A method and system for real-time monitoring of mine pressure using optical fiber.

Abstract

The present invention relates to the field of data processing technology, and in particular to a method and system for real-time monitoring of mine pressure using optical fiber for mining, comprising: obtaining pressure data corresponding to multiple acquisition points on a mine, and dividing the data into a plurality of real-time time window data, obtaining the distribution similarity of the real-time time window data corresponding to each acquisition point, thereby obtaining the comprehensive distribution similarity of the real-time time window data, obtaining the stability of each real-time time window data, thereby obtaining the increase and decrease coefficient of each real-time time window data, obtaining the number of trees in the isolated forest after each real-time time window data is adjusted, thereby using an isolated forest algorithm for calculation to obtain the monitoring result of the mine pressure data. The present invention improves the accuracy of the real-time monitoring result of mine pressure using optical fiber for mining by adaptively adjusting the number of trees in the isolated forest algorithm.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a method and system for real-time monitoring of mine pressure using an optical fiber. Background Art

[0002] Fiber optic sensing technology is a sensor network based on optical fibers that monitors physical quantities in the environment by measuring changes in optical signals. In mining, fiber optic sensors can be used to monitor groundwater pressure, rock stress, changes in mine support structures, etc. Real-time pressure data obtained by mining optical fibers is usually analyzed through isolated forests.

[0003] When using the isolation forest algorithm to monitor mine pressure in real time, as the number of trees increases, the computational cost will also increase. Building and training multiple trees requires more computing resources and time. In real-time analysis, computational efficiency needs to be considered to ensure the model. The number of traditional default trees is difficult to achieve the best balance between response speed and robustness and accuracy, thereby reducing the accuracy of the real-time monitoring results of mine pressure using optical fiber. Summary of the invention

[0004] The present invention provides a method and system for real-time monitoring of mine pressure using optical fiber for mines, so as to solve the existing problems.

[0005] The present invention provides a method and system for real-time monitoring of mine pressure using optical fiber. The method and system adopt the following technical solutions:

[0006] An embodiment of the present invention provides a method for real-time monitoring of mine pressure using an optical fiber, the method comprising the following steps:

[0007] Obtain pressure data corresponding to multiple collection points on the mine; divide the pressure data corresponding to each collection point into several real-time time window data;

[0008] The distribution similarity of each real-time window data corresponding to each collection point is obtained according to the data difference between adjacent real-time window data corresponding to each collection point; the comprehensive distribution similarity of the real-time window data is obtained according to the distribution similarity of the same real-time window data of all collection points;

[0009] The stability of each real-time window data is obtained according to the difference of the same real-time window data of all acquisition points;

[0010] According to the stability of each real-time window data, the increase or decrease coefficient of each real-time window data is obtained;

[0011] According to the comprehensive distribution similarity of each real-time window data and the increase and decrease coefficient of each real-time window data, the number of trees in the isolated forest after each real-time window data is adjusted is obtained;

[0012] According to the number of trees in the isolated forest adjusted according to each real-time time window data corresponding to all the collection points in the mine, the isolation forest algorithm is used to calculate and obtain the monitoring results of the mine pressure data.

[0013] Furthermore, the pressure data corresponding to each acquisition point is divided into a plurality of real-time time window data, including the following specific steps:

[0014] In the pressure data corresponding to the ath collection point, the pressure data of every b minutes is recorded in sequence without duplication as a real-time time window data; b is a preset time window length.

[0015] Furthermore, the method of obtaining the distribution similarity of each real-time window data corresponding to each collection point according to the data difference between adjacent real-time window data corresponding to each collection point includes the following specific steps:

[0016] The first-order derivative method is used to operate each real-time window data corresponding to the a-th collection point to obtain the extreme value points in each real-time window data; in all real-time window data corresponding to the a-th collection point, the distribution similarity of each real-time window data is obtained according to the distance between the extreme value points in adjacent real-time window data.

