Method and system for identifying numerical distortion of gas sensor in coal mining face of coal mine

Abstract

The present application provides a method and system for identifying the numerical distortion of gas sensors in coal mining working faces in coal mines. The method includes: real-time collection of daily measurement point values ​​of multiple gas sensors on the same coal mining working face; noise reduction processing of the collected daily measurement point values ​​of gas sensors; obtaining daily measurement point values ​​of gas sensors whose data volume after noise reduction processing is greater than a preset threshold; inputting the obtained daily measurement point values ​​of gas sensors into a pre-built artificial movement recognition model, and an artificial interruption or occlusion recognition model for abnormal recognition, and obtaining abnormal recognition results of gas sensors. The present application conducts a joint analysis of the monitoring data of multiple gas sensors in the same coal mining working face, and promptly discovers the numerical distortion of gas sensors in coal mining working faces caused by various types of reasons, and has good universality.

Description

Technical Field

[0001] The present application relates to the technical field of coal mine risk monitoring equipment, and in particular to a method and system for identifying numerical distortion of a gas sensor at a coal mining face in a coal mine. Background Art

[0002] In the process of coal mining, coal mining working faces in underground coal mines will emit gas. In order to ensure safe production, a coal mine production safety risk monitoring and early warning system has been established. The data of the underground safety monitoring system of coal mines is collected in real time and uploaded to the higher-level supervision and inspection department. Among the collected safety monitoring system data, the focus is on monitoring data such as underground gas and carbon monoxide, especially the gas monitoring data at the coal mining face.

[0003] At present, there are many studies on risk prediction based on coal mine gas monitoring data, but there are few studies on the distortion of monitoring data caused by illegal behaviors such as human movement of gas sensors (there are strict requirements for the installation location of each gas sensor on the working face), human interruption of monitoring data transmission, and human shielding and wrapping of sensors.

[0004] The existing coal mining face gas sensor numerical distortion identification has the following problems:

[0005] First, the research direction is single. There are some studies on the identification of numerical distortion of gas sensors in coal mining working faces. For example, in the prior art, it is proposed to identify whether the position of gas sensors in coal mining working faces in coal mines has been moved manually to avoid over-limit alarms. It is proposed to determine whether the trend of gas sensor value changes is consistent by checking whether the peaks and troughs of gas sensors at different positions in the same coal mining working face are within a certain period of time, or by calculating whether the correlation coefficient of the monitoring values ​​of upstream and downstream gas sensors in the same long period of time meets the threshold after a certain time difference. In this way, it is determined whether the position of the gas sensor has been moved manually. However, the above identification method cannot identify violations such as artificial obstruction of coal mining working face gas sensors and artificial interruption of monitoring value upload. The above method has low accuracy in detecting abnormalities of gas sensors and low accuracy in risk warning.

[0006] Second, the research is relatively poor in universality. In the existing technology, when judging the consistency of the trend of the value change of the gas sensor, either the ventilation type is obtained by relying on the coal mine map, or the distance and wind speed of the gas sensor in the coal mine are relied on to calculate the wind flow translation time. The scope of application of this method is relatively small.

[0007] At present, there is an urgent need to provide a method that has good universality and can identify the numerical distortion of gas sensors in coal mining faces caused by various types of reasons. Summary of the invention

[0008] The purpose of this application is to provide a method and system for identifying numerical distortion of gas sensors in coal mining faces in coal mines, to jointly analyze the monitoring data of multiple gas sensors in the same coal mining face, to promptly discover numerical distortion of gas sensors in coal mining faces caused by various different types of reasons, and to have good universality.

[0009] To achieve the above-mentioned purpose, the present application provides a method for identifying numerical distortion of gas sensors in coal mining working faces in coal mines, the method comprising:

[0010] Real-time collection of daily measurement point values ​​of multiple gas sensors on the same coal mining face;

[0011] Perform noise reduction on the daily measurement point values ​​collected from the gas sensor;

[0012] Obtain daily measurement point values ​​of gas sensors whose data volume after noise reduction is greater than a preset threshold;

[0013] The daily measurement point values ​​of the gas sensor are input into the pre-built human movement recognition model and the human interruption or occlusion recognition model for abnormality recognition to obtain the abnormality recognition results of the gas sensor.

