Mine roof multi-source data fusion early warning method and device and storage medium

Abstract

The invention provides a mine roof multi-source data fusion early warning method and device and a storage medium, relates to the technical field of engineering safety, and solves the technical problem that it is difficult to realize mine roof instability early warning by fusing multi-source monitoring data in the prior art. The method specifically comprises the following steps: acquiring multi-source data; preprocessing the multi-source data; extracting a normalized characteristic value of the preprocessed multi-source data; basic probability distribution is constructed based on mine roof historical instability data, the D-S evidence theory is adopted to fuse the normalized feature values, and the early warning confidence degree of each early warning level is output; and determining the early warning level of the mine roof based on the early warning confidence. The method is used for mine roof safety early warning.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of engineering safety, and in particular to a mine roof multi-source data fusion early warning method and device and a storage medium. BACKGROUND

[0002] Mine roof instability is a major safety hazard in coal mine production, and accurate early warning is crucial to protect the lives of underground workers and production stability. At present, mine roof early warning mostly uses single type sensors to collect stress, displacement or crack related data, or fuses a small amount of monitoring data through simple threshold comparison, traditional weighted average and other methods, and some schemes rely on manual inspection and static mechanical models for risk judgment. The existing technology generally has the defects of incomplete monitoring dimension, lack of effective processing of data conflicts in multi-source data fusion method, and insufficient accuracy due to the dependence of early warning judgment on experience threshold, and it is difficult to dynamically capture the precursor signals of roof instability. Therefore, the existing technology has the technical problem that it is difficult to realize mine roof instability early warning by fusing multi-source monitoring data. SUMMARY

[0003] The present application provides a mine roof multi-source data fusion early warning method, device and storage medium, which solves the technical problem that the existing technology is difficult to realize mine roof instability early warning by fusing multi-source monitoring data.

[0004] To achieve the above purpose, the present application adopts the following technical solutions: In a first aspect, a mine roof multi-source data fusion early warning method is provided, comprising: acquiring multi-source data; the multi-source data includes roof stress, roof subsidence, roof subsidence rate, crack quantity and crack opening; preprocessing the multi-source data; the preprocessing includes replacing abnormal values and supplementing missing values; extracting normalized characteristic values of the preprocessed multi-source data; the normalized characteristic values include normalized stress mutation rate, stress mean value, subsidence acceleration, subsidence cumulative displacement, opening growth rate and crack quantity increment; constructing a basic probability distribution based on historical instability data of the mine roof, and fusing the normalized characteristic values using D-S evidence theory to output early warning confidence of each early warning level; determining the early warning level of the mine roof based on the early warning confidence.

[0005] In combination with the first aspect described above, in a possible implementation, the preprocessing of the multi-source data includes: replacing abnormal values in the multi-source data by domain median; and supplementing missing values in the multi-source data by time series linear interpolation.

[0006] In a possible implementation manner of the first aspect, the normalized characteristic values of the preprocessed multi-source data are extracted, including: setting a characteristic window length and a collection interval; calculating a stress mutation rate and a stress average value based on preprocessed roof stress data; calculating a subsidence acceleration and a cumulative displacement based on preprocessed roof subsidence and roof subsidence rate data; calculating an opening growth rate and a crack number increment based on preprocessed crack opening and crack number data; performing Min-Max normalization processing on the characteristic values to map to the interval [0, 1] to obtain the normalized characteristic values.

[0007] In a possible implementation manner of the first aspect, the basic probability distribution is constructed based on the historical instability data of the mine roof, and the Dempster evidence theory is used to fuse the normalized characteristic values to output the early warning confidence of each early warning level, including: constructing a recognition framework with the early warning levels of the mine roof as elements; the early warning levels include a first level, a second level, a third level and a fourth level; based on the historical instability data of the mine roof, a mapping relationship between the normalized characteristic value interval and each early warning level is established; based on the normalized characteristic values and the mapping relationship, an m function of each characteristic value is obtained; the m functions of each characteristic value are fused by using the Dempster combination rule to output the early warning confidence corresponding to each early warning level after fusion.

[0008] In a possible implementation manner of the first aspect, the fusion of the m functions of each characteristic value by using the Dempster combination rule satisfies the following formula: wherein, is the early warning confidence of the early warning level A, K is an evidence conflict coefficient, is the basic probability distribution corresponding to the stress characteristic supported early warning level, is the basic probability distribution corresponding to the displacement characteristic supported early warning level, is the basic probability distribution corresponding to the crack characteristic supported early warning level.

[0009] In a possible implementation manner of the first aspect, the first level corresponds to the I-level early warning of the mine roof, the second level corresponds to the II-level early warning, the third level corresponds to the III-level early warning, and the fourth level corresponds to the IV-level early warning; the severity of the first level, the second level, the third level and the fourth level is sequentially increasing.

[0010] In a possible implementation manner of the first aspect, the method further includes: determining the pre-warning level of the mine roof based on the pre-warning confidence, including: taking the maximum value of the pre-warning confidence corresponding to each pre-warning level as a preliminary determination level, and if the acceleration of subsidence or the stress mutation rate exceeds the preset characteristic threshold, increasing the preliminary determination level by one level, and if the maximum pre-warning confidence is lower than the preset confidence threshold, decreasing the preliminary determination level by one level, and taking the preliminary determination level after the increase or decrease as the pre-warning level of the mine roof.

