A digital mine inspection data analysis method, device and medium

Abstract

The application discloses a kind of digital mine inspection data analysis method, equipment and medium, it is related to mine digital monitoring technical field, including, corresponding mine space coordinate system is constructed three-dimensional digital twin model, multi-source mine inspection data are collected, and according to the attribute of collection object, the semantic label of structure safety, equipment operation, environmental safety and personnel safety is written, and the inspection data set with semantic mark is generated;Based on the inspection data set with semantic mark, construct semantic layered tensor, and after stacking according to space unit, the coordinate information of corresponding space unit in three-dimensional digital twin model is fused, and the mine situation field is formed.The application realizes the accurate alignment and structured expression of multi-source inspection data under unified coordinates, and realizes the unified modeling and dynamic mapping of multi-semantic, multi-spatiotemporal data, so that mine safety monitoring is changed from static display to dynamic evaluation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of digital mine monitoring, and in particular to a digital mine inspection data analysis method, device and medium. BACKGROUND

[0002] With the continuous improvement of the digitalization and intelligentization level of mine production, mine inspection and safety monitoring technology has experienced a change from manual inspection to multi-source information perception and digital twin modeling. The wide application of technologies such as laser radar scanning, oblique photogrammetry, infrared thermal imaging, acoustic detection and environmental sensing enables the digital collection of mine spatial structure, equipment operating state and environmental parameters with centimeter-level precision and second-level frequency. In recent years, mine digital twin models based on three-dimensional point cloud reconstruction have become a research focus. By mapping multi-source perception data to a unified spatial coordinate system, the mine structure, equipment layout and operating environment can be reproduced in a virtual space, and real-time visualization of the operating state can be achieved. At the same time, the introduction of artificial intelligence algorithms and big data analysis technology makes the pattern recognition, feature extraction and anomaly detection of inspection data more efficient, and promotes the change of mine monitoring from static display to dynamic evaluation, providing technical support for safety production and intelligent scheduling.

[0003] There are still two problems in the existing mine digital inspection technology in terms of data fusion and risk assessment. Firstly, there are differences in sampling frequency, format structure and coordinate system between multi-source inspection data, and there is a lack of unified semantic organization method, so it is difficult to accurately associate and comprehensively analyze data from different sources under the same spatial reference. Secondly, the existing risk analysis focuses on state recognition at a single time section, lacks dynamic modeling mechanism for time series changes and spatial propagation characteristics, and cannot quantify the duration and diffusion trend of abnormal state, making it difficult to form continuous expression of the risk evolution process. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a digital mine inspection data analysis method to solve the problems of difficulty in fusing multi-source inspection data under a unified spatial semantic framework and lack of spatiotemporal evolution modeling leading to difficulty in dynamically evaluating risk state.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides a digital mine inspection data analysis method, which comprises,

[0008] A three-dimensional digital twin model corresponding to the mine spatial coordinate system is constructed, multi-source mine inspection data is collected, and semantic labels of structure safety, equipment operation, environmental safety and personnel safety are written according to the attributes of the collected objects to generate a set of inspection data with semantic labels;

[0009] A semantic layered tensor is constructed based on the set of inspection data with semantic labels, and after stacking according to the space unit, the coordinate information of the corresponding space unit in the three-dimensional digital twin model is fused to form a mine situation field;

[0010] According to the time series data of each space unit in the mine situation field, the spatio-temporal feature drift index in the continuous inspection period is calculated, and the persistence parameter and the spatial diffusion trend parameter of the abnormal state are extracted to form a risk evolution feature set;

[0011] Based on the risk evolution feature set, the abnormal persistence parameter and the spatial diffusion trend parameter are taken as weighted factors to calculate the risk evolution score of each space unit, and the risk evolution score and the corresponding spatial coordinate are written into the three-dimensional digital twin model to generate a risk evolution analysis layer.