[0014] As above, after the obtained daily measurement point value of the gas sensor is input into the pre-built artificial movement recognition model, the artificial movement recognition model performs abnormal recognition on the gas sensor;

[0015] Among them, the method of using the artificial movement recognition model to identify abnormalities of the gas sensor includes:

[0016] Obtain the daily measurement point values ​​of the gas sensors after noise reduction treatment in the same coal mining face;

[0017] Calculate the evaluation index data of the gas sensor according to the daily measurement point value of the gas sensor, and the evaluation index data includes the mean, standard deviation, mean difference and standard deviation difference;

[0018] According to the obtained evaluation index data, it is determined whether the evaluation index data meets the preset human movement determination conditions. If so, it is determined that the gas sensor has human movement abnormality, otherwise, there is no human movement abnormality.

[0019] As mentioned above, the preset artificial movement determination condition includes two determination steps:

[0020] The first judgment step: If the mean difference , and the standard deviation , it is determined that the gas sensor of the coal mining face has been moved manually, otherwise, the second determination step is executed;

[0021] The second judgment step: If the standard deviation of the daily measurement point value of the T0 gas sensor , mean difference , , it is determined that the gas sensor of the coal mining face has been moved manually, otherwise, there is no human movement; among them, It represents the mean value of the daily measurement point value of T0 gas sensor; It represents the mean value of the daily measurement point values ​​of T1 gas sensor; It represents the average of the daily measurement point values ​​of T2 gas sensor.

[0022] As above, the mean difference is calculated as follows:

[0023] ;

[0024] in, represents the mean difference; Indicates taking , and The maximum value in ;

[0025] Indicates taking , and The minimum value in .

[0026] As above, the standard deviation difference is calculated as:

[0027] ;

[0028] in, represents the standard deviation; Indicates taking , and The maximum value in ; Indicates taking , and The minimum value in .

[0029] As above, after the acquired daily measurement point value of the gas sensor is input into the pre-built artificial interruption or occlusion recognition model, the artificial interruption or occlusion recognition model performs abnormal recognition on the gas sensor;

[0030] Methods for artificially interrupting or blocking the recognition model to identify abnormalities of the gas sensor include:

[0031] Obtain the daily measurement point values ​​of the gas sensors after noise reduction treatment in the same coal mining face;

[0032] Determine whether the daily measurement point value of the gas sensor meets the artificial occlusion condition or the artificial interruption condition. If it meets the artificial occlusion condition, output the artificial occlusion exception. If it meets the artificial interruption condition, output the artificial interruption exception. Otherwise, do not output the exception.

[0033] As above, the method for determining whether the daily measurement point value of the gas sensor meets the artificial shielding condition includes:

[0034] The first occlusion determination method:

[0035] According to the daily measurement point values ​​of the gas sensor collected in a day, the window time period average of the daily measurement point value of a single gas sensor in a preset time period in a day is calculated and window period standard deviation , and the non-window period average of the daily measurement point values ​​of the gas sensor in other time periods and non-window period standard deviation;

[0036] If the window time period average of a single gas sensor ,and ,

[0037] and ; * represents multiplication; then it is determined that the gas sensor is artificially blocked, otherwise, the second blockage determination method is executed.

[0038] As above, the second occlusion determination method: based on the daily measurement point value of a single gas sensor, generate a curve of its change over time, and judge whether the gas sensor is artificially obstructed based on the curve and the preset gas sensor daily measurement point value threshold.

[0039] As above, the method for determining whether the daily measurement point value of the gas sensor meets the conditions of human interruption includes: for a single gas sensor, if the interruption time of the monitoring value of the gas sensor exceeds the preset time, and the value before the interruption is greater than the interruption threshold, and the value after the interruption is less than the interruption threshold, then it is determined that there is an artificial interruption in the gas sensor, otherwise, there is no artificial interruption.