[0017] Furthermore, the distribution similarity of each real-time window data is obtained according to the distance between the extreme value points in the adjacent real-time window data, and the corresponding specific calculation formula is:

[0018]

[0019] Among them, Q i represents the distribution similarity of the data in the i-th real-time window; N represents the number of data in each real-time window; q i,n represents the nth data in the i-th real-time window data; q i-1,n Represents the nth data in the i-1th real-time window data; σ i [d(x)] represents the standard deviation of the distances between all adjacent extreme points of the i-th real-time window data; σ i-1 [d'(x)] represents the standard deviation of the distances between all adjacent extreme points of the i-1th real-time window data; norm() is a linear normalization function, and || is an absolute value function.

[0020] Furthermore, the method of obtaining the comprehensive distribution similarity of the real-time window data according to the distribution similarity of the same real-time window data of all collection points includes the following specific steps:

[0021] Among all the collection points, any two collection points constitute a collection point combination; according to the difference between the distribution similarities of the same real-time window data in the collection point combination, the comprehensive distribution similarity of the real-time window data is obtained.

[0022] Furthermore, the specific calculation formula corresponding to the comprehensive distribution similarity of the real-time window data is obtained according to the difference between the distribution similarities of the same real-time window data in the collection point combination:

[0023]

[0024] Among them, W i Represents the comprehensive distribution similarity of the data in the i-th real-time window; represents the mean value of the distribution similarity of the i-th real-time window data of all acquisition points; L represents the number of acquisition point combinations; Δ c (Q i ) represents the absolute value of the difference in distribution similarity between the i-th real-time window data of two collection points in the c-th collection point combination in the collection point combination; norm() is a linear normalization function; d is a preset difference threshold.

[0025] Furthermore, the step of obtaining the stationarity of each real-time window data according to the difference of the same real-time window data of all acquisition points includes the following specific steps:

[0026] With time as the horizontal axis and pressure data as the vertical axis, a scatter plot of the pressure data corresponding to each acquisition point is constructed; according to the angle between the connecting straight line of adjacent extreme value points in the real-time window data in the scatter plot and the horizontal axis, the specific calculation formula corresponding to the stability of each real-time window data is obtained as follows:

[0027]

[0028] Among them, E i Indicates the stability of the i-th real-time window data; M h,i represents the number of extreme points in the i-th real-time window data of the h-th acquisition point; θ i,h,m Indicates the minimum angle between the connecting line between the mth extreme point and the m+1th extreme point in the scatter plot of the i-th real-time window data of the h-th acquisition point and the horizontal axis; θ' is the preset angle threshold; H is the number of acquisition points; It means that when m belongs to [1,M h,i -1] The maximum value of; norm() is the linear normalization function.

[0029] Furthermore, the step of obtaining the increase or decrease coefficient of each real-time window data according to the stationarity of each real-time window data includes the following specific steps:

[0030] The difference between the stationarity of the i-th real-time window data and the preset difference threshold is recorded as the first difference;

[0031] The absolute value of the difference between the stationarity of the i-th real-time window data and the preset difference threshold is recorded as the second difference;

[0032] The ratio of the first difference to the second difference is calculated, and the product of the ratio, the preset range coefficient and the second difference is recorded as the increase or decrease coefficient of the i-th real-time time window data.

[0033] Furthermore, the method of obtaining the number of trees in the isolated forest after each real-time window data is adjusted according to the comprehensive distribution similarity of each real-time window data and the increase and decrease coefficient of each real-time window data includes the following specific steps:

[0034] According to the comprehensive distribution similarity and increase / decrease coefficient of each real-time window data, the iteration starts from the preset initial number of trees U0, and the iteration formula corresponding to the number of trees in the isolated forest adjusted for each real-time window data is obtained in turn:

[0035]

[0036] Among them, U i represents the number of trees in the isolated forest after the adjustment of the i-th real-time window data; U i-1 represents the number of trees in the isolated forest after the adjustment of the i-1th real-time window data; W i represents the comprehensive distribution similarity of the data in the i-th real-time window; R i represents the increase or decrease coefficient of the i-th real-time window data; y is the preset reference value of the adjustment range; Represents the ceiling function.

[0037] The present invention also proposes a real-time monitoring system for mine pressure using optical fiber, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the steps of the aforementioned real-time monitoring method for mine pressure using optical fiber are implemented.