[0011] In a possible implementation manner of the first aspect, after determining the pre-warning level of the mine roof based on the pre-warning confidence, the method further includes: triggering the corresponding level of the in-situ sound and light alarm in the mine based on the pre-warning level, and pushing the pre-warning information to the mobile phone terminal, and the pre-warning information includes the pre-warning level, the multi-source data, the characteristic value and the fusion confidence, and the pre-warning information is supplemented to the historical instability database of the mine roof.

[0012] In a possible implementation manner of the first aspect, after determining the pre-warning level of the mine roof based on the pre-warning confidence, the method further includes: triggering the corresponding level of the in-situ sound and light alarm in the mine based on the pre-warning level, and pushing the pre-warning information to the mobile phone terminal, and the pre-warning information includes the pre-warning level, the multi-source data, the characteristic value and the fusion confidence, and the pre-warning information is supplemented to the historical instability database of the mine roof.

[0013] In a possible implementation manner of the first aspect, after determining the pre-warning level of the mine roof based on the pre-warning confidence, the method further includes: triggering the corresponding level of the in-situ sound and light alarm in the mine based on the pre-warning level, and pushing the pre-warning information to the mobile phone terminal, and the pre-warning information includes the pre-warning level, the multi-source data, the characteristic value and the fusion confidence, and the pre-warning information is supplemented to the historical instability database of the mine roof.

[0014] The application provides a mine roof multi-source data fusion early warning method, device and storage medium, which can accurately fuse multi-dimensional monitoring data of roof stress, displacement and cracks, improves data effectiveness through preprocessing, efficiently processes multi-source data conflicts with the aid of D-S evidence theory, combines instability sensitive feature extraction and hierarchical early warning judgment rules, realizes accurate hierarchical early warning of mine roof instability, effectively reduces false alarm rate and missed alarm rate, and solves the technical problem that the prior art cannot realize mine roof instability early warning by fusing multi-source monitoring data. Meanwhile, the multi-channel early warning information pushing can help underground operation personnel to have sufficient emergency disposal time, and the historical instability database can be dynamically updated to continuously optimize the adaptability of the early warning model, significantly improve the reliability and pertinence of mine roof safety prevention and control, and ensure underground operation safety and production stability.

[0015] It should be understood that the description of technical features, technical solutions, beneficial effects or similar language in the present application does not imply that all features and advantages can be achieved in any single embodiment. On the contrary, it can be understood that the description of a feature or beneficial effect means that the specific technical feature, technical solution or beneficial effect is included in at least one embodiment. Therefore, the description of technical features, technical solutions or beneficial effects in the specification does not necessarily refer to the same embodiment. Further, the technical features, technical solutions and beneficial effects described in the embodiments can be combined in any appropriate manner. Those skilled in the art will understand that the embodiments can be implemented without one or more specific technical features, technical solutions or beneficial effects of the specific embodiments. In other embodiments, additional technical features and beneficial effects can be identified in specific embodiments that do not embody all embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 A flowchart of a mine roof multi-source data fusion early warning method provided by the embodiments of the present application is shown in the figure; Figure 2 A flowchart of another mine roof multi-source data fusion early warning method provided by the embodiments of the present application is shown in the figure; Figure 3 A flowchart of another mine roof multi-source data fusion early warning method provided by the embodiments of the present application is shown in the figure; Figure 4 A flowchart of another mine roof multi-source data fusion early warning method provided by the embodiments of the present application is shown in the figure; Figure 5 A structural diagram of a mine roof multi-source data fusion early warning device provided by the embodiments of the present application is shown in the figure. DETAILED DESCRIPTION

[0017] In the description of the present application, unless otherwise specified, " / " means "or", for example, A / B can mean A or B. "And / or" herein is only a description of the relationship between the associated objects, which means that there can be three relationships, for example, A and / or B, which can mean: A exists alone, A and B exist together, and B exists alone. In addition, "at least one" means one or more, and "multiple" means two or more. "First", "second", and the like do not limit the quantity and execution order, and "first", "second", and the like do not necessarily mean different.

[0018] It should be noted that in the present application, "exemplary" or "for example" is used to mean example, illustration or description. Any embodiment or design scheme described as "exemplary" or "for example" in the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the use of "exemplary" or "for example" is intended to present the relevant concept in a specific manner.

[0019] To solve the technical problem that it is difficult to realize mine roof instability early warning by fusing multi-source monitoring data in the prior art, the present application provides a mine roof multi-source data fusion early warning method, which comprises: acquiring multi-source data of roof stress, roof subsidence, roof subsidence rate, crack quantity and crack opening; pre-processing the multi-source data by replacing abnormal values and supplementing missing values; extracting normalized characteristic values of the pre-processed data; constructing basic probability distribution based on historical instability data of the mine roof and fusing the normalized characteristic values by D-S evidence theory to output early warning confidence of each early warning level; and finally determining the early warning level of the mine roof based on the early warning confidence. Based on this, the complementary information of multi-source monitoring data can be effectively integrated, the data quality and fusion accuracy can be improved, and reliable early warning of mine roof instability can be realized, thereby providing effective protection for underground operation safety.