[0040] The present application also provides a coal mine coal mining face gas sensor numerical distortion identification system, the system comprising:

[0041] The data acquisition module is used to collect the daily measurement point values ​​of multiple gas sensors on the same coal mining face in real time;

[0042] A data processor, used for performing noise reduction processing on the daily measurement point values ​​collected by the gas sensor;

[0043] A data acquisition module, used to obtain daily measurement point values ​​of gas sensors whose data volume after noise reduction processing is greater than a preset threshold;

[0044] The anomaly recognition model is used to input the daily measurement point values ​​of the gas sensor into a pre-built human movement recognition model and a human interruption or occlusion recognition model to perform anomaly recognition and obtain the anomaly recognition results of the gas sensor.

[0045] The beneficial effects achieved by this application are as follows:

[0046] (1) This application analyzes the frequency, amplitude and trend of the value changes of gas sensors at different positions in the coal mining face of a coal mine, and conducts a joint analysis of the value changes of gas sensors at multiple positions in the same coal mining face. This can timely discover the distortion of gas sensors at the coal mining face of a coal mine caused by human intervention, providing an important means to support coal mine supervision and inspection.

[0047] (2) The present application can not only identify the numerical distortion of the gas sensor caused by human movement, but also identify the numerical distortion of the gas sensor caused by human interruption or obstruction, thereby timely discovering the numerical distortion of the gas sensor in the coal mining face caused by various types of reasons, and improving the accuracy of identifying the numerical distortion of the gas sensor. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For those skilled in the art, other drawings can also be obtained based on these drawings.

[0049] Figure 1 This is a flow chart of a method for determining a coal mining working face according to an embodiment of the present application.

[0050] Figure 2 This is a U-shaped coal mining working face in an embodiment of the present application.

[0051] Figure 3 This is a flow chart of a method for identifying numerical distortion of a gas sensor in a U-shaped coal mining face according to an embodiment of the present application.

[0052] Figure 4 This is a flow chart of a method for determining whether a gas sensor has abnormal movement according to an embodiment of the present application.

[0053] Figure 5 This is a flow chart of a method for identifying abnormalities in a gas sensor by artificially interrupting or blocking the recognition model according to an embodiment of the present application.

[0054] Figure 6 This is a flow chart of a method for determining whether a gas sensor has been manually moved according to an embodiment of the present application.

[0055] Figure 7This is a flow chart of a method for determining artificial interruption and artificial obstruction of a gas sensor according to an embodiment of the present application.

[0056] Figure 8 This is a structural schematic diagram of a numerical distortion identification system for a coal mine coal mining face gas sensor according to an embodiment of the present application.

[0057] Fig. 9 This is a schematic diagram of a curve showing the daily measurement point values ​​of the gas sensor according to an embodiment of the present application changing with time.

[0058] Figure numerals: 10 - data acquisition module, 20 - data processor; 30 - data acquisition module; 40 - abnormality identification module; 100 - gas sensor numerical distortion identification system. DETAILED DESCRIPTION

[0059] The following is a clear and complete description of the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application. Embodiment 1

[0060] The present application provides a method for identifying numerical distortion of a gas sensor in a coal mining face of a coal mine, the method comprising the following steps:

[0061] like Figure 1 As shown, the ventilation type of the coal mining face is judged:

[0062] Step S1, obtaining coal mine information and its gas sensor information.

[0063] Among them, coal mine information includes mine status (normal production or abnormal production) and coal mine code of normal construction. Gas sensor information includes gas sensor code and its location information.

[0064] Step S2, obtaining the location information of the gas sensor in the coal mine in normal production according to the coal mine information and its gas sensor information.

[0065] The location information of the gas sensor includes T0, T1 or T2, and the gas sensors without excavation information (filtered by blacklist) are retained.

[0066] Step S3, identifying the coal mining face according to the coal mine code in the coal mine information or the location information of the gas sensor.