[0038] The beneficial effects of the technical solution of the present invention are as follows: pressure data corresponding to multiple acquisition points on a mine are obtained, and the data are divided into a number of real-time window data, the distribution similarity of the real-time window data corresponding to each acquisition point is obtained, and the comprehensive distribution similarity of the real-time window data is obtained according to the distribution similarity of the same real-time window data of all acquisition points, thereby obtaining the comprehensive distribution similarity through analysis between the same real-time window data of multiple acquisition points, and using this as the calculation parameter for the number of adaptive trees to ensure the accuracy of the number of adaptive trees; the comprehensive distribution similarity of the real-time window data is obtained according to the distribution similarity of the same real-time window data of all acquisition points, the stability of each real-time window data is obtained according to the difference of the same real-time window data of all acquisition points, and the stability of each real-time window data is obtained according to the difference of the same real-time window data of each real-time window data. The increase and decrease coefficient of each real-time window data is obtained according to the stability of the real-time window data, and the number of trees in the isolated forest after adjustment of each real-time window data is obtained according to the comprehensive distribution similarity of each real-time window data and the increase and decrease coefficient. The comprehensive distribution similarity is adjusted by the increase and decrease coefficient, which further improves the accuracy of the number of adaptive trees, thereby ensuring the accuracy of the monitoring results; the number of trees in the isolated forest is adjusted according to the needs of each real-time window data corresponding to all the collection points on the mine, and the isolation forest algorithm is used for calculation to obtain the monitoring results of the mine pressure data. At this point, the number of trees in the isolated forest is dynamically adjusted through real-time analysis, which ensures the accuracy of real-time detection of mine pressure while reducing the increase in computing cost due to too many trees, thereby improving the accuracy of real-time monitoring results of mine pressure by optical fiber. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention 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 of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 The present invention is a flowchart of the steps of a method for real-time monitoring of mine pressure using optical fiber. DETAILED DESCRIPTION

[0041] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of a mine pressure real-time monitoring method and system for optical fiber mines proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0042] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0043] The specific scheme of a method and system for real-time monitoring of mine pressure with optical fiber provided by the present invention is described in detail below with reference to the accompanying drawings.

[0044] See also Figure 1 , which shows a flowchart of a method for real-time monitoring of mine pressure using an optical fiber provided by an embodiment of the present invention, the method comprising the following steps:

[0045] Step S001: Obtain pressure data corresponding to multiple collection points on a mine; divide the pressure data corresponding to each collection point into a number of real-time time window data.

[0046] Pressure acquisition devices are installed at various locations in the mine to obtain multiple pressure data at each location to facilitate subsequent analysis.

[0047] Because rock formation stress is a key parameter for evaluating rock stability and mine safety, this embodiment uses mining optical fibers in key locations such as underground mine tunnels and underground coal mining faces to monitor geological stress, rock formation deformation, and other factors that may affect mine safety and stability. At the same time, the preset acquisition frequency in this embodiment is once per second, which is described as an example. Other implementations may be set to other values, which are not limited in this embodiment.

[0048] Therefore, this embodiment can obtain multiple pressure data of various positions of mine pressure in real time, and each position corresponds to a collection point, thereby obtaining pressure data corresponding to multiple collection points in the mine.

[0049] When analyzing the pressure data of a mine in this embodiment, 5 minutes of pressure data can be collected first as the reference time window data. A time window is set, and when new real-time data is collected, the degree of change between the real-time window data and the previous time window data is analyzed. The degree of change of normal data is relatively small. When the analyzed real-time window data changes greatly compared with the previous time window data, this indicates that the data of the real-time window corresponds to two situations, more stable or more volatile, so the data of the real-time window should be given a larger tree expansion range.

[0050] In this embodiment, the preset time window length b is 5 minutes, which is described as an example. Other implementations may be set to other values, which are not limited in this embodiment. To facilitate the subsequent real-time time window data analysis, this embodiment takes the pressure data corresponding to the a-th collection point as an example. In the pressure data corresponding to the a-th collection point, the pressure data of every b minutes is recorded as a real-time time window data without duplication, which is equivalent to adjusting the number of trees in the isolated forest once every b minutes, and b is the preset time window length.