[0012] As a preferred scheme of the digital mine inspection data analysis method, the three-dimensional digital twin model corresponding to the mine spatial coordinate system is constructed, and the specific steps are as follows,

[0013] Laser radar scanning data and oblique photography image data of the mine survey area are collected, and three-dimensional reconstruction is performed according to a unified spatial coordinate reference to generate a mine point cloud data set;

[0014] The mine point cloud data set is subjected to spatial gridding processing, and the space unit is divided, the point cloud density, normal vector direction, average value of reflection intensity and average value of texture color are calculated, and a space feature table is generated;

[0015] A three-dimensional grid is generated using the geometric topological relationship in the space feature table, and a three-dimensional digital twin model corresponding to the mine spatial coordinate system is constructed through a three-dimensional grid reconstruction algorithm.

[0016] As a preferred scheme of the digital mine inspection data analysis method, the three-dimensional digital twin model corresponding to the mine spatial coordinate system is constructed, and the specific steps are as follows,

[0017] According to the time stamp and spatial coordinate, the multi-source mine inspection data is matched with the corresponding space unit in the three-dimensional digital twin model to form a multi-source inspection data table;

[0018] According to the attributes of the collected objects, the semantic labels of structure safety, equipment operation, environmental safety and personnel safety are written into the multi-source mine inspection data table, and format checking and field alignment are performed to generate a set of inspection data with semantic labels.

[0019] As a preferred embodiment of the digital mine inspection data analysis method of the present invention, the specific steps for constructing a semantically hierarchical tensor based on the semantically labeled inspection data set are as follows:

[0020] The inspection data set with semantic tags is sorted in three ways according to spatial units, timestamps and semantic tags, and a tensor structure index framework is established.

[0021] Extract data fields from the inspection data set with semantic tags and perform normalization processing to generate normalized inspection data;

[0022] Normalized inspection data is written into the index framework of a tensor structure to form a semantically hierarchical tensor.

[0023] As a preferred embodiment of the digital mine inspection data analysis method of the present invention, the specific steps for forming the mine situation field are as follows:

[0024] Based on the semantic hierarchical tensor, a mapping relationship is established between tensor units and corresponding spatial units in the three-dimensional digital twin model to generate a spatially aligned semantic hierarchical tensor.

[0025] The semantic hierarchical tensor of spatial alignment is subjected to integrity verification and fused with the coordinate information of the corresponding spatial unit in the three-dimensional digital twin model to generate the mine situation field.

[0026] As a preferred embodiment of the digital mine inspection data analysis method of the present invention, the specific steps for calculating the spatiotemporal characteristic drift index within a continuous inspection cycle based on the time series data of each spatial unit in the mine situation field are as follows.

[0027] Extract the time series data corresponding to each spatial unit from the mine situation field according to the spatial unit number, and generate a spatial unit time series dataset;

[0028] Based on the spatial unit time series dataset, the numerical difference of two adjacent inspection cycles on the same index is calculated, and the ratio of the numerical difference to the time interval of the inspection cycle is used as the time change rate sequence.

[0029] Based on the spatial positional relationship of each spatial unit in the three-dimensional digital twin model, the set of adjacent spatial units of each spatial unit is obtained, and the spatial diffusion trend parameters are calculated.

[0030] The time rate of change sequence and the spatial diffusion trend parameter are merged according to the spatial unit number to generate a spatiotemporal characteristic drift index.

[0031] As a preferred embodiment of the digital mine inspection data analysis method of the present invention, the specific steps for forming the risk evolution feature set are as follows:

[0032] Based on the time change rate sequence, combined with the index threshold, the abnormal state persistence parameter is sequentially counted according to the inspection cycle.

[0033] The spatiotemporal feature drift index and the abnormal state persistence parameter are combined according to the number of the spatial unit to form a risk evolution feature set.

[0034] As a preferred scheme of the digital mine inspection data analysis method, the risk evolution analysis layer is generated, and the specific steps are as follows,

[0035] The abnormal state persistence parameter and the spatial diffusion trend parameter in the risk evolution feature set are assigned fixed weighting factors according to the number of the spatial unit to generate a weighted factor set;

[0036] For each spatial unit in the risk evolution feature set, the product of the abnormal state persistence parameter and the corresponding weighted factor of the abnormal state persistence parameter is summed to obtain a persistence weighted component;