[0020] As shown in Figure 1 The mine roof multi-source data fusion early warning method provided by the present application comprises: S101, acquiring multi-source data.

[0021] The multi-source data refers to multi-dimensional monitoring data that can reflect the stability state of the mine roof, and specifically includes roof stress, roof subsidence, roof subsidence rate, crack quantity and crack opening.

[0022] In the present application, the multi-source data can be collected by sensors deployed at key monitoring positions underground. The sensor type can be selected according to the mine geological conditions and monitoring requirements, and the collection range covers core dimensions such as roof mechanical state, spatial deformation and structural damage.

[0023] It should be pointed out that data acquisition needs to adapt to the complex environment underground, ensure that the sensor has anti-interference, temperature and humidity resistance and other characteristics, and protect the stability of data transmission.

[0024] As an example, in a certain coal mining face, the roof stress sensor is arranged on the roof to collect the roof stress, the roof displacement monitor collects the roof subsidence amount and the roof subsidence rate, and the borehole imaging instrument collects the crack quantity and the crack opening.

[0025] Based on the above steps, the key data related to the stability of the roof can be comprehensively obtained, and complete data support is provided for subsequent multi-source fusion analysis.

[0026] S102, pre-processing of multi-source data.

[0027] Among them, pre-processing refers to the process of optimizing the collected raw data, including abnormal value replacement and missing value supplement. Abnormal value refers to a value that exceeds the reasonable data range corresponding to the mine geological conditions, and missing value refers to blank data caused by collection interval or equipment failure.

[0028] In the embodiment of the application, the abnormal value processing can adopt the neighborhood median replacement method suitable for the characteristics of underground data, and the missing value supplement can adopt the time series linear interpolation method suitable for the gradual change characteristics of underground data, to ensure that the processed data fit the actual monitoring scene.

[0029] As an example, the neighborhood median replacement satisfies the following formula: And / or, the time series linear interpolation method satisfies the following formula: Wherein, k is the time domain window, x(t) is the original value at time t, is the corrected value, is the effective data before missing, is the effective data after missing, is the linear interpolation.

[0030] Based on the above steps, the noise and blank in the original data can be effectively eliminated, and the accuracy and reliability of the subsequent feature extraction link can be ensured.

[0031] S103, extracting the normalized feature value of the pre-processed multi-source data.

[0032] Among them, the normalized feature value is the standardized data of mapping the original feature data to a unified interval, including stress mutation rate, stress average, subsidence acceleration, subsidence cumulative displacement, opening growth rate and crack quantity increment, which respectively represent the roof instability precursor from the change amplitude, overall state, deformation trend and other dimensions.

[0033] In the embodiments of the present application, the feature extraction needs to set the feature window length and the collection interval adaptive to the mine monitoring frequency, and then calculate the stress mutation rate and the stress average value based on the preprocessed roof stress data of S102; calculate the sinking acceleration and the cumulative displacement of the sinking based on the preprocessed roof subsidence and the roof subsidence rate data; calculate the opening growth rate and the crack number increment based on the preprocessed crack opening and the crack number data; perform Min-Max normalization processing on all feature values, map them to the interval [0, 1], and obtain the normalized feature values.

[0034] Based on the above steps, the original monitoring data can be converted into sensitive features reflecting the roof instability state, and the data representation ability for risk changes can be improved.

[0035] S104, based on the historical instability data of the mine roof, a basic probability distribution is constructed, and the D-S evidence theory is used to fuse the normalized feature values, and the warning confidence of each warning level is output.

[0036] Among them, the basic probability distribution is the confidence mapping relationship between the normalized feature values and the warning levels based on the historical data, the D-S evidence theory is a multi-source data fusion method that can effectively handle uncertain information, and the warning confidence is the probability value supporting each warning level after fusion.

[0037] In the embodiments of the present application, the basic probability distribution needs to be calibrated by the correlation between the feature values and the warning levels in the historical instability cases of the target mine, and the fusion process needs to first construct an identification framework containing each warning level, and then use the evidence combination rule to process the conflict information of multiple source features.

[0038] As an example, based on the roof instability data of a mine for 5 years, the mapping relationship between the feature value interval and the I-IV level warning is established, the m function of the stress, displacement and crack features is obtained, and after fusion using the Dempster combination rule, the warning confidence of each level is output.

[0039] Based on the above steps, the complementary information of multiple source features can be integrated, the credibility of each warning level can be quantified, and the uncertainty of single feature analysis can be reduced.

[0040] S105, based on the warning confidence, the warning level of the mine roof is determined.

[0041] Among them, the warning level is a grading result reflecting the risk degree of the mine roof instability, including I level (no warning), II level (pay attention to warning), III level (general warning), and IV level (major warning), corresponding to different site disposal requirements.

[0042] In the embodiments of the present application, the pre-warning level determination needs to take the pre-warning confidence as the core basis, and can also be verified in combination with the mutation of the key features of the roof instability, and the determination rule needs to adapt to the timeliness requirement of the underground emergency disposal.

[0043] As an example, the pre-warning level is determined based on the pre-warning confidence by the maximum confidence principle, the mutation strengthening rule and the threshold verification.

[0044] Based on the above steps, the quantified confidence can be converted into an explicit pre-warning level, providing a direct and accurate decision basis for the safety disposal of underground operations.