[0051] Step S002: obtaining the distribution similarity of each real-time window data corresponding to each collection point according to the data difference between adjacent real-time window data corresponding to each collection point; obtaining the comprehensive distribution similarity of the real-time window data according to the distribution similarity of the same real-time window data of all collection points.

[0052] This embodiment determines the comprehensive distribution similarity of the real-time window data based on the distribution similarity of the real-time window data and its corresponding previous time window data. The distribution similarity of the data refers to characteristic information such as the fluctuation amplitude of the data, the extreme value of the data, and the average value of the data.

[0053] The first-order derivative method is used to calculate each real-time window data to obtain the extreme value point in each real-time window data. The first-order derivative method is a well-known technology, and the specific method is not introduced here.

[0054] Taking the i-th real-time window data corresponding to the a-th collection point as an example, for the real-time window data and the previous time window data corresponding to it, mark the extreme point data, and count the distances between adjacent extreme points, that is, the number of data between adjacent extreme points.

[0055]

[0056] Formula explanation: Q i represents the distribution similarity of the data in the ith real-time window, which is determined by analyzing the corresponding previous time window; N represents the number of data in each real-time window; q i,nrepresents the nth data in the i-th real-time window data; q i-1,n Represents the nth data in the i-1th real-time window data; σ i [d(x)] represents the standard deviation of the distances between all adjacent extreme points of the i-th real-time window data; σ i-1 [d'(x)] represents the standard deviation of the distances between all adjacent extreme points of the i-1th real-time window data. norm() is a linear normalization function that normalizes the data value to the interval [0,1]; || is an absolute value function. It should be noted that for the first time window, the corresponding i-1th time window refers to the base time window.

[0057] In the formula, It means that the average of the cumulative sum of the data differences between the same sequence number points in adjacent time windows is used to compare the similarity of the corresponding data in adjacent time windows; It means analyzing the similarity of data in adjacent time windows from an overall perspective, which is reflected by the standard deviation of the two data. The closer the standard deviations are, the more similar the data are.

[0058] Thus, the distribution similarity of each real-time window data corresponding to each collection point is obtained.

[0059] When collecting data in this embodiment, this embodiment collects multiple data, that is, the distribution similarity of multiple data corresponds to the same time window. When mining optical fiber collects mine pressure data, when an abnormality occurs, there is a certain correlation between the pressure collection points at various locations. In layman's terms, when one collection point senses a pressure abnormality, the other collection points should also sense a certain degree of pressure abnormality. Even if the degree is not completely consistent, there is a certain correlation.

[0060] The distribution similarity of the i-th real-time window data obtained in the above process is only the distribution similarity of the real-time window data of one collection point. Next, the comprehensive distribution similarity of each real-time window data needs to be determined based on the distribution similarity of the same real-time window data of each collection point.

[0061] Among all the collection points, any two collection points constitute a collection point combination, thereby obtaining several collection point combinations.

[0062]

[0063] Formula explanation: W i Represents the comprehensive distribution similarity of the data in the i-th real-time window; represents the mean value of the distribution similarity of the i-th real-time window data of all acquisition points; L represents the number of acquisition point combinations; Δc (Q i ) represents the absolute value of the difference in distribution similarity between the i-th real-time window data of two collection points in the c-th collection point combination in the collection point combination; norm() is a linear normalization function that normalizes the data value to the interval [0,1]; d is the preset difference threshold.

[0064] In this embodiment, the preset difference threshold d is 0.5, which is used as an example for description. In other implementations, it can be set to other values, which are not limited in this embodiment.

[0065] In the formula, This is used as the benchmark value, and the benchmark is adjusted by analyzing the data distribution similarity differences of each collection point in pairs. c (Q i ) are smaller, indicating that the similarity of their corresponding distributions is more consistent, d-Δ c (Q i ), which means the more consistent (or greater) the similarity of the comprehensive distribution is. The correlation between the collection points is determined by averaging the two-by-two analysis. The stronger the correlation, the greater the adjustment of the final comprehensive distribution similarity in this embodiment. That is, Corresponding to the embodiment of this part.