[0037] For each spatial unit in the risk evolution feature set, the product of the spatial diffusion trend parameter and the corresponding weighted factor of the spatial diffusion trend parameter is summed to obtain a diffusion weighted component;

[0038] The persistence weighted component and the diffusion weighted component are summed to obtain a risk evolution score of the corresponding spatial unit, and a spatial unit risk evolution score set is generated;

[0039] The risk evolution score of each spatial unit in the spatial unit risk evolution score set is bound with the coordinate information of the corresponding spatial unit in the three-dimensional digital twin model to generate a spatial unit risk mapping table;

[0040] The risk evolution score in the spatial unit risk mapping table is written into the three-dimensional digital twin model according to the number of the spatial unit to generate the risk evolution analysis layer.

[0041] In a second aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and wherein the computer program, when executed by the processor, implements any step of the digital mine inspection data analysis method according to the first aspect of the present application.

[0042] In a third aspect, the present application provides a computer readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements any step of the digital mine inspection data analysis method according to the first aspect of the present application.

[0043] The present application has the beneficial effects that: by constructing a three-dimensional digital twin model corresponding to a mine space coordinate system, accurate alignment and structured expression of multi-source inspection data in a unified coordinate are realized, and the mine space structure and equipment operation state have computability and traceability; by constructing a semantic layered tensor and fusing three-dimensional digital twin model coordinate information, a mine situation field capable of simultaneously representing structure safety, equipment operation, environment safety and personnel safety is formed, unified modeling and dynamic mapping of multi-semantic and multi-time and space data are realized, thereby supporting risk evolution analysis and trend prediction, and making mine safety monitoring change from static display to dynamic evaluation. BRIEF DESCRIPTION OF DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0045] Fig. 1 The flowchart of the digital mine inspection data analysis method.

[0046] Fig. 2 The flowchart of multi-source inspection data semantic modeling.

[0047] Fig. 3 The flowchart of semantic layered tensor construction and mine situation field generation.

[0048] Fig. 4 The flowchart of risk evolution feature extraction and risk evolution analysis. DETAILED DESCRIPTION

[0049] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings of the specification.

[0050] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the connotation of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0051] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0052] REFERENCE Figs. 1-4For an embodiment of the present application, the embodiment provides a digital mine inspection data analysis method, comprising the following steps:

[0053] S1: a three-dimensional digital twin model corresponding to the mine space coordinate system is constructed, multi-source mine inspection data is collected, and semantic labels of structure safety, equipment operation, environment safety and personnel safety are written according to the attributes of the collected objects to generate a set of inspection data with semantic labels;

[0054] S1.1: laser radar scanning data and oblique photography image data of the mine survey area are collected, and three-dimensional reconstruction is performed according to a unified spatial coordinate reference to generate a mine point cloud data set;

[0055] Further, laser radar scanning is performed in the mine survey area to obtain laser radar scanning data, which includes spatial form information of the mine tunnel, mining area, supporting structure and equipment location. Oblique photogrammetry is performed in the mine survey area to obtain oblique photography image data, which includes texture and color information of the mine tunnel surface, surrounding rock surface and equipment surface. Spatial coordinates are established for the laser radar scanning data and the oblique photography image data, and the laser radar scanning data and the oblique photography image data are registered using a unified spatial coordinate reference, so that the laser radar scanning data and the oblique photography image data correspond to the same spatial position under the unified spatial coordinate reference. Three-dimensional reconstruction is performed on the registered laser radar scanning data and oblique photography image data to generate a three-dimensional point set containing spatial points of the mine survey area, and a mine point cloud data set is output.

[0056] S1.2: spatial gridding processing is performed on the mine point cloud data set, and spatial units are divided, point cloud density, normal vector direction, average value of reflectance intensity and average value of texture color are calculated, and a spatial feature table is generated;

[0057] Further, the mine point cloud data set is divided into multiple spatial units through spatial gridding processing. For each spatial unit, points located in the spatial unit are extracted from the mine point cloud data set, the number of points in the spatial unit is calculated, and the point cloud density is recorded. A local plane model is constructed based on the neighboring points of each point, and the plane is fitted through the least squares method to obtain the normal vector direction of the points in the spatial unit. The average value of the reflectance intensity values of all points in the spatial unit is calculated after accumulation, and the average value of the reflectance intensity is recorded. Based on the oblique photography image data, the average value of the texture color of the region corresponding to the spatial unit is calculated. The point cloud density, normal vector direction, average value of reflectance intensity, average value of texture color, spatial coordinates and spatial unit number associated with the spatial unit are recorded respectively according to the spatial unit to generate a spatial feature table.