[0045] Based on the above technical solution, a closed loop is formed from multi-source data acquisition to pre-warning level output, which not only realizes the comprehensive coverage and quality optimization of stress, displacement and fracture data, but also effectively integrates the complementary information of each dimension data through sensitive feature extraction and D-S evidence theory fusion, solves the one-sidedness problem of single data pre-warning, significantly improves the accuracy and reliability of the mine roof instability pre-warning, and solves the technical problem that the existing technology cannot realize the mine roof instability pre-warning by fusing multi-source monitoring data.

[0046] In a possible implementation manner of the embodiments of the present application, the pre-warning confidence is determined based on the stress, displacement and fracture data of the mine roof. Figure 1 As shown in Figure 2 The above S104 can be implemented by the following S201 to S204, which will be described in detail below: S201, constructing an identification framework with the pre-warning level of the mine roof as an element.

[0047] The identification framework is a set used to define the decision target in the D-S evidence theory, and the pre-warning level includes a first level, a second level, a third level and a fourth level, which correspond to I level, II level, III level and IV level respectively, and the severity increases in turn.

[0048] In the embodiments of the present application, when constructing the identification framework, the core risk features of each level need to be determined to ensure that the level division is consistent with the actual risk evolution process of the mine roof instability, and corresponds to the requirements of the underground emergency disposal.

[0049] As an example, the identification framework is constructed as ={I level, II level, III level, IV level}, wherein the I level corresponds to the stable state of the roof, and the IV level corresponds to the extremely high risk state of instability.

[0050] Based on the above steps, the decision target of multi-source data fusion can be determined, providing a clear judgment dimension for subsequent confidence calculation.

[0051] S202, based on the historical instability data of the mine roof, a mapping relationship between the normalized feature value interval and each pre-warning level is established.

[0052] wherein the mapping relationship refers to the association rule between different normalized characteristic value intervals and each early warning level, and is used to quantify the support degree of the characteristic value to the risk level.

[0053] In the embodiment of the present application, the occurrence frequency of different early warning levels in each normalized characteristic value interval in the historical instability data of the target mine is counted, and the association strength is determined based on the frequency ratio to form a standardized mapping rule table.

[0054] As an example, based on the 5-year historical data of a mine, the mapping rule is established: the normalized characteristic value [0, 0.2] corresponds to I-level early warning, (0.2, 0.4] corresponds to II-level early warning, (0.4, 0.7] corresponds to III-level early warning, and (0.7, 1.0] corresponds to IV-level early warning.

[0055] Based on the above steps, the abstract characteristic value can be converted into a clear risk association basis, and the characteristic and early warning level can be quantitatively linked.

[0056] S203, obtaining the m function of each characteristic value based on the normalized characteristic value and the mapping relationship.

[0057] wherein the m function is a basic probability assignment function, which is used to represent the support confidence of a single characteristic value to each early warning level, and meets the fusion requirements of the Dempster combination rule, and its calculation needs to match the uncertainty characteristics of multi-source data.

[0058] In the embodiment of the present application, the confidence of each early warning level is assigned according to the interval to which the normalized characteristic value belongs and the association strength in the mapping relationship, the sum of the confidence is 1, and a small amount of uncertainty can be reserved to deal with data deviation.

[0059] As an example, the normalized value of the stress characteristic at a certain moment is 0.879 (belongs to the interval (0.7, 1.0]), according to the mapping relationship, m1(I)=0.05, m1(II)=0.05, m1(III)=0.0, and m1(IV)=0.9, wherein m1 is the m function of the stress characteristic.

[0060] Based on the above steps, the characteristic information can be converted into a fusionable probabilistic data, and the uncertainty characteristics of multi-source data are retained.

[0061] S204, the m functions of each characteristic value are fused by using the Dempster combination rule, and the early warning confidence corresponding to each early warning level after fusion is output.

[0062] wherein the Dempster combination rule is the core fusion algorithm of the D-S evidence theory, which processes the contradictory information of multi-source data by calculating the evidence conflict coefficient, and finally integrates the unified confidence result.

[0063] In the embodiments of the present application, the conflict coefficient K of the three types of characteristic m functions of stress, displacement and fracture is calculated first, if K is less than a preset threshold, then direct fusion is performed, if K is too large, then fusion is performed after adjusting the weight, so as to ensure that the fusion result is reliable.

[0064] Preferably, the m functions of each characteristic value are fused by using the Dempster combination rule, which satisfies the following formula: wherein, is the early warning confidence of the early warning level A (I, II, III, IV), K is the evidence conflict coefficient, is the basic probability assignment corresponding to the stress characteristic supported early warning level, is the basic probability assignment corresponding to the displacement characteristic supported early warning level, is the basic probability assignment corresponding to the fracture characteristic supported early warning level.

[0065] It should be pointed out that the summation sign in the formula is the addition operation of the product of the basic probability assignment functions, which is one of the core steps of the D-S evidence theory fusion of multi-source data, and is used to satisfy the intersection of the early warning levels A supported by the stress, displacement and fracture characteristics respectively , , equal to the target early warning level A.

[0066] As an example, based on the m1 (stress), m2 (displacement) and m3 (fracture) function values obtained in S203, the conflict coefficient K = 0.1175, and after fusion according to the formula, m(I) = 0.0, m(II) = 0.088, m(III) = 0.008, and m(IV) = 0.904 are output.

