes a target detection unit, a posture extraction unit, a behavior detection unit, a matching unit, and a generation unit; wherein, The target detection unit is used to perform target detection on people and objects in the video surveillance information at the current moment using the YOLOv5 algorithm, and obtain the target recognition result; the target recognition result includes the pixel region where the detected person is located, the pixel region where the object is located, and the object category. The pose extraction unit is used to extract key point features from the pixel region where the person is located and obtain key point feature information. The behavior detection unit is used to perform behavior detection based on the key point feature information, the pixel region where the object is located, and the object category, and to obtain the target person with abnormal behavior and the content of the abnormal behavior. The matching unit is used to match the coordinate information determined based on the pixel region where the target person is located with the positioning information to determine the identity information of the target person. The generation unit is used to generate a diagnostic result of abnormal behavior of the target person based on the target person's identity information and the content of the target person's abnormal behavior.
5. The system according to claim 1, characterized in that, The security risk prediction module includes a data preprocessing unit, a feature extraction unit, and a risk prediction unit; wherein... The data preprocessing unit is used to perform time and space alignment processing on the abnormal personnel behavior diagnosis results, equipment status diagnosis results, environmental information and equipment operating parameter information at each moment in the current time period; The feature extraction unit is used to extract time-frequency domain features from the data processed by the data preprocessing unit to obtain the extracted feature information. The risk prediction unit allows the user to input the feature information into the time-series risk prediction model and obtain the security risk prediction results output by the time-series risk prediction model.
6. The system according to claim 1, characterized in that, The device status diagnostic module includes a first diagnostic unit, a second diagnostic unit, and a diagnostic result output unit; wherein... The first diagnostic unit is used to compare the current device operating parameter information with the corresponding threshold to obtain a first diagnostic result; The second diagnostic unit is used to extract device-related video surveillance data from video surveillance data, and to perform anomaly identification on the device-related video surveillance data to obtain a second diagnostic result. The diagnostic result output unit is used to output the device status diagnostic result based on the first diagnostic result and the second diagnostic result.
7. The system according to claim 1, characterized in that, The emergency dispatch module is used to send early warning notifications to the wearable devices of relevant personnel.
8. The system according to claim 1, characterized in that, The warning notification may take any of the following forms: alarm sound, text, light, or voice broadcast.
9. A method for predicting mine safety risks, characterized in that, The mine safety risk prediction system applied to any one of claims 1 to 8 comprises: The sensing module collects real-time location information of personnel underground, environmental information, equipment operating parameters, and video surveillance information from underground. The abnormal behavior diagnosis module diagnoses abnormal behavior based on the video surveillance information and location information at the current moment, and obtains the abnormal behavior diagnosis results at the current moment. The device status diagnosis module performs status diagnosis based on the device operating parameter information at the current moment and obtains the device status diagnosis result at the current moment. The safety risk prediction module performs risk prediction based on the abnormal personnel behavior diagnosis results, equipment status diagnosis results, environmental information, and equipment operating parameter information at various times within the current time period, and obtains the safety risk prediction result for the current time period; the current time period includes the current time and multiple consecutive times preceding the current time. Based on the current personnel anomaly diagnosis results, equipment status diagnosis results, and safety risk prediction results, the emergency dispatch module sends early warning notifications and / or equipment linkage control.
10. The method according to claim 9, characterized in that, The environmental information includes at least one of the following: temperature, smoke concentration, gas concentration, dust concentration, oxygen concentration, humidity, and air pressure.