[0058] The point cloud density is calculated, and the expression is:

[0059] ;

[0060] in, spatial unit Point cloud density, spatial unit The number of points, spatial unit The spatial volume of the corresponding spatial unit after spatial rasterization processing. Number the spatial unit.

[0061] It should be noted that a spatial unit refers to a discrete spatial region division unit obtained after spatial rasterization of a mine point cloud dataset. Spatial rasterization is the process of dividing continuously distributed mine point cloud data into regular discrete spatial units under a unified spatial coordinate system, used to achieve structured management of spatial data and subsequent quantitative calculations. Specifically, in the corresponding mine spatial coordinate system, fixed intervals along the X, Y, and Z axes are used to divide the entire spatial range of the mine survey area into several voxel-shaped cubic spatial units according to the intervals. For each spatial unit, points located within the boundary coordinate range of the spatial unit are retrieved from the mine point cloud dataset based on the boundary coordinate range of the spatial unit, and these points are recorded as the point set of the spatial unit. The continuous point cloud space is discretized into multiple spatial units with fixed numbers and coordinate boundaries, enabling subsequent calculation of spatial feature indicators such as point cloud density, normal vector direction, average reflection intensity, and average texture color within each spatial unit.

[0062] S1.3: Generate a three-dimensional mesh using the geometric topological relationships in the spatial feature table, and construct a three-dimensional digital twin model of the corresponding mine spatial coordinate system using a three-dimensional mesh reconstruction algorithm;

[0063] Furthermore, the geometric topological relationships in the spatial feature table are read, and a 3D mesh is generated using these relationships. These relationships represent the adjacency connections between spatial units in the spatial feature table, including contact relationships, connection order, and spatial organization. A 3D mesh reconstruction algorithm is then executed based on the mesh, mapping it to the spatial unit numbers, coordinates, and geometric topological relationships in the spatial feature table. This outputs a spatial structure representation corresponding to the mining survey area, serving as a 3D digital twin model of the corresponding mining spatial coordinate system.

[0064] It should be noted that the generation of the three-dimensional grid is guided to splice the space units into a continuous three-dimensional grid according to the geometric topological relationship. The three-dimensional grid refers to the continuous space surface and space body formed after connecting the space units according to the geometric topological relationship. The three-dimensional digital twin model corresponding to the mine space coordinate system refers to the three-dimensional space expression constructed based on the three-dimensional grid reconstruction algorithm, which is used to reflect the spatial form, structural layout and space unit distribution of the mine survey area.

[0065] S1.4: Collecting multi-source mine inspection data, the multi-source mine inspection data including high-resolution visible light images, infrared thermal imaging images, sound waveform signals, vibration acceleration sequences, gas concentration values, temperature and humidity values, dust concentration values, and personnel positioning trajectory data;

[0066] Further, high-resolution visible light images are collected to record the state of the roadway surface, the state of the support member, and the state of the equipment surface. Infrared thermal imaging images are collected to record the thermal distribution of the equipment surface. Sound waveform signals are collected to record the sound performance of the equipment during operation. Vibration acceleration sequences are collected to record the vibration performance of the equipment during operation. Gas concentration values, temperature values, humidity values, and dust concentration values are collected to record the environmental state. Personnel positioning trajectory data are collected to record the actual activity area of the personnel. While collecting the multi-source mine inspection data, a timestamp and a space coordinate corresponding to the collection position of each piece of multi-source mine inspection data are recorded to form multi-source mine inspection data containing a timestamp, a space coordinate, and corresponding observation content.

[0067] It should be noted that the space coordinate refers to the coordinate position of the position where the multi-source mine inspection data is collected in the corresponding mine space coordinate system, which is used to locate in the three-dimensional digital twin model corresponding to the mine space coordinate system.