[0067] Based on the above steps, the complementary information of multi-source characteristics can be integrated, the one-sidedness of a single characteristic can be eliminated, and the comprehensiveness and reliability of the early warning confidence can be improved.

[0068] Based on the above technical scheme, a complete multi-source data fusion link is formed from the identification framework construction to the m function fusion, the quantitative correlation between the characteristics and the risk is established by means of the historical data, the data conflict is effectively processed by using the Dempster combination rule, and finally the output early warning confidence can accurately reflect the actual risk state of the roof, thereby providing a reliable quantitative basis for the subsequent early warning level determination.

[0069] In a possible implementation of the embodiment of the present application, in combination with Figure 1 As shown in Figure 3 S105 can be specifically implemented through the following S301 to S304, which will be described in detail below: S301, take the maximum value in the early warning confidence corresponding to each early warning level as the preliminary determination level corresponding to the early warning confidence.

[0070] Among them, the maximum early warning confidence refers to the highest value in the confidence value corresponding to the I-IV level early warning level after multi-source data fusion, and the preliminary determination level is the basic early warning level based on the confidence dominance.

[0071] In the embodiment of the present application, the maximum value is filtered by traversing the confidence data of all early warning levels.

[0072] As an example, the confidence of each level after fusion S204 is m(I)=0.0, m(II)=0.088, m(III)=0.008, and m(IV)=0.904. The maximum value 0.904 corresponding to the IV level is taken as the preliminary determination level.

[0073] Based on the above steps, the core risk level jointly supported by multi-source data can be quickly locked, and the pertinence and conciseness of the determination logic are ensured.

[0074] S302, if the subsidence acceleration or stress mutation rate exceeds the preset feature threshold, the preliminary determination level is promoted by one level.

[0075] Among them, the feature threshold is a critical value set based on the historical instability data of the mine and the geomechanical parameters, which is used to identify the dangerous precursor of the rapid deformation of the roof. The subsidence acceleration and stress mutation rate are sensitive features reflecting the instability trend of the roof.

[0076] In the embodiment of the present application, if the normalized feature values of the two parameters of the subsidence acceleration or stress mutation rate exceed the preset feature threshold, the early warning level is promoted by one level. When the level is promoted, the upper limit needs to be limited, and the highest level does not exceed IV.

[0077] As an example, the subsidence acceleration feature threshold of a certain mine is set to 2 mm / h 2 If the preliminary determination level is III, and the measured subsidence acceleration is 2.3 mm / h 2 , it is promoted to IV; if the preliminary is IV, it remains unchanged.

[0078] Based on the above steps, the roof burst instability signal can be captured, the lag caused by pure confidence determination can be avoided, and the timeliness of the early warning can be improved.

[0079] S303, if the maximum early warning confidence is lower than the preset confidence threshold, the preliminary determination level is reduced by one level.

[0080] The confidence threshold is a critical standard for measuring the reliability of the early warning confidence, and is used to filter high-level early warning results with insufficient reliability. The confidence threshold is usually set based on the average confidence level of mine multi-source data fusion.

[0081] In the embodiments of the present application, the confidence threshold can be dynamically adjusted according to the data quality. If the maximum early warning confidence is lower than the preset confidence threshold, the early warning level is reduced by one level. When the level is reduced, the lower limit needs to be limited, and the minimum level should not be lower than level I.

[0082] It should be pointed out that this rule is only for the maximum confidence, and does not affect the reference value of other levels of confidence, ensuring the rigor of the determination.

[0083] As an example, a mine sets the confidence threshold to 0.6. If the preliminary determination level is level II and the maximum early warning confidence is 0.55, it is reduced to level I. If the preliminary level is I, it remains unchanged.

[0084] Based on the above steps, the misjudgment early warning with low reliability can be eliminated, unnecessary production interruption can be reduced, and the practicability of the early warning system can be improved.

[0085] S304, the preliminary determination level after being raised or reduced is used as the early warning level of the mine roof.

[0086] The final early warning level is a clear risk level formed after double verification, and is directly used to guide the execution of emergency disposal and safety control measures in the underground site.

[0087] In the embodiments of the present application, the results of S301-S303 are integrated. If the upgrade and downgrade rules are triggered at the same time, the upgrade logic is preferentially executed to prevent major risks. If no verification trigger occurs, the preliminary determination level is used.

[0088] Based on the above steps, accurate and executable early warning results can be output, and clear decision basis can be provided for safety control of underground operation.

[0089] Based on the above technical solutions, the maximum confidence is taken as the core and guide, combined with sensitive feature mutation verification and confidence reliability screening, the accurate determination of the early warning level of the mine roof is realized. The one-sidedness of single determination logic is avoided, and the timeliness and reliability of the early warning are balanced. The safety control demand under the complex geological conditions in the underground is adapted.

[0090] In a possible implementation manner, the sensitive feature mutation verification and the confidence reliability screening are combined. Figure 1 For example, Figure 4As shown, after S105, the mine roof multi-source data fusion early warning method provided in the embodiments of the present application further includes the following S401 and S402: S401, based on the early warning level, triggering the corresponding level of underground sound and light alarm, and pushing the early warning information to the mobile terminal.

