An intelligent early warning system and method for coal mine water hazard monitoring
By calculating the comprehensive anomaly confidence and anomaly score correction factor of coal mine environmental data, the deviation of the isolated forest algorithm when distinguishing real anomaly data from noise data is solved, and the warning accuracy of the coal mine water damage monitoring and early warning system is improved.
Patent Information
- Application Number
- CN202311709753.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-12-13
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2043-12-13
AI Technical Summary
The high degree of isolation of noise data in coal mine environmental data makes it difficult for the isolated forest algorithm to distinguish between real abnormal data and noise data, resulting in deviations in the early warning system analysis and unable to issue early warnings in a timely manner.
By obtaining the abnormal scores of coal mine environmental data, the comprehensive abnormality confidence and abnormality score correction factors are calculated, and the abnormality score is corrected to distinguish the true abnormality data and filter out the true abnormality data in the coal mine.
It improves the early warning accuracy of the coal mine water damage monitoring and early warning system, and can effectively distinguish between real abnormal data and noise data to ensure timely monitoring and early warning.
Smart Images

Figure CN117605539B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data processing, and in particular to an intelligent early warning system and method for monitoring water hazards in coal mines. Background Art
[0002] During the construction and production process of coal mines, it is possible to encounter water seepage and water gushing underground. Surface water and groundwater flow into the mine through various channels. When the water gushing in the mine exceeds the normal drainage capacity of the mine, it will cause mine flooding, which will not only affect the normal production of the mine, but also cause casualties, which is very harmful. Therefore, real-time monitoring of hydrological, geological, environmental and other data in coal mines and taking early warning measures for water disasters are important links to ensure the smooth progress of the project. Among them, monitoring environmental data such as humidity, temperature, and gas concentration in coal mines can timely detect environmental anomalies in the mine, prevent the occurrence of phenomena such as water level rise, rock deformation, and gas explosion that may cause water disasters, and effectively prevent and reduce the occurrence of coal mine water disasters. Therefore, the monitoring of coal mine environmental data is of great significance to the prevention of water disasters. When monitoring and analyzing the anomalies of coal mine environmental data, a commonly used algorithm is the isolation forest algorithm. The algorithm starts from the abnormal point, divides by formulating rules, and obtains the abnormal score of all data according to multiple division times, so as to judge the abnormal data.
[0003] Environmental factors and equipment problems inside coal mines can cause noise data to be generated in the coal mine environmental data. When using the isolation forest algorithm to monitor and analyze the coal mine environmental data, the core of the algorithm is to judge whether it is abnormal based on the degree of isolation of the data. The degree of isolation of these noise data is often high, even higher than the real abnormal data, so their anomaly scores are also high, making it impossible to distinguish between real abnormal data and noise data, which will cause deviations in the system's analysis and prediction, resulting in the system's inability to issue early warnings through the control system's alarm. Summary of the invention
[0004] In order to solve the above problems, the present invention provides a coal mine water hazard monitoring intelligent early warning system and method.
[0005] An embodiment of the present invention provides a method for intelligent early warning of coal mine water hazard monitoring, the method comprising the following steps:
[0006] Acquire a coal mine environment data sequence, the coal mine environment data sequence includes coal mine environment data at a number of sampling moments, the coal mine environment data at each sampling moment refers to coal mine temperature data, coal mine humidity data and coal mine gas concentration data at each sampling moment;
[0007] The anomaly score of the coal mine environmental data at each sampling moment is obtained by using the isolation forest algorithm, and the coal mine environmental data is screened according to the anomaly score to obtain all target anomaly data;
[0008] Arrange all target abnormal data according to the time sequence of sampling time to obtain a target abnormal data sequence; obtain the adjacent abnormal data segment of each target abnormal data in the target abnormal data sequence; obtain the adjacent range data segment of each target abnormal data in the coal mine environment data sequence; obtain the comprehensive abnormal confidence of each target abnormal data according to the difference in abnormal scores between each target abnormal data and other target abnormal data in the adjacent abnormal data segment and the difference in coal mine humidity data between each target abnormal data and coal mine environment data at other sampling times in the adjacent range data segment; obtain the abnormal score correction factor of each target abnormal data according to the comprehensive abnormal confidence of each target abnormal data and the difference between each target abnormal data and coal mine environment data at other sampling times in the adjacent range data segment;
[0009] The anomaly score of each target anomaly data is corrected according to the anomaly score correction factor, and the coal mine environmental data is screened according to the corrected anomaly score to obtain all the real anomaly data; according to the distribution of the real anomaly data in the coal mine environmental data sequence, the construction environment in the coal mine is monitored.
