Periodic pressure identification method based on hydraulic support pressure data in fully mechanized mining working face

Abstract

The present invention relates to a method for identifying periodic pressure based on the pressure data of the hydraulic support of the fully-mechanized mining working face, and belongs to the field of identifying periodic pressure on the fully-mechanized mining working face. The method can identify periodic pressure based on the offline pressure data and real-time pressure data of the hydraulic support of the fully-mechanized mining working face. The periodic pressure identification method of the method includes: collecting the pressure monitoring data of each hydraulic support of the fully-mechanized mining working face; pre-processing the data to form a data set; setting a pressure judgment threshold, identifying periodic pressure according to the data set, and counting the number of pressures, pressure dates and pressure duration. The present invention can cope with multi-factor interference and improve the accuracy of identifying periodic pressure. In addition, constructing a periodic pressure prediction and control system based on the present invention can provide a more solid safety guarantee for coal mine production and promote the efficient and sustainable conduct of production activities.

Description

Technical Field

[0001] The present invention belongs to the field of comprehensive mining face pressure identification, and relates to a periodic pressure identification method based on the pressure data of a hydraulic support of a comprehensive mining working face. Background Art

[0002] Cyclic pressure, an inherent mechanical response of the roof strata during coal mining, exhibits periodic fracture and collapse characteristics as the working face continues to advance. This phenomenon not only poses a severe challenge to the underground support system but also, by imposing significant impact loads, poses the risk of inducing safety accidents such as roof collapse, directly threatening the safety and continuity of coal mine operations and the lives of workers. Currently, identification of cyclic pressure relies on the experience of frontline workers. For example, a 200-300 meter long working face is generally assumed to experience pressure around 20-30 meters of excavation. However, cyclic pressure exhibits non-uniform periodicity depending on the working face, geological structure, and production intensity. Therefore, accurately identifying and effectively managing cyclic pressure has become a core issue in coal mine safety management systems.

[0003] Modern coal mines have widely adopted electric-hydraulic-controlled hydraulic support systems, incorporating a high-tech platform called a column pressure monitoring sensor network. This network enables real-time, continuous monitoring of the pressure status of the support columns, providing a solid data foundation for the study and analysis of cyclical pressure events. Traditional mine pressure analysis methods attempt to use parameters such as the final resistance and time-weighted working resistance of a single or a few hydraulic supports as a basis for determining roof pressure, or to aggregate and analyze support pressure data across the entire working face within a specific time window. However, these methods have limitations and fail to comprehensively and objectively reflect the overall pressure distribution and dynamics of the working face. In particular, the asynchronous nature of cyclical pressure events—in which a few supports initially respond, then gradually spread to the majority, and finally return to a state where only a few supports bear the pressure—further exacerbates the difficulty of accurately determining the timing of roof pressure events.

[0004] When the working face encounters roof pressure, hydraulic supports, particularly in the central area, experience significant pressure increases, prompting frequent activation of safety valves to relieve overload pressure. While using this characteristic as an indicator for pressure on the face is effective, it faces multiple complexities: uneven force distribution across supports leads to varying safety valve opening times and frequencies; during normal production, localized pressure surges can also trigger the safety valves of individual supports; and during both pressure and normal production, adjacent supports with activated safety valves may either also open their own valves or experience pressure increases that do not reach the threshold, creating a chain reaction.

[0005] Therefore, there is an urgent need to develop a comprehensive method that can accurately identify periodic pressure, so as to effectively deal with multi-factor interference and improve recognition accuracy. Summary of the Invention

[0006] In view of this, the purpose of the present invention is to provide a method for identifying periodic pressure based on the pressure data of hydraulic supports in fully-mechanized mining working faces, effectively deal with multi-factor interference, improve recognition accuracy, and at the same time construct a periodic pressure prediction and control system to provide a more solid safety guarantee for coal mine production and promote efficient and sustainable production activities.

[0007] In order to achieve the above object, the present invention provides the following technical solutions:

[0008] A method for identifying periodic pressure based on the pressure data of the hydraulic support of a fully mechanized mining working face can be used to identify periodic pressure based on the offline pressure data and real-time pressure data of the hydraulic support of the fully mechanized mining working face.