[0091] Among them, the corresponding level of sound and light alarm refers to the differentiated alarm mode matched with I-IV level early warning, and the early warning information includes early warning level, multi-source data, characteristic value and fusion confidence.

[0092] In the embodiments of the present application, the underground sound and light alarm device is deployed in personnel-intensive areas such as working face, roadway intersection, etc., and the alarm volume, light color and flicker frequency are strengthened with the upgrade of early warning level, and the mobile terminal pushes the information to the management personnel, on-site operating personnel and technical person in charge.

[0093] As an example, when triggering IV level early warning, the underground red high-frequency flashing sound and light alarm (volume ≥ 85dB) is started, and the mobile terminal pushes the SMS "302 working face IV level early warning, subsidence acceleration exceeds the standard, immediately evacuate operating personnel".

[0094] Based on the above steps, the risk information can be quickly transmitted to ensure that relevant personnel know and start emergency disposal at the first time, and the probability of accident occurrence is reduced.

[0095] S402, supplementing the early warning information to the mine roof historical instability database.

[0096] In the embodiments of the present application, the data supplement adopts an automatic synchronization mode, and is stored in the mine roof historical instability database according to "time-area-early warning information".

[0097] Based on the above steps, the historical data samples can be continuously enriched to provide real scene data support for subsequent basic probability distribution function optimization and early warning model iteration.

[0098] Based on the above technical solutions, the rapid response of underground operating personnel to risks is ensured, and the scene adaptability and accuracy of the early warning system are continuously improved through data accumulation, further strengthening the mine roof safety prevention and control capability.

[0099] The scheme of the embodiments of the present application is introduced from the perspective of device implementation. It can be understood that each device, for example, the mine roof multi-source data fusion early warning device, includes at least one of the corresponding hardware structure and software module for executing each function. Those skilled in the art should easily realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized in the form of hardware or a combination of hardware and computer software. Whether a certain function is executed in the form of hardware or computer software driven hardware depends on the specific application and design constraints of the technical solution. The skilled person can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0100] The embodiments of the present application can divide the functional units of the mine roof multi-source data fusion early warning device according to the above method examples. For example, each functional unit can be divided according to each function, or two or more functions can be integrated in one processing unit. The integrated unit can be realized in the form of hardware or software functional unit. It should be noted that the division of units in the embodiments of the present application is illustrative, and is only a logical functional division. Actual implementation can have another division method.

[0101] In the case of using integrated units, Figure 5 A possible structure diagram of the mine roof multi-source data fusion early warning device (denoted as mine roof multi-source data fusion early warning device 50) involved in the above embodiments is shown, which includes a processing unit 501 and a communication unit 502, and can also include a storage unit 503. Figure 5 The structure diagram shown can be used to illustrate the structure of the mine roof multi-source data fusion early warning device involved in the above embodiments.

[0102] When Figure 5 When the structure diagram shown is used to illustrate the structure of the mine roof multi-source data fusion early warning device involved in the above embodiments, the processing unit 501 is used to control and manage the actions of the mine roof multi-source data fusion early warning device, the communication unit 502 is used for communication between the mine roof multi-source data fusion early warning device and other devices, and the storage unit 503 is used to store the program code and data of the mine roof multi-source data fusion early warning device.

[0103] For example, the communication unit 502 is used to obtain multi-source data; the multi-source data includes roof stress, roof subsidence amount, roof subsidence rate, crack number and crack opening; The processing unit 501 is configured to preprocess the multi-source data, including replacing abnormal values, supplementing missing values, extracting normalized characteristic values of the preprocessed multi-source data, and fusing the normalized characteristic values based on historical instability data of the mine roof to output early warning confidence of each early warning level.

[0104] In a possible implementation, the processing unit 501 is further configured to preprocess the multi-source data, including replacing abnormal values in the multi-source data by a median in the field, and supplementing missing values in the multi-source data by a time series linear interpolation.

[0105] In a possible implementation, the processing unit 501 is further configured to extract normalized characteristic values of the preprocessed multi-source data, including setting a characteristic window length and a collection interval, calculating a stress mutation rate and a stress average based on preprocessed roof stress data, calculating a subsidence acceleration and a subsidence cumulative displacement based on preprocessed roof subsidence data and roof subsidence rate data, calculating an opening growth rate and a crack number increment based on preprocessed crack opening and crack number data, and performing Min-Max normalization processing on the characteristic values to map to the interval [0, 1] to obtain the normalized characteristic values.

[0106] In a possible implementation, the processing unit 501 is further configured to construct a basic probability distribution based on historical instability data of the mine roof, fuse the normalized characteristic values by using D-S evidence theory, and output early warning confidence of each early warning level, including constructing a recognition framework with early warning levels of the mine roof as elements, the early warning levels including a first level, a second level, a third level, and a fourth level, establishing a mapping relationship between normalized characteristic value intervals and each early warning level based on the historical instability data of the mine roof, obtaining m functions of each characteristic value based on the normalized characteristic values and the mapping relationship, and fusing the m functions of each characteristic value by using a Dempster combination rule to output early warning confidence corresponding to each early warning level after fusion.

[0107] In a possible implementation, fusing the m functions of each characteristic value by using the Dempster combination rule satisfies the following formula: wherein, is the early warning confidence of the early warning level A, K is an evidence conflict coefficient, is a basic probability distribution corresponding to a stress characteristic supported early warning level, is a basic probability distribution corresponding to a displacement characteristic supported warning levels, basic probability distribution corresponding to the crack feature supported warning levels.