[0010] Preferably, the method of screening the coal mine environmental data according to the abnormality score to obtain all target abnormal data includes:
[0011] A threshold T1 is preset. For the coal mine environment data at any sampling time in the coal mine environment data sequence, if the anomaly score of the coal mine environment data at the sampling time is greater than or equal to T1, the coal mine environment data at the sampling time is recorded as the target anomaly data of the coal mine environment data sequence, and then all the target anomaly data of the coal mine environment data sequence are obtained.
[0012] Preferably, the step of obtaining the adjacent abnormal data segments of each target abnormal data in the target abnormal data sequence includes the following specific methods:
[0013] A parameter a is preset, and for the ith target abnormal data in the target abnormal data sequence, a data segment consisting of the first a target abnormal data of the ith target abnormal data and the last a target abnormal data of the ith target abnormal data is used as the adjacent abnormal data segment of the ith target abnormal data.
[0014] Preferably, the method of obtaining the adjacent range data segment of each target abnormal data in the coal mine environment data sequence includes:
[0015] For the kth target abnormal data in the coal mine environmental data sequence, the data segment consisting of the coal mine environmental data at the a sampling moments before the kth target abnormal data and the coal mine environmental data at the a sampling moments after the kth target abnormal data is used as the adjacent range data segment of the kth target abnormal data, and a is a preset parameter.
[0016] Preferably, the comprehensive abnormality confidence of each target abnormal data is obtained according to the abnormality score difference between each target abnormal data and other target abnormal data in the adjacent abnormal data segment and the coal mine humidity data difference between each target abnormal data and the coal mine environmental data at other sampling times in the adjacent range data segment, including the specific method of:
[0017] According to the difference in abnormal scores between each target abnormal data and other target abnormal data in the adjacent abnormal data segment, the degree of abnormality in the adjacent area of each target abnormal data in the coal mine environmental data sequence is obtained; according to the difference in coal mine humidity data between each target abnormal data and the coal mine environmental data at other sampling times in the adjacent range data segment, the degree of range abnormality of each target abnormal data in the coal mine environmental data sequence is obtained;
[0018] The product of the abnormality degree of the adjacent area of each target abnormal data in the coal mine environmental data sequence and the abnormality degree of the range of each target abnormal data in the coal mine environmental data sequence is taken as the comprehensive abnormality confidence of each target abnormal data.
[0019] Preferably, the specific formula for obtaining the abnormality degree of the adjacent area of each target abnormal data in the coal mine environment data sequence according to the abnormality score difference between each target abnormal data and other target abnormal data in the adjacent abnormal data segment is:
[0020]
[0021] In the formula, represents the abnormal degree of the neighboring area of the ith target abnormal data in the coal mine environmental data sequence; t i represents the sampling time corresponding to the i-th target abnormal data in the coal mine environment data sequence; t i,j represents the sampling time corresponding to the jth target abnormal data in the adjacent abnormal data segment of the i-th target abnormal data in the coal mine environment data sequence; h i represents the abnormal score of the i-th target abnormal data in the coal mine environment data sequence; h i,j It represents the anomaly score of the jth target anomaly data in the adjacent anomaly data segment of the i-th target anomaly data in the coal mine environment data sequence; a is a preset parameter; || means taking the absolute value; norm() represents the linear normalization function.
[0022] Preferably, the method of obtaining the range abnormality degree of each target abnormal data in the coal mine environmental data sequence according to the difference in coal mine humidity data between each target abnormal data and the coal mine environmental data at other sampling moments in the adjacent range data segment includes the following specific methods:
[0023] For the coal mine environmental data at the vth sampling moment in the adjacent range data segment of the kth target abnormal data, the absolute value of the difference between the coal mine humidity data at the vth sampling moment and the coal mine humidity data at the v+1th sampling moment is used as the first-order difference of humidity of the coal mine environmental data at the vth sampling moment; the absolute value of the difference between the first-order difference of humidity of the coal mine environmental data at the vth sampling moment and the first-order difference of humidity of the coal mine environmental data at the v+1th sampling moment is used as the second-order difference of humidity of the coal mine environmental data at the vth sampling moment; then the range abnormality degree calculation method of the kth target abnormal data of the coal mine environmental data sequence is:
[0024]
[0025] In the formula, Indicates the range of abnormality of the kth target abnormal data in the coal mine environmental data sequence; ΔH1 k,v Represents the first-order difference of humidity of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔH2 k,v It represents the second-order difference of humidity of the coal mine environmental data at the vth sampling moment in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; a is a preset parameter; norm() represents a linear normalization function.