[0009] Among them, for the offline pressure data of the hydraulic support of the fully mechanized mining working face, the periodic pressure identification method of this method is as follows:

[0010] S1. Collect offline pressure monitoring data of each hydraulic support in the fully mechanized mining working face to form a data set;

[0011] S2. Setting a time interval, and resampling the formed data set at the time interval;

[0012] S3. During the resampling process, if there are multiple original data points within the time interval, an average value is calculated; if there are no original data points within the time interval, an interpolation method is performed to fill in the empty values, thereby obtaining a resampled data set;

[0013] S4. Identify periodic pressure according to the resampled data set, and count the number of pressures, pressure dates, and pressure duration.

[0014] Furthermore, in step S3, the average value is calculated as:

[0015]

[0016] Where T represents the set time interval, N represents the number of original data points within the time interval T, and D(t i ) represents the original data point, t i <t j <t i +T; the calculated average value is taken as time point t i The sampling data of .

[0017] Furthermore, in step S3, the method of determining whether the original data point is missing in the time interval includes: if the original data point D(t m ) at time t m exists, then at the time point t after resampling i Make a judgment at this point. If t i If the time does not match the original data point, interpolation is used to fill in the empty values.

[0018] Furthermore, in step S4, the method of performing periodic pressure identification according to the resampled data set includes: setting an opening threshold Thre of the hydraulic support safety valve of the fully mechanized mining working face;

[0019] Counting the number of stents whose pressure value P at each recording moment in the resampled data set satisfies 42≤P<Thre, recorded as count1; counting the number of stents whose pressure value P at each recording moment in the resampled data set satisfies Thre≤P, recorded as count2;

[0020] Sum count1 and count2 at each moment to get the total at each moment count , and find the count2 in total at each moment count The proportion of open rate ;

[0021] When open rate >0.2 and total count When ≥0.15*M, add the current time to the time data set T to be judged D In which M represents the total number of hydraulic supports;

[0022] Time dataset T to be judged D Perform statistics on every day, and meet the conditions open rate >0.2 and total count The number of occurrences of moments ≥0.15*M and the duration of pressure on each day; days with less than 10 occurrences and duration less than 60 minutes are excluded.

[0023] The number of pressure calls is counted based on the periodic pressure call identification result, wherein consecutive dates are regarded as one pressure call, and the final number of pressure calls, pressure call date and pressure call duration are obtained.

[0024] For the real-time pressure data between the hydraulic pressure of the fully mechanized mining working face, the method for periodic pressure identification includes:

[0025] S1, real-time collection of pressure data between each hydraulic pressure in the fully mechanized mining working face through pressure sensors;

[0026] S2. Set a time window, store the real-time collected pressure data within the time window, and aggregate the stored data when the data within the time window reaches the accumulation upper limit;

[0027] S3. Set a pressure judgment threshold, and for the aggregated data in each time window, determine whether the data is in a pressure state in a certain time window according to the pressure judgment threshold.

[0028] The cumulative upper limit in step S2 is the length of time set in the time window, or the amount of data stored in the time window.

[0029] In step S3, the method for determining whether a certain time window is in a pressure incoming state includes:

[0030] According to the pre-set pressure segments 42≤P<Thre and Thre≤P, the number of hydraulic supports that meet the conditions of each pressure segment within the time window is recorded; the number of hydraulic supports with pressure values ​​between 42≤P<Thre is recorded as count1, and the number of hydraulic supports with pressure values ​​between Thre≤P is recorded as count2; where P represents the pressure value of the hydraulic support, and Thre represents the opening threshold of the hydraulic support safety valve;

[0031] Calculate total based on the statistical results within the time window count = count1 + count2, and open rate =count 2 total count ;

[0032] When open rate >0.2 and total count When ≥0.15*M, the current time window is judged to be in a pressure state, and a pressure event is generated for recording.

[0033] The beneficial effects of the present invention are as follows: the present invention can cope with multi-factor interference, improve the recognition accuracy of periodic pressure, and has a simple principle and is easy to implement. It can complete the processing of a large amount of data in a short period of time and achieve rapid response. Moreover, the present invention makes full use of existing monitoring equipment and data acquisition systems, does not require additional hardware investment, and greatly reduces the implementation cost. In addition, the construction of a periodic pressure prediction and control system based on the present invention can help coal mining enterprises better control maintenance costs, provide more solid safety guarantees for coal mine production, and promote efficient and sustainable production activities.