[0068] S1.5: According to the timestamp and the space coordinate, the multi-source mine inspection data is matched with the corresponding space unit in the three-dimensional digital twin model to form a multi-source inspection data table;

[0069] Further, the timestamp and the space coordinate corresponding to each piece of multi-source mine inspection data are read, the space unit corresponding to the space coordinate is searched in the three-dimensional digital twin model corresponding to the mine space coordinate system using the space coordinate, and the space unit number of the space unit is obtained. The space unit number, the timestamp, and the observation content corresponding to the multi-source mine inspection data are recorded as a record row, and the observation content includes at least one of the high-resolution visible light images, the infrared thermal imaging images, the sound waveform signals, the vibration acceleration sequences, the gas concentration values, the temperature values, the humidity values, the dust concentration values, and the personnel positioning trajectory data. A plurality of record rows are accumulated to form a multi-source inspection data table, and the multi-source inspection data table is output.

[0070] S1.6: According to the attribute of the collection object, write the semantic labels of structure safety, equipment operation, environment safety and personnel safety into the multi-source mine inspection data table, and perform format verification and field alignment to generate an inspection data set with semantic labels;

[0071] Further, read each record in the multi-source inspection data table, and write semantic labels according to the attribute of the collection object. For example, write a structure safety semantic label for a record describing the stability state of the surrounding rock of the roadway, the support state of the roadway, and the deformation condition of the roadway. Write an equipment operation semantic label for a record describing the operation condition, operation vibration condition and operation acoustic condition of the conveyor belt, fan, pump station and electrical equipment. Write an environment safety semantic label for a record describing the gas concentration value, temperature value, humidity value and dust concentration value. Write a personnel safety semantic label for a record describing personnel positioning trajectory data. Perform format verification and field alignment on the multi-source inspection data table with written semantic labels, so that each record contains a space unit number, a timestamp, a space coordinate, an observation content and a semantic label. Output the data set after format verification and field alignment to generate an inspection data set with semantic labels.

[0072] It should be noted that the attribute of the collection object refers to the safety concern category to which the object described by the multi-source mine inspection data belongs, including structure safety, equipment operation, environment safety and personnel safety, which is used to determine which type of semantic label to write. The structure safety semantic label, the equipment operation semantic label, the environment safety semantic label and the personnel safety semantic label respectively represent the safety semantic category to which the record belongs. The inspection data set with semantic labels refers to the data set obtained after writing semantic labels and completing format verification and field alignment.

[0073] S2: Construct a semantic layered tensor based on the inspection data set with semantic labels, and fuse the coordinate information of the corresponding space unit in the three-dimensional digital twin model after stacking by space unit to form a mine situation field;

[0074] S2.1: Triple-sort the inspection data set with semantic labels according to space unit, timestamp and semantic label, and establish an index framework of tensor structure;

[0075] Further, read the inspection data set with semantic labels, and extract the space unit, timestamp and semantic label from the records therein. Group the records according to the space unit, arrange them in the same space unit according to the timestamp order, and arrange them according to the semantic label order under the same timestamp to complete the triple-sorting of the space unit, timestamp and semantic label; combine the spatial index position of the space unit, the time index position of the timestamp and the semantic index position of the semantic label to generate an index framework of tensor structure.

[0076] It should be noted that triple ordering refers to ordering the semantic labeled inspection data set according to the order of spatial unit, timestamp and semantic label, so that the records under the same spatial unit are arranged according to the timestamp, and the records of different semantic labels under the same timestamp have a fixed arrangement order.

[0077] S2.2: Extract data fields from the semantic labeled inspection data set and perform normalization processing to generate normalized inspection data;

[0078] Further, the data fields are extracted from the semantic labeled inspection data set, including the observation values corresponding to the structure safety semantic label, the equipment operation semantic label, the environment safety semantic label and the personnel safety semantic label. The numerical observation values under each semantic label are normalized according to a unified standard to eliminate dimensional differences and generate dimensionless data that can be directly compared, forming normalized inspection data.

[0079] It should be noted that the data field refers to the numerical information, waveform feature information, vibration feature information, image extraction feature information and personnel positioning trajectory information corresponding to the observation content in the semantic labeled inspection data set.