[0026] Preferably, the method of obtaining the abnormal score correction factor of each target abnormal data according to the comprehensive abnormal confidence of each target abnormal data and the difference between each target abnormal data and the coal mine environmental data at other sampling times in the adjacent range data segment includes the following specific methods:
[0027] For the kth target abnormal data in the coal mine environmental data sequence, the environmental data change speed and environmental data change amount of the kth target abnormal data in the coal mine environmental data sequence are obtained; then the calculation method of the abnormal score correction factor of the kth target abnormal data in the coal mine environmental data sequence is:
[0028]
[0029] In the formula, δ k represents the abnormal score correction factor of the kth target abnormal data in the coal mine environmental data sequence; Q k W represents the environmental data change speed of the kth target abnormal data in the coal mine environmental data sequence;k Indicates the environmental data change of the kth target abnormal data in the coal mine environmental data sequence; It represents the comprehensive anomaly confidence of the kth target anomaly data in the coal mine environmental data sequence; norm() represents the linear normalization function.
[0030] Preferably, the specific method of obtaining the environmental data change speed and environmental data change amount of the kth target abnormal data in the coal mine environmental data sequence includes:
[0031] For the coal mine environmental data at the vth sampling moment in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence, the absolute value of the difference between the coal mine temperature data at the vth sampling moment and the coal mine temperature data at the v+1th sampling moment is used as the first-order temperature difference of the coal mine environmental data at the vth sampling moment; the absolute value of the difference between the first-order temperature difference of the coal mine environmental data at the vth sampling moment and the first-order temperature difference of the coal mine environmental data at the v+1th sampling moment is used as the second-order temperature difference of the coal mine environmental data at the vth sampling moment; The absolute value of the difference between the coal mine gas concentration data of the coal mine environmental data at the vth sampling moment and the coal mine gas concentration data of the coal mine environmental data at the v+1th sampling moment is taken as the first-order difference of the gas concentration of the coal mine environmental data at the vth sampling moment; the absolute value of the difference between the first-order difference of the gas concentration of the coal mine environmental data at the vth sampling moment and the first-order difference of the gas concentration of the coal mine environmental data at the v+1th sampling moment is taken as the second-order difference of the gas concentration of the coal mine environmental data at the vth sampling moment; then the calculation method of the environmental data change speed and environmental data change amount of the kth target abnormal data in the coal mine environmental data sequence is:
[0032]
[0033]
[0034] Where, ΔH1 k,v Represents the first-order difference of humidity of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔH2 k,v Represents the second-order difference of humidity of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔC1 k,v The first-order difference of the gas concentration of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔC2 k,vIndicates the second-order difference of gas concentration of coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔF1 k,v Indicates the first-order difference of the temperature of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔF2 k,v It represents the second-order difference of the temperature of the coal mine environmental data at the vth sampling moment in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; a is a preset parameter; || represents the absolute value.
[0035] The present invention also proposes a coal mine water hazard monitoring intelligent early warning system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any one of the steps of the coal mine water hazard monitoring intelligent early warning method when executing the computer program.
[0036] The beneficial effect of the technical solution of the present invention is as follows: the present invention obtains the anomaly score correction factor of each target anomaly data according to the comprehensive anomaly confidence of each target anomaly data and the difference between each target anomaly data and the coal mine environmental data at other sampling moments in the adjacent range data segment, obtains the corrected anomaly score of each target anomaly data according to the anomaly score correction factor of each target anomaly data, screens the coal mine environmental data according to the corrected anomaly score to obtain all real anomaly data, and then distinguishes the real anomaly data and the noise data in the coal mine environmental data sequence, monitors the construction environment in the coal mine shaft according to the distribution of the real anomaly data in the coal mine environmental data sequence, and makes the early warning effect of the coal mine early warning system more accurate. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] 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.
[0038] Figure 1 The present invention is a flowchart of the steps of a coal mine water hazard monitoring intelligent early warning method. DETAILED DESCRIPTION
[0039] 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 coal mine water hazard monitoring intelligent early warning system and method 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.
[0040] 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.
[0041] The specific scheme of the intelligent early warning system and method for coal mine water hazard monitoring provided by the present invention is described in detail below with reference to the accompanying drawings.