[0034] Other advantages, objects, and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art upon examination of the following description or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below with reference to the accompanying drawings, in which:

[0036] Figure 1 Schematic diagram of the process of the present invention;

[0037] Figure 2 Resampling process for the dataset. DETAILED DESCRIPTION

[0038] The following describes the embodiments of the present invention by means of specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0039] See also Figure 1 , is a method for identifying periodic pressure inflow based on the pressure data of hydraulic supports in the fully mechanized mining working face. The method mainly consists of first collecting the pressure monitoring data of each hydraulic support in the fully mechanized mining working face to form a data set containing time and pressure values; resampling the data set to unify the time interval, and using average value calculation and linear interpolation to fill the missing values; applying the periodic pressure inflow identification algorithm, based on the safety valve opening ratio and duration, combined with statistical analysis methods, to identify and count the pressure inflow events.

[0040] Specifically, the method is as follows:

[0041] 1. Data Collection

[0042] The pressure monitoring data of each hydraulic support in the fully mechanized mining working face is collected to form a data set D, which contains the collection time and pressure value of each support pressure data.

[0043] 2. Resampling the dataset

[0044] Due to the complex production environment, the stent pressure data collection process may suffer from data loss or inconsistent data collection frequency. Therefore, after collecting the dataset D, it is first converted into a regularly spaced time series, where the interval t is customizable and is typically set to 30 seconds or 1 minute. In this example, the resampling frequency is set to 1 minute.

[0045] During the resampling process, since the original data may not be collected completely at the frequency of the set interval time T, many missing time points will be generated. In order to deal with these missing time points, the resampled data is first averaged. This is because there may be multiple original data points in the same time interval, and the average value can provide comprehensive information within the time period. However, since the original data may be irregular, there may be no data points in certain time intervals, so after calculating the average value, there will still be missing values. In order to solve this problem, the present invention uses linear interpolation to fill in the null values ​​to obtain a new data set D′. The linear interpolation method estimates the missing values ​​based on the linear relationship between the two closest valid values ​​before and after the missing value. In this way, not only is the data converted into a regular time series, but the missing values ​​are also filled, ensuring the continuity and integrity of the data.

[0046] like Figure 2 As shown, the data resampling calculation method is: for time series data, set a fixed time interval T. If the original data point D(t m ) at time t m exists, then at the time point t after resampling i Make a judgment at this point. If t i If the time matches the original data point, the data point is used directly; otherwise, if t i If the value falls between two original data points, interpolation is needed to fill it.

[0047] The average is calculated as follows:

[0048] If at a certain resampling time interval [t i ,t i +T] there are multiple original data points D(t j ), where t i <t j <t i +T, then the data value of this interval can be expressed as the average value of these points, which can be expressed by the following formula:

[0049]

[0050] Where N is the number of original data points in the time interval. The calculated average value is taken as the time point t i The sampling data of .

[0051] The linear interpolation filling calculation method is as follows:

[0052] At two known data points D(t a ) and D(t b Any time point t between c (where t a <t c <t b ), the interpolated value D(t c ) can be expressed as:

[0053]

[0054] 3. Cycle pressure judgment

[0055] The cycle pressure judgment needs to rely on the opening ratio and duration of all the support safety valves in the entire working face.

[0056] The force changes on the hydraulic support during the opening of the safety valve are as follows: when the top plate pressure is too large and exceeds the threshold for opening the hydraulic support safety valve (the threshold is related to the specific model of the hydraulic support, usually between 42MPa and 46MPa), the safety valve of the hydraulic support will be activated, the column will drop by a few millimeters to a few centimeters, and the pressure value will drop (usually within 3MPa). When it drops below the threshold, the safety valve will close.

[0057] When the working surface is under pressure, the support will continue to be under strong pressure, resulting in the pressure value not having enough time to drop below the threshold. The safety valve will continue to be open, and the pressure value will continue to rise. When it reaches a certain peak, the support column and pressure value will drop, and the safety valve will be closed when it drops below the threshold.

[0058] The specific methods for periodic pressure identification are as follows:

[0059] 1) Set the threshold Thre for opening the safety valve of the hydraulic support on the working face, which is usually 42MPa~46MPa.

[0060] 2) Count the number of supports in the statistical data set D' whose pressure value P satisfies 42 ≤ P < Thre at each recording moment, and record it as count1. Count the number of supports whose pressure value P satisfies Thre ≤ P at each recording moment, and record it as count2. For example, at a certain recording moment, among all the supports on the working surface, the number of supports that satisfies 42 ≤ P < Thre is 2, so count1 = 2, and the number of supports that satisfies Thre ≤ P is 0, so count2 = 0.