[0080] S2.3: Write the normalized inspection data into the index framework of the tensor structure to form a semantic layered tensor;

[0081] Further, according to the index framework of the tensor structure, the normalized inspection data is written into the corresponding position. For each record in the semantic labeled inspection data set, read its spatial index position, time index position and semantic index position, and write the normalized inspection data corresponding to the record into the corresponding unit in the index framework of the tensor structure. Repeat the writing until all the data corresponding to the spatial unit, timestamp and semantic label are filled to generate a semantic layered tensor.

[0082] It should be noted that the semantic layered tensor is a multi-dimensional data structure jointly positioned by spatial index, time index and semantic index, which is used to express the distribution of multi-semantic data under different spatial units and time nodes.

[0083] S2.4: According to the semantic layered tensor, establish a mapping relationship between the tensor unit and the corresponding spatial unit in the three-dimensional digital twin model, and generate a spatially aligned semantic layered tensor;

[0084] Further, the mapping relationship between the tensor unit and the corresponding spatial unit in the three-dimensional digital twin model is established. The spatial index position in the semantic layered tensor is read, and the spatial unit number recorded in the three-dimensional digital twin model is corresponded, so that each tensor unit is bound with the corresponding spatial unit number. The binding relationship is written into the semantic layered tensor to obtain a spatially aligned semantic layered tensor.

[0085] S2.5: integrity check is performed on the spatially aligned semantic hierarchical tensor, and the coordinate information of the corresponding spatial unit in the three-dimensional digital twin model is fused to generate a mine situation field;

[0086] Further, it is checked whether there is data record corresponding to all semantic labels under all time stamps for each spatial unit number, and the missing tensor unit is recorded. The spatially aligned semantic hierarchical tensor that passes the integrity check is fused with the coordinate information of the corresponding spatial unit in the three-dimensional digital twin model, and the normalized inspection data corresponding to each spatial unit number is bound with the spatial coordinates to form a spatial state expression containing spatial unit number, time stamp, semantic label and normalized inspection data, and output as a mine situation field.

[0087] It should be noted that the missing tensor unit refers to that, in the semantic hierarchical tensor, when the data record of any one or more of the four semantic labels (structure safety, equipment operation, environmental safety and personnel safety) corresponding to a spatial unit at a specific time stamp is not collected or not completely recorded, the tensor unit of the spatial unit at the time stamp is considered as missing. For example, if the spatial unit A lacks the data of the "personnel safety" label at time T1, or simultaneously lacks the data of the "equipment operation" and "environmental safety" labels, the corresponding tensor unit is a missing unit. The integrity check checks the existence of data of all semantic labels by spatial unit and time stamp, records the missing situation and excludes it, to ensure that the subsequent fusion operation is based on complete data units.

[0088] S3: According to the time series data of each spatial unit in the mine situation field, the spatio-temporal feature drift index in the continuous inspection period is calculated, and the persistence parameter and spatial diffusion trend parameter of the abnormal state are extracted to form a risk evolution feature set;

[0089] S3.1: Extract the time series data corresponding to each spatial unit from the mine situation field according to the number of spatial units to generate a spatial unit time series data set;

[0090] Further, all spatial unit numbers are read from the mine situation field. For each spatial unit number, the corresponding semantic hierarchical data values are extracted in time stamp order, including the normalized inspection data under the structure safety semantic label, the equipment operation semantic label, the environmental safety semantic label and the personnel safety semantic label. The data arranged in time stamp order under the same spatial unit number is combined into a time series, recorded as spatial unit time series data. All spatial unit number corresponding time series data is summarized to generate a spatial unit time series data set.

[0091] S3.2: Based on the spatial unit time series dataset, calculate the numerical difference of the same index in the adjacent two inspection periods, and take the ratio of the numerical difference and the time interval of the inspection period as the time change rate sequence;

[0092] Further, from the spatial unit time series dataset, select the data of the adjacent two inspection periods under the same spatial unit number, calculate the numerical difference of the normalized inspection data of the same index. Take the ratio of each numerical difference and the time interval of the adjacent two inspection periods as the time change rate. Repeat the calculation for all inspection periods under the same spatial unit number to form the time change rate sequence.