[0042] See also Figure 1 , which shows a flow chart of a method for intelligent early warning of coal mine water hazard monitoring provided by an embodiment of the present invention, the method comprising the following steps:
[0043] Step S001: Acquire coal mine environment data sequence.
[0044] It should be noted that monitoring environmental data such as humidity, temperature, and gas concentration in coal mines can promptly detect environmental anomalies in the mines, prevent phenomena such as rising water levels, rock deformation, and gas explosions that may cause water disasters, and effectively prevent and reduce the occurrence of coal mine water disasters. Therefore, the monitoring of coal mine environmental data is even more important for the prevention of water disasters.
[0045] Specifically, in order to implement the coal mine water hazard monitoring intelligent early warning method proposed in this embodiment, it is first necessary to collect the coal mine environmental data sequence, and the specific process is as follows:
[0046] This embodiment presets a sampling moment every 3 minutes, and each time, the three dimensional data types of coal mine temperature data, coal mine humidity data and coal mine gas concentration data are collected in turn through the mine humidity sensor, the mine temperature sensor and the gas concentration detector, and the collection is carried out for a total of 24 hours; the three dimensional data of coal mine temperature data, coal mine humidity data and coal mine gas concentration data at each sampling moment are used as the coal mine environment data sequence.
[0047] The coal mine environment data sequence includes coal mine environment data at several sampling moments, and the coal mine environment data at each sampling moment refers to the coal mine temperature data, coal mine humidity data and coal mine gas concentration data at each sampling moment.
[0048] So far, the coal mine environmental data sequence is obtained through the above method.
[0049] Step S002: Acquire all target abnormal data of the coal mine environment data sequence.
[0050] It should be noted that when monitoring and analyzing anomalies in coal mine environmental data, a commonly used algorithm is the isolation forest algorithm. This algorithm starts from the anomaly point, divides it by formulating rules, and obtains the anomaly score of all data based on multiple division times, thereby determining the abnormal data.
[0051] A threshold value T1 is preset, wherein this embodiment is described by taking T1=0.75 as an example, and this embodiment is not specifically limited, wherein T1 is determined according to specific implementation conditions.
[0052] Specifically, the anomaly score of the coal mine environmental data at each sampling time in the coal mine environmental data sequence is obtained by the isolation forest algorithm; for the coal mine environmental data at any sampling time in the coal mine environmental data sequence, if the anomaly score of the coal mine environmental data at the sampling time is greater than or equal to T1, the coal mine environmental data at the sampling time is recorded as the target anomaly data of the coal mine environmental data sequence, and then all the target anomaly data of the coal mine environmental data sequence are obtained.
[0053] So far, all target abnormal data of the coal mine environmental data sequence have been obtained through the above method.
[0054] Step S003: Obtain an anomaly score correction factor for each target anomaly data according to the comprehensive anomaly confidence of each target anomaly data in the coal mine environment data sequence.
[0055] 1. Obtain the comprehensive anomaly confidence of each target anomaly data in the coal mine environmental data sequence.
[0056] It should be noted that for all target abnormal data in the coal mine environmental data sequence, there will be real abnormal data and noise data in the time series, which can be distinguished from the following two perspectives: real abnormal data often appear collectively, while noise data appear alone or in small quantities, so it can be judged by the relationship between adjacent abnormal data in the time series; the change trend of real abnormal data conforms to the law of physical change, and its change in time series is relatively gentle, while noise data is unstable and irregular, and its change in time series is usually more drastic, so it can be judged by the change trend of data in time series. The above two characteristics can be used to comprehensively derive the confidence of the data anomaly score.
[0057] It should be noted that when analyzing the relationship between target anomaly data, all target anomaly data need to be arranged in time series, and the degree of anomaly of the corresponding adjacent area is calculated by the difference and time interval between several target anomaly data that are close to it and their anomaly scores.
[0058] A parameter a is preset, wherein this embodiment is described by taking a=5 as an example, and this embodiment is not specifically limited, wherein a is determined according to specific implementation conditions.