[0067] Specifically, for gas sensors in the same coal mine, the working face number of the gas sensor is obtained and combined with the coal mine code to uniquely identify a coal mining working face of the coal mine. According to the correspondence between the location information of the gas sensor and the working face, the working face number of the gas sensor with known location information is obtained. If the working face number cannot be obtained, the first two characters of the gas sensor location information are used to replace the coal mining working face of the gas sensor.

[0068] Step S4, obtaining all gas sensors contained in the same coal mining face.

[0069] Step S5, judging the ventilation type of the coal mining face according to the number of gas sensors included in the same coal mining face, the number of gas sensors and preset judgment conditions.

[0070] Specifically, if the same coal mining working face contains and only contains one T0 gas sensor, one T1 gas sensor and one T2 gas sensor (determination condition of U-shaped ventilation coal mining working face), then the coal mining working face is determined to be a U-shaped ventilation coal mining working face.

[0071] like Figure 2 The figure shows a U-shaped ventilated coal mining face, in which a T0 gas sensor without excavation information, a T1 gas sensor and a T2 gas sensor are located in the same coal mining face.

[0072] U-type ventilation is the most commonly used ventilation method for coal mining faces in coal mines. When judging coal mining faces of other ventilation types, it is sufficient to change the judgment conditions of the name and number of the gas sensors of the face. This application can not only identify coal mining faces of U-type ventilation type, but also identify other coal mining faces other than U-type ventilation type. It has a wide range of applications and good universality. After judging the ventilation type of the coal mining face, save the ventilation type information and gas sensor number of the coal mining face.

[0073] According to the ventilation type of the coal mining face, the corresponding gas sensor numerical distortion identification steps are performed.

[0074] like Figure 3 As shown, if the ventilation type of the coal mining face is U-shaped, the method for performing the gas sensor numerical distortion identification includes the following steps:

[0075] Step T1, real-time collection of daily measurement point values ​​of multiple gas sensors on the same coal mining face.

[0076] The daily measurement point value refers to the set of 24-hour gas sensor measurement point values. Preferably, within 24 hours a day, a piece of monitoring value data of each gas sensor is obtained every 30 seconds, and all the obtained monitoring value data constitute the daily measurement point value.

[0077] Among them, a coal mining face of U-type ventilation type includes three gas sensors, namely the first gas sensor (T0 gas sensor), the second gas sensor (T1 gas sensor) and the third gas sensor (T2 gas sensor). Among them, the first gas sensor is located in the upper corner. Under normal circumstances, gas first accumulates at the T0 gas sensor in the upper corner, and then gradually diffuses to the T1 gas sensor and T2 gas sensor. The change trends of the T0 gas sensor, T1 gas sensor and T2 gas sensor are consistent, and the concentration at the T0 gas sensor reaches the highest first, and then diffuses to the T1 gas sensor and T2 gas sensor, and the concentration gradually decreases. Specifically, the daily measurement point value of the gas sensor is the gas concentration data detected by the gas sensor.

[0078] Step T2, performing noise reduction processing on the daily measurement point values ​​collected by the gas sensor.

[0079] The noise reduction method includes: removing negative values, sorting the values ​​(daily measurement point values ​​of the gas sensor), and then removing 5% of the head and tail data.

[0080] Specifically, the monitoring values ​​(measurement point values) of each gas sensor within a day are processed to remove negative values, then sorted from small to large, and 5% of the data at the head and tail after sorting are removed.

[0081] Step T3, obtaining the daily measurement point value of the gas sensor whose data volume after noise reduction processing is greater than a preset threshold.

[0082] For example, the preset threshold is 50, and it is determined whether the data volume of the daily measurement point values ​​of the gas sensor after noise reduction processing is greater than 50. If so, the daily measurement point values ​​of the gas sensor will be obtained.

[0083] Step T4, input the obtained daily measurement point value of the gas sensor into the pre-built human movement recognition model and the human interruption or occlusion recognition model to perform abnormality recognition, and obtain the abnormality recognition result of the gas sensor.