[0093] S3.3: Based on the spatial position relationship of each spatial unit in the three-dimensional digital twin model, obtain the adjacent spatial unit set of each spatial unit, and calculate the spatial diffusion trend parameter;

[0094] Further, read the spatial position relationship of the spatial unit recorded in the three-dimensional digital twin model, retrieve its spatial coordinates in the three-dimensional digital twin model according to the spatial unit number, identify the adjacent spatial units of the spatial coordinates, and form the adjacent spatial unit set. Calculate the difference value of the normalized inspection data of each spatial unit and its adjacent spatial unit set in the continuous inspection period, take the average value of all adjacent spatial unit difference values, and record it as the spatial diffusion trend parameter. Perform the same operation for all spatial unit numbers to generate the spatial diffusion trend parameter set.

[0095] S3.4: Merge the time change rate sequence and the spatial diffusion trend parameter according to the number of spatial units to generate the space-time feature drift index;

[0096] Further, merge the time change rate sequence and the spatial diffusion trend parameter according to the number of spatial units. Read the time change rate sequence and the corresponding spatial diffusion trend parameter under each spatial unit number, combine the time change rate sequence and the spatial diffusion trend parameter into the same record row. All record rows are arranged in order of spatial unit number to generate the space-time feature drift index.

[0097] It should be noted that the space-time feature drift index is a quantitative index composed of the time change rate sequence and the spatial diffusion trend parameter, which is used to describe the change trend of the spatial unit in the time dimension and the space dimension.

[0098] S3.5: Based on the time change rate sequence, combine the index threshold, and statistically calculate the abnormal state persistence parameter according to the inspection period order;

[0099] Further, based on the time change rate sequence, an index threshold is set for each spatial unit number. The time change rate of each inspection period is compared with the index threshold, and when the time change rate exceeds the index threshold, it is recorded as an abnormal state. When the time change rate of a plurality of consecutive inspection periods exceeds the index threshold, the number of consecutive inspection periods is counted, and the abnormal state persistence parameter is recorded. Repeat the statistics for all spatial unit numbers to generate a set of abnormal state persistence parameters.

[0100] It should be noted that the setting of the index threshold is determined according to the statistical distribution characteristics of the index corresponding to each type of semantic label in the historical inspection period. Specifically, for the time change rate sequence under each type of semantic label, the mean and standard deviation of the time change rate sequence in a plurality of inspection periods are counted, and a range of the weighted standard deviation of the mean is taken as the threshold interval. The lower limit of the threshold is the mean of the time change rate, and the upper limit of the threshold is the sum of the mean and the standard deviation of the time change rate. When the time change rate exceeds the upper limit threshold, it is determined to be an abnormal state. The threshold range can be applied to the time change rate judgment of the four types of indexes of structure safety, equipment operation, environmental safety and personnel safety, so that the threshold can cover the normal fluctuation interval and effectively distinguish the abnormal change trend.

[0101] S3.6: Merge the spatio-temporal feature drift index and the abnormal state persistence parameter according to the number of spatial units to form a risk evolution feature set;

[0102] Further, the spatio-temporal feature drift index and the abnormal state persistence parameter are merged according to the spatial unit number. The spatio-temporal feature drift index and the abnormal state persistence parameter corresponding to each spatial unit number are read and recorded as the same record row. All record rows are arranged in order of spatial unit number to form a risk evolution feature set.

[0103] S4: Based on the risk evolution feature set, the abnormal persistence parameter and the spatial diffusion trend parameter are used as weighted factors to calculate the risk evolution score of each spatial unit, and the risk evolution score and the corresponding spatial coordinates are written into the three-dimensional digital twin model to generate a risk evolution analysis layer;

[0104] S4.1: Assign fixed weighting factors to the abnormal state persistence parameter and the spatial diffusion trend parameter in the risk evolution feature set according to the number of spatial units to generate a set of weighting factors;

[0105] Further, read all spatial unit numbers in the risk evolution feature set. For each spatial unit number, extract the abnormal state persistence parameter and the spatial diffusion trend parameter from the risk evolution feature set. Set a fixed weighting factor for the abnormal state persistence parameter and a fixed weighting factor for the spatial diffusion trend parameter, and record the two types of weighting factors as the abnormal state persistence parameter weighting factor and the spatial diffusion trend parameter weighting factor, respectively. Record the two types of weighting factors corresponding to each spatial unit number as a group to form a weighting factor set.