[0059] Specifically, all target abnormal data of the coal mine environment data sequence are arranged according to the time sequence of the sampling time to obtain the target abnormal data sequence; for the i-th target abnormal data in the target abnormal data sequence, the data segment consisting of the first a target abnormal data of the i-th target abnormal data and the last a target abnormal data of the i-th target abnormal data is used as the adjacent abnormal data segment of the i-th target abnormal data; then the calculation method of the abnormal degree of the adjacent area of the i-th target abnormal data of the coal mine environment data sequence is:
[0060]
[0061] In the formula, represents the abnormal degree of the neighboring area of the ith target abnormal data in the coal mine environmental data sequence; t i represents the sampling time corresponding to the i-th target abnormal data in the coal mine environment data sequence; t i,j represents the sampling time corresponding to the jth target abnormal data in the adjacent abnormal data segment of the i-th target abnormal data in the coal mine environment data sequence; h i represents the abnormal score of the i-th target abnormal data in the coal mine environment data sequence; h i,j It represents the anomaly score of the jth target anomaly data in the adjacent anomaly data segment of the i-th target anomaly data in the coal mine environment data sequence; a is a preset parameter; || means taking the absolute value; norm() represents the linear normalization function.
[0062] Among them, if the number of all target abnormal data before the i-th target abnormal data is less than a, the number of all target abnormal data before the i-th target abnormal data is recorded as b, and the data segment consisting of the b target abnormal data before the i-th target abnormal data and the 2a-b target abnormal data after the i-th target abnormal data is used as the adjacent abnormal data segment of the i-th target abnormal data; if the number of all target abnormal data after the i-th target abnormal data is less than a, the number of all target abnormal data after the i-th target abnormal data is recorded as c, and the data segment consisting of the c target abnormal data after the i-th target abnormal data and the 2a-c target abnormal data before the i-th target abnormal data is used as the adjacent abnormal data segment of the i-th target abnormal data; the adjacent abnormal data segment of the i-th target abnormal data does not include the i-th target abnormal data.
[0063] At this point, the abnormality degree of the neighboring area of each target abnormal data in the coal mine environmental data sequence is obtained.
[0064] It should be noted that it is known that all target abnormal data are mixed with noise data, and the changes of noise data in time series are unstable and irregular, while the changes of real abnormal data in time series are relatively more stable and gentle. Therefore, we can derive the corresponding range of abnormality by the changing trend of target abnormal data in time series.
[0065] Specifically, for the kth target abnormal data in the coal mine environmental data sequence, the data segment consisting of the coal mine environmental data at the a sampling moments before the kth target abnormal data and the coal mine environmental data at the a sampling moments after the kth target abnormal data is used as the neighboring range data segment of the kth target abnormal data.
[0066] For the coal mine environmental data at the vth sampling moment in the adjacent range data segment of the kth target abnormal data, the absolute value of the difference between the coal mine humidity data at the vth sampling moment and the coal mine humidity data at the v+1th sampling moment is used as the first-order difference of humidity of the coal mine environmental data at the vth sampling moment; the absolute value of the difference between the first-order difference of humidity of the coal mine environmental data at the vth sampling moment and the first-order difference of humidity of the coal mine environmental data at the v+1th sampling moment is used as the second-order difference of humidity of the coal mine environmental data at the vth sampling moment; then the range abnormality degree calculation method of the kth target abnormal data of the coal mine environmental data sequence is:
[0067]
[0068] In the formula, Indicates the range of abnormality of the kth target abnormal data in the coal mine environmental data sequence; ΔH1 k,v Represents the first-order difference of humidity of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔH2 k,v It represents the second-order difference of humidity of the coal mine environmental data at the vth sampling moment in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; a is a preset parameter; norm() represents a linear normalization function.
[0069] Among them, referring to the method for obtaining the neighboring abnormal data segment of the i-th target abnormal data, the neighboring range data segment of the k-th target abnormal data is obtained; the neighboring range data segment of the k-th target abnormal data includes the k-th target abnormal data; the first-order difference of humidity represents the changing speed of humidity data, and the second-order difference of humidity represents the changing amount of the changing speed of humidity data.
[0070] Furthermore, the product of the abnormality degree of the neighboring area of each target abnormal data in the coal mine environmental data sequence and the abnormality degree of the range of each target abnormal data is taken as the comprehensive abnormality confidence of each target abnormal data.
[0071] At this point, the comprehensive anomaly confidence of each target anomaly data in the coal mine environmental data sequence is obtained.
[0072] 2. Obtain the anomaly score correction factor for each target anomaly data in the coal mine environmental data sequence.