[0108] In a possible implementation, the first level corresponds to a first-level warning of the mine roof, the second level corresponds to a second-level warning, the third level corresponds to a third-level warning, and the fourth level corresponds to a fourth-level warning; the first level, the second level, the third level, and the fourth level are in an order of increasing severity.

[0109] In a possible implementation, the processing unit 501 is further configured to determine a warning level of the mine roof based on the warning confidence, including: taking a maximum value in the warning confidence corresponding to each warning level as a preliminary determination level, and taking a warning level corresponding to the maximum value as the preliminary determination level; if the subsidence acceleration or the stress mutation rate exceeds a preset feature threshold, increasing the preliminary determination level by one level; the preliminary determination level does not exceed the fourth level; if the maximum warning confidence is lower than a preset confidence threshold, decreasing the preliminary determination level by one level; and taking the preliminary determination level after the increase or decrease as the warning level of the mine roof.

[0110] In a possible implementation, the processing unit 501 is further configured to, after determining the warning level of the mine roof based on the warning confidence, trigger a corresponding level of underground site sound and light alarm based on the warning level, and push a warning information to a mobile phone terminal; the warning information includes the warning level, the multi-source data, the feature value, and the fusion confidence; and the warning information is supplemented to a mine roof historical instability database.

[0111] The processing unit 501 can be a processor or a controller, and the communication unit 502 can be a communication interface, a transceiver, a transceiver, a transceiver circuit, a transceiver device, etc. The communication interface is a general term and can include one or more interfaces. The storage unit 503 can be a memory. When the mine roof multi-source data fusion warning device 50 is a chip, the processing unit 501 can be a processor or a controller, and the communication unit 502 can be an input interface and / or an output interface, a pin, or a circuit, etc. The storage unit 503 can be a storage unit (for example, a register, a cache, etc.) in the chip, or can be a storage unit (for example, a read-only memory (ROM), a random access memory (RAM), etc.) located outside the chip.

[0112] The communication unit can also be referred to as a transceiver unit. The antenna and the control circuit with transceiving functions in the mine roof multi-source data fusion early warning device 50 can be regarded as a communication unit 502 of the mine roof multi-source data fusion early warning device 50, and the processor with processing functions can be regarded as a processing unit 501 of the mine roof multi-source data fusion early warning device 50. Optionally, the device for realizing the receiving function in the communication unit 502 can be regarded as a communication unit, which is used for executing the receiving steps in the embodiments of the present application, and the communication unit can be a receiver, a receiver, a receiving circuit, etc. The device for realizing the sending function in the communication unit 502 can be regarded as a sending unit, which is used for executing the sending steps in the embodiments of the present application, and the sending unit can be a transmitter, a sender, a sending circuit, etc.

[0113] Figure 5 The integrated units in the above embodiments can be stored in a computer readable storage medium if they are realized in the form of software function modules and sold or used as independent products. Based on such understanding, the technical solutions of the embodiments of the present application or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the methods described in the embodiments of the present application. The storage medium for storing the computer software product includes a U disk, a mobile hard disk, a read-only memory, a random access memory, a magnetic disk or an optical disk, and various media that can store program codes.

[0114] Figure 5 The units in the above embodiments can also be referred to as modules, for example, the processing unit can be referred to as a processing module.

[0115] The embodiments of the present application also provide a computer readable storage medium, including instructions, when running on a computer, causing the computer to execute any of the above methods.

[0116] In the above embodiments, all or part of the embodiments can be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present application are generated. The computer can be a general purpose computer, a special purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through wired (such as coaxial cable, optical fiber, digital subscriber line (digital subscriber line, DSL)) or wireless (such as infrared, wireless, microwave, etc.) mode. The computer readable storage medium can be any available medium that can be accessed by a computer or data storage device including one or more servers, data centers, etc. integrated with the medium. The available medium can be magnetic medium (such as floppy disk, hard disk, magnetic tape), optical medium (such as DVD), or semiconductor medium (such as solid state disk (solid state disk, SSD)) and the like.

[0117] Although the present application is described herein in conjunction with various embodiments, other variations and modifications of the disclosed embodiments are possible in light of this disclosure, and it is intended that the scope of the application encompass such variations and modifications as can come within the scope of the claims. In the claims, the term "comprising" does not exclude other elements or steps, the term "one" or "a" does not exclude a plurality, and the term "one" or "a" means "one or more but can include only one". Individual features or units can implement several functions of several claim items. Measures described in mutually different dependent claims can be combined and implemented to produce desirable results.

[0118] Although the present application is described herein in conjunction with various embodiments, other variations and modifications of the disclosed embodiments are possible in light of this disclosure, and it is intended that the scope of the application encompass such variations and modifications as can come within the scope of the claims. In the claims, the term "comprising" does not exclude other elements or steps, the term "one" or "a" does not exclude a plurality, and the term "one" or "a" means "one or more but can include only one". Individual features or units can implement several functions of several claim items. Measures described in mutually different dependent claims can be combined and implemented to produce desirable results.