[0084] Specifically, after the acquired daily measurement point values ​​of the gas sensor are input into a pre-built human motion recognition model, the human motion recognition model performs abnormality recognition on the gas sensor.

[0085] like Figure 4 and Figure 6 As shown, the method for identifying abnormalities of a gas sensor using a human movement recognition model includes:

[0086] Step T410, obtaining the daily measurement point value of the gas sensor after noise reduction processing in the same coal mining face.

[0087] Step T420: Calculate the evaluation index data of the gas sensor according to the daily measurement point value of the gas sensor. The evaluation index data includes the mean, standard deviation, mean difference and standard deviation difference.

[0088] Specifically, the mean refers to the mean of the daily measurement point values ​​of each gas sensor; the standard deviation refers to the standard deviation of the daily measurement point values ​​of each gas sensor; the mean difference refers to the mean difference of all gas sensors; and the standard deviation difference refers to the standard deviation difference of all gas sensors.

[0089] The mean value is calculated by dividing the sum of all data in a set of data by the number of data in the set. Specifically, the mean value of a gas sensor's daily measurement point value is calculated as follows:

[0090] ;

[0091] in, represents the mean, Indicates the total number of measurement point values ​​of the gas sensor after noise reduction processing; Indicates the gas sensor after noise reduction processing. The value of a measuring point.

[0092] The calculation method of the standard deviation of the daily measurement point value of a gas sensor is:

[0093] ;

[0094] Where S represents the standard deviation of the daily measurement point values ​​of the gas sensor; represents the mean, Indicates the total number of measurement point values ​​of the gas sensor after noise reduction processing; Indicates the gas sensor after noise reduction processing. The standard deviation can reflect the degree of dispersion of a data set and is most commonly used as a measure of the degree of statistical distribution in probability statistics. Two sets of data with the same average (mean) may not have the same standard deviation.

[0095] The mean difference is calculated as follows:

[0096] ;

[0097] in, represents the mean difference; Indicates taking , and The maximum value in ;

[0098] Indicates taking , and the minimum value in; represents the mean value of the daily measurement points of the T0 gas sensor; represents the mean value of the daily measurement points of the T1 gas sensor; represents the mean value of the daily measurement points of the T2 gas sensor;

[0099] Among them, the calculation method of the standard deviation difference is:

[0100] ;

[0101] Among them, represents the standard deviation difference; represents taking , and the maximum value in; represents taking , and the minimum value in; represents the standard deviation of the daily measurement points of the T0 gas sensor; represents the standard deviation of the daily measurement points of the T1 gas sensor; represents the standard deviation of the daily measurement points of the T2 gas sensor.

[0102] Step T430, according to the obtained evaluation index data, judge whether the evaluation index data meets the preset artificial movement judgment condition. If so, it is determined that there is an artificial movement anomaly in the gas sensor. Otherwise, there is no artificial movement anomaly.

[0103] As Figure 6 shown, the preset artificial movement judgment condition includes two judgment steps:

[0104] The first judgment step: If the mean difference , and the standard deviation difference , it is considered that the change trend of the measurement points in the coal mining face is inconsistent, and the judgment result is that there is an artificial movement of the gas sensor in the coal mining face. Otherwise, the second judgment step is executed.

[0105] The second judgment step: If the standard deviation of the daily measurement points of the T0 gas sensor , the mean difference and , or T0 mean < T1 mean or T0 mean < T2 mean; it is considered that the monitoring value (daily measurement point value) of the gas sensor in the coal mining face does not conform to the data change that should be in the correct installation position, and it is determined that there is an artificial movement of the position of the gas sensor in the coal mining face. Otherwise, there is no artificial movement; among them, represents the mean value of the daily measurement points of the T0 gas sensor; It represents the mean value of the daily measurement point values ​​of T1 gas sensor; It represents the average of the daily measurement point values ​​of T2 gas sensor.

[0106] As a specific embodiment of the present invention, after determining whether the gas sensor is abnormally moved by human beings, it is identified whether the gas sensor is abnormally interrupted or blocked by human beings.