[0073] It should be noted that since the change of humidity data is related to other environmental data such as temperature and gas concentration, the correction of the abnormal score not only needs to rely on the confidence of humidity, but also needs to conduct a comprehensive analysis of other dimensional data at the same time. In the same coal mine environment, humidity, temperature and gas concentration are interrelated and jointly affect the stability of the environment. Among them, humidity and temperature are usually inversely proportional. When the air humidity increases, the temperature decreases, and when the humidity decreases, the temperature increases; humidity and gas concentration are also inversely proportional. When the humidity increases, the air becomes more humid, the gas diffusion rate slows down, and the gas concentration decreases. Conversely, when the humidity decreases, the gas concentration increases. Therefore, when the humidity data is abnormal, it is often accompanied by changes in temperature and gas concentration.
[0074] Specifically, for the coal mine environmental data at the vth sampling moment in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence, the absolute value of the difference between the coal mine temperature data at the vth sampling moment and the coal mine temperature data at the v+1th sampling moment is used as the first-order temperature difference of the coal mine environmental data at the vth sampling moment; the absolute value of the difference between the first-order temperature difference of the coal mine environmental data at the vth sampling moment and the first-order temperature difference of the coal mine environmental data at the v+1th sampling moment is used as the second-order temperature difference of the coal mine environmental data at the vth sampling moment. ; The absolute value of the difference between the coal mine gas concentration data of the coal mine environmental data at the vth sampling moment and the coal mine gas concentration data of the coal mine environmental data at the v+1th sampling moment is taken as the first-order difference of the gas concentration of the coal mine environmental data at the vth sampling moment; The absolute value of the difference between the first-order difference of the gas concentration of the coal mine environmental data at the vth sampling moment and the first-order difference of the gas concentration of the coal mine environmental data at the v+1th sampling moment is taken as the second-order difference of the gas concentration of the coal mine environmental data at the vth sampling moment; Then the calculation method of the abnormal score correction factor of the kth target abnormal data in the coal mine environmental data sequence is:
[0075]
[0076]
[0077]
[0078] In the formula, δ krepresents the abnormal score correction factor of the kth target abnormal data in the coal mine environmental data sequence; Q k W represents the environmental data change speed of the kth target abnormal data in the coal mine environmental data sequence; k Indicates the environmental data change of the kth target abnormal data in the coal mine environmental data sequence; represents the comprehensive abnormal confidence of the kth target abnormal data in the coal mine environmental data sequence; ΔH1 k,v Represents the first-order difference of humidity of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔH2 k,v Represents the second-order difference of humidity of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔC1 k,v The first-order difference of the gas concentration of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔC2 k,v Indicates the second-order difference of gas concentration of coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔF1 k,v Indicates the first-order difference of the temperature of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔF2 k,v It represents the second-order difference of the temperature of the coal mine environmental data at the vth sampling moment in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; a is a preset parameter; || represents the absolute value; norm() represents the linear normalization function.
[0079] At this point, the anomaly score correction factor of each target anomaly data in the coal mine environmental data sequence is obtained.
[0080] Step S004: According to the abnormal score correction factor of each target abnormal data, all real abnormal data in the coal mine environment data sequence are obtained, and then the construction environment in the coal mine is monitored.
[0081] Two parameters N1 and N2 are preset, wherein this embodiment is described by taking N1=3 and N2=5 as an example, and this embodiment is not specifically limited, wherein N1 and N2 are determined according to specific implementation conditions.
[0082] Specifically, for any target abnormal data in the coal mine environmental data sequence, the product of the abnormality score of the target abnormal data and the abnormality score correction factor of the target abnormal data is used as the corrected abnormality score of the target abnormal data; if the corrected abnormality score of the target abnormal data is greater than T1, the target abnormal data is recorded as the real abnormal data, and then all the real abnormal data in the coal mine environmental data sequence are obtained.
[0083] Furthermore, the coal mine environmental data of any 20 consecutive sampling moments in the coal mine environmental data sequence are recorded as the target data segment of the coal mine environmental data sequence; if real abnormal data appear N1 times in a row in the coal mine environmental data sequence, or if there are more than N2 real abnormal data in any target data segment of the coal mine environmental data sequence, the alarm of the control system will issue a warning to remind the staff to check the construction environment in the coal mine in time.
[0084] Through the above steps, an intelligent early warning method for coal mine water hazard monitoring is completed.
[0085] The present invention also proposes a coal mine water hazard monitoring intelligent early warning system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any one of the steps of the coal mine water hazard monitoring intelligent early warning method when executing the computer program.