[0061] 3) Sum count1 and count2 at each moment to get the total at each moment count, and find out the count2 in total at each moment count The proportion of open rate .

[0062] 4) When open rate >0.2 and total count When ≥0.15*M, the specific time at this moment is formed into the time data set T to be judged D , where M represents the total number of hydraulic supports.

[0063] 5) Time dataset T to be judged D Statistics are performed on a daily basis. Every day, the conditions of open are met. rate >0.2 and total count The number of occurrences of moments ≥0.15*M and the duration of pressure in each day (in minutes) are calculated. Days with less than 10 occurrences and a duration less than 60 minutes are excluded.

[0064] 4. Statistics of pressure results

[0065] The number of pressure calls obtained through periodic pressure call identification is counted, and consecutive dates are regarded as one pressure call to obtain the final number of pressure calls, pressure call date, and pressure call duration.

[0066] The present invention can perform offline calculations on large amounts of historical data, or it can perform real-time calculations on real-time data sets. Both scenarios will be described below.

[0067] 1. Offline computing

[0068] Taking 119 hydraulic supports at the 81308 working face of a mine in the Xinzhou mining area of ​​Shanxi Province as an example, pressure data from 0:00 on May 15, 2023, to 24:00 on June 15, 2023, was selected. The acquisition frequency of the collected raw data set ranged from 1 second to 10 minutes, with an average acquisition frequency of 56 seconds.

[0069] Therefore, we first resampled the dataset, set the acquisition frequency to 1 minute, calculated the average value first, and used linear interpolation to fill in the missing values ​​to obtain a new dataset D′, and finally generated a dataset with 46,080 rows and 119 columns.

[0070] Set the opening threshold Thre of the hydraulic support safety valve to 45 MPa, set two segments of 42≤P<45 and 45≤P, count each row of data (i.e., each recording moment), and record the number of supports whose pressure value P meets 42≤P<Thre as count1, and the number of supports whose pressure value P meets Thre≤P as count2.

[0071] Sum count1 and count2 at each moment to get the total at each moment count , and find the count2 in total at each moment count The proportion of open rate .

[0072] When open rate >0.2 and total count ≥0.15*119, the specific time at this moment is formed into the time data set T to be judged D . Next, the time dataset T to be judged D Statistics are performed on a daily basis, including the number of occurrences per day and the time span per day (in minutes), as shown in Table 1:

[0073] Table 1

[0074] date count time_span (minutes) 2023-05-16 3 2 2023-05-18 30 134 2023-05-19 337 884 2023-05-22 156 628 2023-05-25 103 1095 2023-05-26 54 731 2023-05-28 52 525 2023-05-29 750 895 2023-06-04 15 131 2023-06-06 50 105 2023-06-09 365 518 2023-06-10 673 811 2023-06-12 1 0 2023-06-14 35 319

[0075] Then, we remove days with less than 10 occurrences and less than 60 minutes of span, and obtain Table 2:

[0076] Table 2

[0077]

[0078]

[0079] Finally, the results obtained by the periodic pressure identification algorithm are counted, and the consecutive dates are regarded as one pressure, and the final pressure number, pressure date and pressure duration are obtained. The pressure data is shown in Table 3:

[0080] Table 3

[0081] Date of press Total duration (minutes) 2023-05-18～2023-05-19 1018 2023-05-22 628 2023-05-25～2023-05-26 1826 2023-05-28～2023-05-29 1420 2023-06-04 131 2023-06-06 105 2023-06-09～2023-06-10 1329 2023-06-14 319

[0082] According to Table 3, the total number of pressures is 8.

[0083] The human observation data of the 81308 working face are shown in Table 4:

[0084] Table 4

[0085] Date of press Footprint of the day (m) Total footage (m) 2023-05-18 3.2 147.6 2023-05-22 4.0 172.2 2023-05-25 0 188.2 2023-05-29 0 207.4 2023-06-04 4.8 242.6 2023-06-09 3.2 272.2 2023-06-14 4.8 301

[0086] Combining Table 3 and Table 4, it can be seen that the calculation results obtained by the method of the present invention are consistent with the manually observed pressure conditions 7 times, with a probability of 87.5%.