[0107] like Figure 5 and Figure 7 As shown, the method of artificially interrupting or blocking the recognition model to identify abnormalities of the gas sensor includes:

[0108] Step T510, obtaining the daily measurement point value of the gas sensor after noise reduction processing in the same coal mining working face.

[0109] Step T520, determine whether the daily measurement point value of the gas sensor meets the artificial occlusion condition and / or the artificial interruption condition. If it meets the artificial occlusion condition, output the artificial occlusion exception. If it meets the artificial interruption condition, output the artificial interruption exception. Otherwise, do not output the exception.

[0110] Among them, the method for judging whether the daily measurement point value of the gas sensor meets the artificial shielding condition includes:

[0111] The first occlusion determination method:

[0112] Step T521, based on the daily measurement point values ​​of the gas sensor collected in one day, calculate the window time period average of the daily measurement point values ​​of a single gas sensor in a preset time period (window time period, such as 8:00-11:00) in one day and window period standard deviation , and the non-window period average of the daily measurement point values ​​of the gas sensor in other time periods The calculation method of the window period mean and window period standard deviation is the same as the calculation method of the mean and standard deviation described above.

[0113] Step T522: If the window time period average of a single gas sensor is ,and ,

[0114] And the standard deviation of the window period ; * indicates multiplication; then it is determined that the gas sensor is blocked by human factors, otherwise, the second blocking determination method is executed. Among them, the determination parameters: 0.05, 0.02 and 0.2 can be adjusted according to experience.

[0115] The second occlusion determination method:

[0116] Step T523, based on the daily measurement point value of a single gas sensor, generate a curve of its change over time, and determine whether the gas sensor is artificially blocked based on the curve and the preset daily measurement point value threshold of the gas sensor.

[0117] like Fig. 9 As shown, this is a curve of the daily measurement point value of the gas sensor changing with time, the horizontal axis is time, and the vertical axis is the daily measurement point value R of the gas sensor.

[0118] Specifically, the preset daily measurement point value threshold of the gas sensor includes an upper limit value R2 and a lower limit value R1. If a single gas sensor has a daily measurement point value that fluctuates from a measurement point value G1 above the upper limit value to a measurement point value G2 below the lower limit value, and continues to rise to a measurement point value G3 above the lower limit value after a period of time, and the standard deviation of the measurement point value in the period below the lower limit value (T3-T5) is <0.02, and (the mean value of the period after the lower limit value rises - the mean value of the period below the lower limit value)>0.1, then it is determined that the gas sensor is artificially blocked. Among them, the calculation method of the standard deviation and variance adopts the existing calculation method.

[0119] As another specific embodiment of the present invention, by analyzing data that suddenly rises and then suddenly drops, and then rises back to a normal range after a period of time, it is determined whether there is a human interruption.

[0120] Among them, the method for judging whether the daily measurement point value of the gas sensor meets the artificial interruption condition includes:

[0121] Step T524, for a single gas sensor, if the interruption time of the monitoring value of the gas sensor exceeds the preset time, and the value before the interruption is greater than the interruption threshold, and the value after the interruption is less than the interruption threshold, it is determined that the gas sensor has an artificial interruption, otherwise, there is no artificial interruption.

[0122] Specifically, for a single gas sensor, if the monitoring value interruption time exceeds the preset time (for example: 10 minutes), and the value before the interruption is greater than the interruption threshold, and the value after the interruption is less than the interruption threshold (that is, the value before the interruption is higher and the value after the interruption is lower), then it is determined that the gas sensor has been artificially interrupted.

[0123] As a specific embodiment of the present invention, if the monitoring value (daily measurement point value) of the gas sensor is above 0.5, and a data interruption occurs suddenly, and the value interruption time interval is greater than 10 minutes, the monitoring value before the interruption is above 0.5, and the monitoring value measured after the interruption is less than 0.5, then it is determined that the gas sensor has been artificially interrupted. It can be understood that after artificially processing the gas sensor during the interruption time, the value before the interruption is greater than the interruption threshold, and the value after the interruption measured after the data is restored is less than the interruption threshold.