[0118] Although the present application is described herein in conjunction with various embodiments, other variations and modifications of the disclosed embodiments are possible in light of this disclosure, and it is intended that the scope of the application encompass such variations and modifications as can come within the scope of the claims. In the claims, the term "comprising" does not exclude other elements or steps, the term "one" or "a" does not exclude a plurality, and the term "one" or "a" means "one or more but can include only one". Individual features or units can implement several functions of several claim items. Measures described in mutually different dependent claims can be combined and implemented to produce desirable results.

Claims

1. A method for early warning based on multi-source data fusion of mine roof, characterized in that, include: Acquire multi-source data; The multi-source data includes roof stress, roof subsidence, roof subsidence rate, number of cracks, and crack aperture. The multi-source data is preprocessed; The preprocessing includes replacing outliers and filling in missing values; Extract normalized feature values ​​from the preprocessed multi-source data; the normalized feature values ​​include normalized stress mutation rate, mean stress, subsidence acceleration, cumulative subsidence displacement, aperture growth rate, and crack number increment. A basic probability allocation is constructed based on historical instability data of the mine roof, and the normalized feature values ​​are fused using DS evidence theory to output the warning confidence level of each warning level. The warning level of the mine roof is determined based on the aforementioned warning confidence level.

2. The method according to claim 1, characterized in that, Preprocessing the multi-source data includes: Replace outliers in multi-source data with the neighborhood median; Time-series linear interpolation is used to supplement missing values ​​in multi-source data.

3. The method according to claim 1, characterized in that, The extraction of normalized feature values ​​from preprocessed multi-source data includes: Set the feature window length and acquisition interval; Based on the preprocessed top plate stress data, the stress mutation rate and the mean stress are calculated. Based on the pre-processed data on the amount and rate of top slab subsidence, the subsidence acceleration and the cumulative subsidence displacement are calculated. Based on the preprocessed fracture aperture and fracture number data, the aperture growth rate and the fracture number increment are calculated. The eigenvalues ​​are normalized using Min-Max and mapped to the [0,1] interval to obtain the normalized eigenvalues.

4. The method according to claim 1, characterized in that, The basic probability allocation is constructed based on historical mine roof instability data, and the normalized feature values ​​are fused using DS evidence theory to output the warning confidence level for each warning level, including: An identification framework is constructed using the early warning level of the mine roof as an element; the early warning level includes a first level, a second level, a third level, and a fourth level; Based on historical instability data of mine roof, a mapping relationship between normalized characteristic value intervals and various early warning levels is established; Based on the normalized eigenvalues ​​and the mapping relationship, the m-functions of each eigenvalue are obtained; The Dempster combination rule is used to fuse the m-functions of each feature value, and the fused warning confidence level is output for each warning level.

5. The method according to claim 4, characterized in that, The Dempster combination rule is used to fuse the m-functions of each eigenvalue, satisfying the following formula: in, K represents the confidence level of an alert with an alert level of A, and K is the coefficient of evidence conflict. Basic probability assignment for stress characteristics Supported warning levels, Basic probability assignment for displacement features Supported warning levels, Basic probability assignment for crack features Supported warning levels.

6. The method according to claim 4, characterized in that, The first level corresponds to a Level I warning for the mine roof, the second level corresponds to a Level II warning, the third level corresponds to a Level III warning, and the fourth level corresponds to a Level IV warning; the severity of the first, second, third, and fourth levels increases sequentially.

7. The method according to claim 6, characterized in that, Determining the early warning level of the mine roof based on the aforementioned early warning confidence level includes: Take the maximum value of the warning confidence scores corresponding to each warning level, and use the corresponding warning level as the preliminary judgment level. If the sinking acceleration or the stress mutation rate exceeds a preset characteristic threshold, the preliminary judgment level will be raised by one level; the preliminary judgment level will not exceed the fourth level. If the maximum warning confidence level is lower than the preset confidence level threshold, the preliminary judgment level will be reduced by one level; The initial assessment level, whether raised or lowered, will be used as the early warning level for the mine roof.

8. The method according to claim 1, characterized in that, After determining the warning level of the mine roof based on the aforementioned warning confidence level, the method further includes: Based on the aforementioned warning level, a corresponding downhole on-site audible and visual alarm is triggered, and warning information is pushed to the mobile device; the warning information includes the warning level, multi-source data, feature values, and fusion confidence level. The aforementioned early warning information will be added to the historical instability database of the mine roof.

9. A multi-source data fusion early warning device for mine roof, characterized in that, The device includes: a communication unit and a processing unit; The communication unit is used to acquire multi-source data, including roof stress, roof subsidence, roof subsidence rate, number of cracks, and crack aperture. The processing unit is used to preprocess the multi-source data; the preprocessing includes replacing outliers and supplementing missing values; extracting normalized feature values ​​from the preprocessed multi-source data; the normalized feature values ​​include normalized stress mutation rate, mean stress, subsidence acceleration, cumulative subsidence displacement, opening growth rate, and crack number increment; constructing a basic probability allocation based on historical instability data of the mine roof, and using DS evidence theory to fuse the normalized feature values ​​to output the warning confidence level for each warning level; and determining the warning level of the mine roof based on the warning confidence level.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed on the mine roof multi-source data fusion early warning device, cause the mine roof multi-source data fusion early warning device to perform the method as described in any one of claims 1-8.

Images