[0066] Thus, this embodiment can obtain the comprehensive distribution similarity of the real-time time window, which will subsequently correspond to the expansion range of the number of trees in the isolated forest.

[0067] Step S003: obtaining the stability of each real-time window data according to the difference of the same real-time window data of all acquisition points.

[0068] Through the above process, this embodiment determines the expansion range of the number of trees in the isolated forest. Next, it is necessary to analyze the stability of the real-time window data. It can be determined by paying attention to its data waveform performance, extreme point distribution, etc. At the same time, it is also necessary to comprehensively determine it based on the stability of the real-time window data of different collection points at the same time.

[0069] With time as the horizontal axis and pressure data as the vertical axis, a scatter plot of pressure data corresponding to each acquisition point is constructed. In this embodiment, the preset angle threshold θ' is 15°, which is used as an example for description. Other implementations may be set to other values, which are not limited in this embodiment.

[0070] The specific process is as follows:

[0071]

[0072] Formula explanation: E iIndicates the stability of the i-th real-time window data; M h,i represents the number of extreme points in the i-th real-time window data of the h-th acquisition point; θ i,h,m Indicates the minimum angle between the connecting line between the mth extreme point and the m+1th extreme point in the scatter plot in the i-th real-time window data of the h-th acquisition point and the horizontal axis; θ' is the preset angle threshold; It means that when m belongs to [1,M h,i -1] The maximum value of; norm() is a linear normalization function that normalizes the data value to the interval [0,1]. H represents the number of collection points.

[0073] In the formula, The stability of the data is reflected by comparing the angle formed by adjacent extreme points with the angle threshold, which is used as the benchmark value. The largest angle difference ratio is extracted to adjust the data stability by amplifying or reducing the data. Such operation will make the final data relatively stable more stable and the chaotic more chaotic. Thus, the present embodiment can obtain the stability of each real-time time window data.

[0074] Step S004: Obtain the increase or decrease coefficient of each real-time window data according to the stability of each real-time window data.

[0075] Because the expansion range of the number of trees in the isolated forest of this embodiment is determined based on the real-time window data and the previous time window data, the more consistent the distribution uniformity of the two is, the larger the corresponding expansion range is, but it is unclear whether this distribution uniformity corresponds to abnormal or normal data performance, so the increase and decrease coefficient of the real-time window is determined by the stability of the data in the real-time window. For abnormal data, an increasing coefficient is determined; and for normal data, a decreasing coefficient is determined. The specific process of determining the increase and decrease coefficient of each real-time window according to the stability of each real-time window data is as follows:

[0076]

[0077] Formula explanation: R i represents the increase or decrease coefficient of the i-th real-time window data; E i represents the stability of the data in the ith real-time window; the preset range coefficient z is 2, and the preset difference threshold d is 0.5. This is taken as an example for description. Other implementations may be set to other values, which are not limited in this embodiment. It is an absolute value function.

[0078] In the formula, Determine the increase or decrease of the increase or decrease coefficient of the real-time window data. Data greater than d is considered stable data, and data less than d is considered chaotic data.

[0079] In the formula, z×|E i -d| represents the coefficient size of the increase or decrease coefficient of the real-time window data. Because for more stable data, the number of trees in the isolated forest should be reduced to a greater extent to improve the efficiency of real-time detection; and for more chaotic data, the number of trees in the isolated forest should be increased to a greater extent to ensure the accuracy of real-time detection of mine pressure. The multiplication of z is to ensure that the expansion range of the number of trees in the isolated forest is still 0-1.

[0080] Step S005: Obtain the number of trees in the isolated forest after adjustment of each real-time window data according to the comprehensive distribution similarity of each real-time window data and the increase / decrease coefficient of each real-time window data.

[0081] The number of trees in the final adjustment of the isolated forest is determined by the comprehensive distribution uniformity of the real-time window data and the increase and decrease coefficient of the real-time window data through the above process. i The larger the value is, the larger the actual expansion range obtained from the reference adjustment range is. i The larger the value (positive or negative), the greater the actual expansion value in the actual expansion range.