[0124] As a specific embodiment of the present invention, the threshold of the artificial interruption or occlusion identification model (for example: the upper limit value R2 and the lower limit value R1 of the daily measurement point value of the gas sensor) or various parameters used for condition judgment are predetermined, and when applied to other ventilation types of coal mining faces, appropriate adjustments are made accordingly.

[0125] After the abnormal results are stored in the database, data support is provided for remote monitoring.

[0126] Embodiment 2

[0127] like Figure 8 As shown, the present application also provides a coal mine coal mining face gas sensor numerical distortion identification system 100, the system comprising:

[0128] The data acquisition module 10 is used to collect the daily measurement point values ​​of multiple gas sensors on the same coal mining face in real time;

[0129] A data processor 20 is used to perform noise reduction processing on the daily measurement point values ​​collected by the gas sensor;

[0130] The data acquisition module 30 is used to obtain the daily measurement point value of the gas sensor whose data volume after noise reduction processing is greater than a preset threshold;

[0131] The abnormality recognition model 40 is used to input the acquired daily measurement point values ​​of the gas sensor into a pre-built human movement recognition model and a human interruption or occlusion recognition model to perform abnormality recognition and obtain the abnormality recognition result of the gas sensor.

[0132] The beneficial effects achieved by this application are as follows:

[0133] (1) This application analyzes the frequency, amplitude and trend of the value changes of gas sensors at different positions in the coal mining face of a coal mine, and conducts a joint analysis of the value changes of gas sensors at multiple positions in the same coal mining face. This can timely discover the distortion of gas sensors at the coal mining face of a coal mine caused by human intervention, providing an important means to support coal mine supervision and inspection.

[0134] (2) The present application can not only identify the numerical distortion of the gas sensor caused by human movement, but also identify the numerical distortion of the gas sensor caused by human interruption or obstruction, thereby timely discovering the numerical distortion of the gas sensor in the coal mining face caused by various types of reasons, and improving the accuracy of identifying the numerical distortion of the gas sensor.

[0135] The above description is only an embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A method for identifying numerical distortion of gas sensors in coal mining working faces in coal mines, characterized in that: The method includes: Real-time collection of daily measurement point values ​​of multiple gas sensors on the same coal mining face; Perform noise reduction on the daily measurement point values ​​collected from the gas sensor; Obtain daily measurement point values ​​of gas sensors whose data volume after noise reduction is greater than a preset threshold; The obtained daily measurement point values ​​of the gas sensor are input into the pre-built human movement recognition model and the human interruption or occlusion recognition model to perform abnormality recognition and obtain the abnormality recognition result of the gas sensor; After the daily measurement point values ​​of the gas sensor are input into the pre-built artificial movement recognition model, the artificial movement recognition model recognizes the abnormality of the gas sensor; Among them, the method of using the artificial movement recognition model to identify abnormalities of the gas sensor includes: Obtain the daily measurement point values ​​of the gas sensors after noise reduction treatment in the same coal mining face; Calculate the evaluation index data of the gas sensor according to the daily measurement point value of the gas sensor, and the evaluation index data includes the mean, standard deviation, mean difference and standard deviation difference; According to the acquired evaluation index data, it is determined whether the evaluation index data meets the preset human movement determination condition. If so, it is determined that the gas sensor has human movement abnormality, otherwise, there is no human movement abnormality; The preset artificial movement determination condition includes two determination steps: The first judgment step: If the mean difference D avg >0.06, and the standard deviation difference D s >0.03, it is determined that the gas sensor of the coal mining face has been moved manually, otherwise, the second determination step is executed; The second judgment step: If the standard deviation S of the daily measurement point value of the T0 gas sensor T0 ≥0.3, mean difference D avg >0.1, It is determined that the gas sensor of the coal mining face has been moved manually, otherwise, it has not been moved manually; It represents the mean value of the daily measurement point value of T0 gas sensor; It represents the mean value of the daily measurement point values ​​of T1 gas sensor; It represents the mean value of the daily measurement point values ​​of T2 gas sensor; The mean difference is calculated as follows: Among them, D avg represents the mean difference; Indicates taking and The maximum value in ; Indicates taking and The minimum value in ; The standard deviation difference is calculated as: D S =max(S T0 ,S T1 ,S T2 )-min(S T0 ,S T1 ,S T2 ); Among them, D s Indicates the standard deviation; max(S T0 , S T1 , S T2 ) means taking S T0 , S T1 and S T2 The maximum value in min(S T0 , S T1 , S T2 ) means taking S T0 , S T1 and S T2 The minimum value among them, S T0 Indicates the standard deviation of the daily measurement point value of T0 gas sensor; S T1 Indicates the standard deviation of the daily measurement point value of T1 gas sensor; S T2 Indicates the standard deviation of the daily measurement point values ​​of the T2 gas sensor.