[0087] 2. Real-time computing

[0088] Real-time computing requires design based on on-site conditions. The real-time computing architecture needs to be able to process continuous data streams. For example, Apache Kafka can be used as a message middleware to receive real-time data from hydraulic support sensors, and then a stream processing framework such as Apache Flink or Spark Streaming can be used for data processing.

[0089] Pressure data is collected in real time from 119 sets of hydraulic supports at the 81308 working face of a mine in the Xinzhou mining area of ​​Shanxi Province. Since data may be generated at different frequencies, it is necessary to adjust the frequency of the data stream in real time in the stream processing framework. For example, a fixed time window (such as 1 minute) can be set, and a cache (such as an in-memory database, a message queue, etc.) can be used to temporarily store data from the most recent period so that batch processing can be performed when the time window is reached. When the data in the time window accumulates to a certain extent (such as reaching a set time length or data volume), the data is aggregated. The aggregation method can be to calculate the average value, and linear interpolation or other suitable interpolation methods can be used to handle missing values.

[0090] According to the set pressure segments (such as 42≤P<45 and 45≤P), count the number of stents that meet each segment in each time window (count1 and count2). Based on the statistical results of each time window, calculate the total count and open rate According to the set threshold (such as open rate >0.2 and total count ≥0.15*M), determine whether the current time window may be in a pressure state. Once the total count and open rate When the conditions are met, an event is generated immediately.

[0091] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for identifying periodic pressure based on pressure data of hydraulic supports in fully mechanized mining working faces, characterized by: Collect offline pressure monitoring data of each hydraulic support in the fully mechanized mining face to form a data set; Setting a time interval, and resampling the formed data set at the time interval; During the resampling process, if there are multiple original data points within the time interval, an average value is calculated; if there are missing original data points within the time interval, an interpolation method is performed to fill in the empty values ​​to obtain a resampled data set; Identify periodic pressure surges based on the resampled dataset and count the number of pressure surges, the date of the surges, and the duration of the surges. The method of periodic pressure identification based on the resampled data set includes: setting the opening threshold of the hydraulic support safety valve of the fully mechanized mining working face ; Count the pressure values ​​at each recording moment in the resampled data set P conform to The number of brackets is recorded as ; Count the pressure values ​​at each recording moment in the resampled data set P conform to The number of brackets is recorded as ; For every moment and Sum up and get the value at each moment , and find the exist The proportion of ; when and When the time is added to the time data set to be judged among M Indicates the total number of hydraulic supports; Time dataset to be judged Perform statistics and count the number of people who meet the conditions every day and The number of times the moment occurs and the duration of pressure on each day; exclude days with less than 10 occurrences and duration less than 60 minutes; The number of pressure calls is counted based on the periodic pressure call identification result, wherein consecutive dates are regarded as one pressure call, and the final number of pressure calls, pressure call date and pressure call duration are obtained.

2. The method for identifying periodic incoming pressure according to claim 1, characterized in that: The mean value calculation is expressed as: Where, Indicates the set time interval. N Indicates time interval T The number of original data points in represents the original data points, ; The calculated average value is used as the time point The sampling data of .

3. The method for identifying periodic incoming pressure according to claim 1, characterized in that: The method of determining whether the original data point is missing in the time interval includes: if the original data point In time exists, then at the time point after resampling Make a judgment, if If the time does not match the original data point, interpolation is used to fill in the empty values.

4. A method for identifying periodic pressure based on hydraulic support pressure data of a fully mechanized mining working face, characterized by: The pressure data of each hydraulic support in the fully mechanized mining working face is collected in real time through pressure sensors; Set a time window, store the real-time collected pressure data within the time window, and aggregate the stored data when the data within the time window reaches the accumulation upper limit; Set the pressure judgment threshold, for the aggregated data in each time window, judge whether it is in the pressure state in a certain time window according to the pressure judgment threshold: according to the pre-set pressure segmentation and , record the number of hydraulic supports that meet the conditions of each pressure segment within the time window; The number of hydraulic supports between , set the pressure value to The number of hydraulic supports is recorded as ;in, P Indicates the pressure value of the hydraulic support. Indicates the opening threshold of the hydraulic support safety valve; Based on the statistical results within the time window, calculate ,as well as ; when and When the current time window is in a pressure state, it is determined that the pressure event is generated for recording. M Indicates the total number of hydraulic supports.

5. The method for identifying periodic incoming pressure according to claim 4, characterized in that: The cumulative upper limit is the time length set in the time window, or the data storage volume set in the time window.

Images