[0086] 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 coal mine water hazard monitoring intelligent early warning method, characterized in that: The method comprises the following steps: Acquire a coal mine environment data sequence, the coal mine environment data sequence includes coal mine environment data at a number of sampling moments, the coal mine environment data at each sampling moment refers to coal mine temperature data, coal mine humidity data and coal mine gas concentration data at each sampling moment; The anomaly score of the coal mine environmental data at each sampling moment is obtained by using the isolation forest algorithm, and the coal mine environmental data is screened according to the anomaly score to obtain all target anomaly data; Arrange all target abnormal data according to the time sequence of sampling time to obtain a target abnormal data sequence; obtain the adjacent abnormal data segment of each target abnormal data in the target abnormal data sequence; obtain the adjacent range data segment of each target abnormal data in the coal mine environment data sequence; obtain the comprehensive abnormal confidence of each target abnormal data according to the difference in abnormal scores between each target abnormal data and other target abnormal data in the adjacent abnormal data segment and the difference in coal mine humidity data between each target abnormal data and coal mine environment data at other sampling times in the adjacent range data segment; The method of obtaining the comprehensive abnormal confidence of each target abnormal data according to the abnormal score difference between each target abnormal data and other target abnormal data in the adjacent abnormal data segment and the coal mine humidity data difference between each target abnormal data and the coal mine environmental data at other sampling moments in the adjacent range data segment includes the following specific methods: According to the difference in abnormal scores between each target abnormal data and other target abnormal data in the adjacent abnormal data segment, the degree of abnormality in the adjacent area of each target abnormal data in the coal mine environmental data sequence is obtained; according to the difference in coal mine humidity data between each target abnormal data and the coal mine environmental data at other sampling times in the adjacent range data segment, the degree of range abnormality of each target abnormal data in the coal mine environmental data sequence is obtained; The product of the abnormal degree of the adjacent area of each target abnormal data in the coal mine environment data sequence and the abnormal degree of the range of each target abnormal data in the coal mine environment data sequence is taken as the comprehensive abnormal confidence of each target abnormal data; The specific formula for obtaining the abnormality degree of the adjacent area of each target abnormal data in the coal mine environment data sequence according to the abnormality score difference between each target abnormal data and other target abnormal data in the adjacent abnormal data segment is: In the formula, represents the abnormal degree of the neighboring area of the ith target abnormal data in the coal mine environmental data sequence; t i represents the sampling time corresponding to the i-th target abnormal data in the coal mine environment data sequence; t i,j represents the sampling time corresponding to the jth target abnormal data in the adjacent abnormal data segment of the i-th target abnormal data in the coal mine environment data sequence; i represents the abnormal score of the i-th target abnormal data in the coal mine environment data sequence; h i,j It represents the abnormal score of the jth target abnormal data in the adjacent abnormal data segment of the i-th target abnormal data in the coal mine environment data sequence; a is a preset parameter; | | represents taking the absolute value; norm() represents the linear normalization function; The specific method of obtaining the range abnormality degree of each target abnormal data in the coal mine environmental data sequence according to the difference in coal mine humidity data between each target abnormal data and the coal mine environmental data at other sampling times in the adjacent range data segment is as follows: For the coal mine environmental data at the vth sampling moment in the adjacent range data segment of the kth target abnormal data, the absolute value of the difference between the coal mine humidity data at the vth sampling moment and the coal mine humidity data at the v+1th sampling moment is used as the first-order difference of humidity of the coal mine environmental data at the vth sampling moment; the absolute value of the difference between the first-order difference of humidity of the coal mine environmental data at the vth sampling moment and the first-order difference of humidity of the coal mine environmental data at the v+1th sampling moment is used as the second-order difference of humidity of the coal mine environmental data at the vth sampling moment; then the range abnormality degree calculation method of the kth target abnormal data of the coal mine environmental data sequence is: In the formula, Indicates the range of abnormality of the kth target abnormal data in the coal mine environmental data sequence; ΔH1 k,v Represents the first-order difference of humidity of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔH2 k,v represents the second-order difference of humidity of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; a is a preset parameter; norm() represents a linear normalization function; According to the comprehensive abnormal confidence of each target abnormal data and the difference between each target abnormal data and the coal mine environmental data at other sampling moments in the adjacent range data segment, the abnormal score correction factor of each target abnormal data is obtained; The method of obtaining the abnormal score correction factor of each target abnormal data according to the comprehensive abnormal confidence of each target abnormal data and the difference between each target abnormal data and the coal mine environmental data at other sampling times in the adjacent range data segment includes the following specific methods: For the kth target abnormal