2. The method for identifying numerical distortion of gas sensors in coal mining working faces in coal mines according to claim 1 is characterized in that: After the acquired daily measurement point values ​​of the gas sensor are input into a pre-built artificial interruption or occlusion recognition model, the artificial interruption or occlusion recognition model performs abnormality recognition on the gas sensor; Methods for artificially interrupting or blocking the recognition model to identify abnormalities of the gas sensor include: Obtain the daily measurement point values ​​of the gas sensors after noise reduction treatment in the same coal mining face; Determine whether the daily measurement point value of the gas sensor meets the artificial occlusion condition and / or artificial interruption condition. If it meets the artificial occlusion condition, output the artificial occlusion exception. If it meets the artificial interruption condition, output the artificial interruption exception. Otherwise, do not output the exception.

3. The method for identifying numerical distortion of gas sensors in coal mining working faces of coal mines according to claim 2 is characterized in that: Methods for determining whether the daily measurement point value of the gas sensor meets the artificial shielding condition include: The first occlusion determination method: According to the daily measurement point values ​​of the gas sensor collected in a day, the window time period average of the daily measurement point value of a single gas sensor in a preset time period in a day is calculated and the standard deviation of the window period S window , and the non-window period average of the daily measurement point values ​​of the gas sensor in other time periods and non-window period standard deviation; If the window time period average of a single gas sensor and And Swindow<0.02; * represents multiplication; then it is determined that the gas sensor is artificially blocked, otherwise, the second blocking determination method is executed.

4. The method for identifying numerical distortion of gas sensors in coal mining working faces in a coal mine according to claim 3 is characterized in that: The second occlusion determination method: based on the daily measurement point value of a single gas sensor, generate a curve of its change over time, and determine whether the gas sensor is artificially blocked based on the curve and the preset gas sensor daily measurement point value threshold.

5. The method for identifying numerical distortion of gas sensors in coal mining working faces in a coal mine according to claim 3, characterized in that: The method for determining whether the daily measurement point value of a gas sensor meets the conditions for an artificial interruption includes: for a single gas sensor, if the interruption time of the monitoring value of the gas sensor exceeds the preset time, and the value before the interruption is greater than the interruption threshold, and the value after the interruption is less than the interruption threshold, then it is determined that the gas sensor has an artificial interruption; otherwise, there is no artificial interruption.

6. A coal mine coal mining face gas sensor numerical distortion identification system, characterized in that: The method according to any one of claims 1 to 5 is implemented, the system comprising: The data acquisition module is used to collect the daily measurement point values ​​of multiple gas sensors on the same coal mining face in real time; A data processor, used for performing noise reduction processing on the daily measurement point values ​​collected by the gas sensor; A data acquisition module, used to obtain daily measurement point values ​​of gas sensors whose data volume after noise reduction processing is greater than a preset threshold; The anomaly recognition model is used to input the daily measurement point values ​​of the gas sensor into a pre-built human movement recognition model and a human interruption or occlusion recognition model to perform anomaly recognition and obtain the anomaly recognition results of the gas sensor.

Images