[0082] Then the iterative formula for the number of trees in the isolated forest adjusted for each real-time window data is:

[0083]

[0084] Formula explanation: U i represents the number of trees in the isolated forest after the adjustment of the i-th real-time window data; U i-1 Indicates the number of trees in the isolated forest after the adjustment of the i-1th real-time time window data. When i=1, the i-1th refers to the number of trees in the reference time window. In this embodiment, the preset initial number of trees U0 is 120, and the preset reference value y of the adjustment range is 20. This is used as an example for description. Other implementations can be set to other values, which are not limited in this embodiment. W i represents the comprehensive distribution similarity of the data in the i-th real-time window. i The larger the value, the larger the expansion range. i Indicates the increase or decrease coefficient of the i-th real-time time window data. Represents the ceiling function.

[0085] Thus, the number of trees in the isolated forest adjusted for each real-time window data can be obtained.

[0086] Step S006: Based on the number of trees in the isolated forest adjusted according to each real-time window data corresponding to all the collection points on the mine, the isolation forest algorithm is used to perform calculations to obtain the monitoring results of the mine pressure data.

[0087] According to the number of trees in the isolated forest adjusted according to each real-time time window data corresponding to all the collection points in the mine, the isolation forest algorithm is used to calculate the pressure data of all the collection points in the mine to obtain the monitoring results of the mine pressure data.

[0088] It should be noted that the isolation forest algorithm is a well-known technology and the specific method will not be introduced here. The output results of the algorithm are abnormal data and normal data. Therefore, the monitoring results of the mine pressure data are normal mine pressure data or abnormal mine pressure data. When the mine pressure data is abnormal, the monitoring results will be fed back to the monitoring personnel.

[0089] So far, the present invention is completed.

[0090] To sum up, in an embodiment of the present invention, the present invention also provides a real-time monitoring system for mine pressure using optical fiber for mining, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program stored in the memory to implement the steps of the aforementioned real-time monitoring method for mine pressure using optical fiber for mining.

[0091] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for real-time monitoring of mine pressure using optical fiber, characterized in that: The method comprises the following steps: Obtain pressure data corresponding to multiple collection points on the mine; divide the pressure data corresponding to each collection point into several real-time time window data; The distribution similarity of each real-time window data corresponding to each collection point is obtained according to the data difference between adjacent real-time window data corresponding to each collection point; the comprehensive distribution similarity of the real-time window data is obtained according to the distribution similarity of the same real-time window data of all collection points; The stability of each real-time window data is obtained according to the difference of the same real-time window data of all acquisition points; According to the stability of each real-time window data, the increase or decrease coefficient of each real-time window data is obtained; According to the comprehensive distribution similarity of each real-time window data and the increase and decrease coefficient of each real-time window data, the number of trees in the isolated forest after each real-time window data is adjusted is obtained; According to the number of trees in the isolated forest adjusted for each real-time window data corresponding to all the collection points on the mine, the isolated forest algorithm is used to calculate and obtain the monitoring results of the mine pressure data; The step of obtaining the stationarity of each real-time window data according to the difference of the same real-time window data of all acquisition points includes the following specific steps: With time as the horizontal axis and pressure data as the vertical axis, a scatter plot of the pressure data corresponding to each acquisition point is constructed; according to the angle between the connecting straight line of adjacent extreme value points in the real-time window data in the scatter plot and the horizontal axis, the specific calculation formula corresponding to the stability of each real-time window data is obtained as follows: in, Indicates the stationarity of the i-th real-time window data; Represents the number of extreme points in the i-th real-time window data of the h-th acquisition point; Indicates the minimum angle between the connecting line between the mth extreme point and the m+1th extreme point in the ith real-time window data of the hth acquisition point and the horizontal axis in the scatter plot; is the preset angle threshold; H is the number of acquisition points; Indicates when belong hour The maximum value of is a linear normalization function.

2. According to claim 1, a method for real-time monitoring of mine pressure using optical fiber, characterized in that: The specific steps of dividing the pressure data corresponding to each acquisition point into a plurality of real-time time window data are as follows: In the pressure data corresponding to the ath collection point, the pressure data of every b minutes is recorded in sequence without duplication as a real-time time window data; b is a preset time window length.