data in the coal mine environmental data sequence, the environmental data change speed and environmental data change amount of the kth target abnormal data in the coal mine environmental data sequence are obtained; then the calculation method of the abnormal score correction factor of the kth target abnormal data in the coal mine environmental data sequence is: In the formula, δ k represents the abnormal score correction factor of the kth target abnormal data in the coal mine environmental data sequence; Q k W represents the environmental data change speed of the kth target abnormal data in the coal mine environmental data sequence; k Indicates the environmental data change of the kth target abnormal data in the coal mine environmental data sequence; represents the comprehensive abnormal confidence of the kth target abnormal data in the coal mine environmental data sequence; norm() represents the linear normalization function; The specific method of obtaining the environmental data change speed and environmental data change amount of the kth target abnormal data in the coal mine environmental data sequence includes: For the coal mine environmental data at the vth sampling moment in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence, the absolute value of the difference between the coal mine temperature data at the vth sampling moment and the coal mine temperature data at the v+1th sampling moment is used as the first-order temperature difference of the coal mine environmental data at the vth sampling moment; the absolute value of the difference between the first-order temperature difference of the coal mine environmental data at the vth sampling moment and the first-order temperature difference of the coal mine environmental data at the v+1th sampling moment is used as the second-order temperature difference of the coal mine environmental data at the vth sampling moment; The absolute value of the difference between the coal mine gas concentration data of the coal mine environmental data at the vth sampling moment and the coal mine gas concentration data of the coal mine environmental data at the v+1th sampling moment is taken as the first-order difference of the gas concentration of the coal mine environmental data at the vth sampling moment; the absolute value of the difference between the first-order difference of the gas concentration of the coal mine environmental data at the vth sampling moment and the first-order difference of the gas concentration of the coal mine environmental data at the v+1th sampling moment is taken as the second-order difference of the gas concentration of the coal mine environmental data at the vth sampling moment; then the calculation method of the environmental data change speed and environmental data change amount of the kth target abnormal data in the coal mine environmental data sequence is: Where, ΔH1 k,v Represents the first-order difference of humidity of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔH2 k,v Represents the second-order difference of humidity of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔC1 k,v The first-order difference of the gas concentration of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔC2 k,v Indicates the second-order difference of gas concentration of coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔF1 k,v Indicates the first-order difference of the temperature of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; ΔF2 k,v It represents the second-order difference of the temperature of the coal mine environmental data at the vth sampling time in the adjacent range data segment of the kth target abnormal data in the coal mine environmental data sequence; a is a preset parameter; | | represents the absolute value; The anomaly score of each target anomaly data is corrected according to the anomaly score correction factor, and the coal mine environmental data is screened according to the corrected anomaly score to obtain all the real anomaly data; according to the distribution of the real anomaly data in the coal mine environmental data sequence, the construction environment in the coal mine is monitored.
2. According to claim 1, a coal mine water hazard monitoring intelligent early warning method is characterized in that: The specific method of screening the coal mine environmental data according to the abnormality score to obtain all target abnormal data includes: A threshold T1 is preset. For the coal mine environment data at any sampling time in the coal mine environment data sequence, if the anomaly score of the coal mine environment data at the sampling time is greater than or equal to T1, the coal mine environment data at the sampling time is recorded as the target anomaly data of the coal mine environment data sequence, and then all the target anomaly data of the coal mine environment data sequence are obtained.
3. According to claim 1, a coal mine water hazard monitoring intelligent early warning method is characterized in that: The specific method of obtaining the adjacent abnormal data segments of each target abnormal data in the target abnormal data sequence includes: A parameter a is preset, and for the ith target abnormal data in the target abnormal data sequence, a data segment consisting of the first a target abnormal data of the ith target abnormal data and the last a target abnormal data of the ith target abnormal data is used as the adjacent abnormal data segment of the ith target abnormal data.
4. According to claim 1, a coal mine water hazard monitoring intelligent early warning method is characterized in that: The specific method of obtaining the adjacent range data segment of each target abnormal data in the coal mine environment data sequence includes: For the kth target abnormal data in the coal mine environmental data sequence, the data segment consisting of the coal mine environmental data at the a sampling moments before the kth target abnormal data and the coal mine environmental data at the a sampling moments after the kth target abnormal data is used as the adjacent range data segment of the kth target abnormal data, and a is a preset parameter.
5. A coal mine water hazard monitoring intelligent early warning system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the intelligent early warning method for coal mine water hazard monitoring as described in any one of claims 1-4 are implemented.