3. According to claim 1, a method for real-time monitoring of mine pressure using optical fiber, characterized in that: The specific steps of obtaining the distribution similarity of each real-time window data corresponding to each collection point according to the data difference between adjacent real-time window data corresponding to each collection point are as follows: The first-order derivative method is used to operate each real-time window data corresponding to the a-th collection point to obtain the extreme value points in each real-time window data; in all real-time window data corresponding to the a-th collection point, the distribution similarity of each real-time window data is obtained according to the distance between the extreme value points in adjacent real-time window data.

4. According to claim 3, a method for real-time monitoring of mine pressure using optical fiber, characterized in that: The distribution similarity of each real-time window data is obtained according to the distance between the extreme value points in the adjacent real-time window data, and the corresponding specific calculation formula is: in, represents the distribution similarity of the data in the i-th real-time window; N represents the number of data in each real-time window; Represents the nth data in the i-th real-time time window data; Represents the nth data in the i-1th real-time time window data; Represents the standard deviation of the distances between all adjacent extreme points of the i-th real-time window data; Represents the standard deviation of the distances between all adjacent extreme points of the i-1th real-time window data; is the linear normalization function, is the absolute value function.

5. According to claim 1, a method for real-time monitoring of mine pressure using optical fiber, characterized in that: The method of obtaining the comprehensive distribution similarity of the real-time window data according to the distribution similarity of the same real-time window data of all the collection points includes the following specific steps: Among all the collection points, any two collection points constitute a collection point combination; according to the difference between the distribution similarities of the same real-time window data in the collection point combination, the comprehensive distribution similarity of the real-time window data is obtained.

6. A method for real-time monitoring of mine pressure using optical fiber according to claim 5, characterized in that: The specific calculation formula corresponding to the comprehensive distribution similarity of the real-time window data is obtained according to the difference between the distribution similarities of the same real-time window data in the collection point combination: in, Represents the comprehensive distribution similarity of the data in the i-th real-time time window; Represents the mean of the distribution similarity of the i-th real-time window data of all acquisition points; Indicates the number of collection point combinations; Indicates the absolute value of the difference between the distribution similarity of the i-th real-time window data of two collection points in the c-th collection point combination in the collection point combination; is a linear normalization function; d is the preset difference threshold.

7. The method for real-time monitoring of mine pressure using optical fiber according to claim 1, characterized in that: The specific steps of obtaining the increase or decrease coefficient of each real-time window data according to the stability of each real-time window data are as follows: The difference between the stationarity of the i-th real-time window data and the preset difference threshold is recorded as the first difference; The absolute value of the difference between the stationarity of the i-th real-time window data and the preset difference threshold is recorded as the second difference; The ratio of the first difference to the second difference is calculated, and the product of the ratio, the preset range coefficient and the second difference is recorded as the increase or decrease coefficient of the i-th real-time time window data.

8. The method for real-time monitoring of mine pressure using optical fiber according to claim 1, characterized in that: The specific steps of obtaining the number of trees in the isolated forest after each real-time window data is adjusted according to the comprehensive distribution similarity of each real-time window data and the increase and decrease coefficient of each real-time window data are as follows: According to the comprehensive distribution similarity and increase / decrease coefficient of each real-time window data, the initial number of trees is preset. Start iterating, and get the iterative formula corresponding to the number of trees in the isolated forest adjusted for each real-time window data: in, represents the number of trees in the isolated forest after adjusting the data of the i-th real-time window; represents the number of trees in the isolated forest after the adjustment of the i-1th real-time window data; Represents the comprehensive distribution similarity of the data in the i-th real-time time window; represents the increase or decrease coefficient of the i-th real-time window data; y is the preset reference value of the adjustment range; Represents the ceiling function.

9. A real-time monitoring system for mine pressure using optical fiber, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the computer program is executed by the processor, the steps of the method for real-time monitoring of mine pressure using optical fiber are implemented as described in any one of claims 1 to 8.

Images