[0106] It should be noted that when assigning weighting factors to the abnormal state persistence parameter and the spatial diffusion trend parameter, the global parameter distribution characteristics are calculated based on the historical inspection data, the average values and standard deviations of the two types of parameters are calculated, and the average value ratio is used as the initial weighting proportion benchmark. For each spatial unit, the deviation degree of the abnormal state persistence parameter and the spatial diffusion trend parameter from the global average value is calculated (represented by the ratio of the absolute value of the difference to the standard deviation). After normalization, the deviation degree is used as a correction coefficient, which is multiplied by the initial weighting proportion benchmark to obtain the abnormal state persistence parameter weighting factor and the spatial diffusion trend parameter weighting factor of the spatial unit, respectively. Through the combination of global benchmark and local difference, the weighting factor assignment reflects both the overall statistical law and the regional feature difference, and finally forms the weighting factor set corresponding to the spatial unit number.

[0107] S4.2: For each spatial unit in the risk evolution feature set, the product of the abnormal state persistence parameter and the weighting factor corresponding to the abnormal state persistence parameter is summed to obtain the persistence weighted component;

[0108] Further, for each spatial unit number in the risk evolution feature set, the corresponding abnormal state persistence parameter and abnormal state persistence parameter weighting factor are read. The product of the abnormal state persistence parameter and the abnormal state persistence parameter weighting factor is obtained. The multiple product results under the same spatial unit number are summed to obtain the persistence weighted component. The persistence weighted component corresponding to each spatial unit number is recorded as the persistence weighted result.

[0109] S4.3: For each spatial unit in the risk evolution feature set, the product of the spatial diffusion trend parameter and the weighting factor corresponding to the spatial diffusion trend parameter is summed to obtain the diffusion weighted component;

[0110] Further, for each spatial unit number in the risk evolution feature set, the product of the multiple spatial diffusion trend parameters and the weighting factors corresponding to the spatial diffusion trend parameters under the same spatial unit number is summed to obtain the diffusion weighted component. The diffusion weighted component corresponding to each spatial unit number is recorded as the diffusion weighted result.

[0111] S4.4: Sum the persistence weighted component and the diffusion weighted component to obtain the risk evolution score of the corresponding spatial unit, and generate a spatial unit risk evolution score set;

[0112] Further, for each spatial unit number, sum the persistence weighted component and the diffusion weighted component, and record the risk evolution score of the spatial unit. Record the risk evolution scores corresponding to all spatial unit numbers in order to generate a spatial unit risk evolution score set.

[0113] S4.5: Bind the risk evolution score of each spatial unit in the spatial unit risk evolution score set with the coordinate information of the corresponding spatial unit in the three-dimensional digital twin model to generate a spatial unit risk mapping table;

[0114] Further, read each spatial unit number in the spatial unit risk evolution score set, and retrieve the spatial unit coordinate information corresponding to the spatial unit number from the three-dimensional digital twin model. Record the spatial unit number, spatial unit coordinate information and corresponding risk evolution score as a record row, and accumulate the records in order of spatial unit number to generate a spatial unit risk mapping table.

[0115] S4.6: Write the risk evolution score in the spatial unit risk mapping table into the three-dimensional digital twin model according to the spatial unit number to generate a risk evolution analysis layer;

[0116] Further, read the spatial unit number, spatial unit coordinate information and risk evolution score from the spatial unit risk mapping table, and write the risk evolution score into the attribute field of the corresponding spatial unit in the three-dimensional digital twin model according to the spatial unit number. Output a structure containing the spatial unit number, spatial unit coordinate information and risk evolution score to generate a risk evolution analysis layer.

[0117] It should be noted that the risk evolution analysis layer refers to a data layer formed after embedding the risk evolution score in the three-dimensional digital twin model, which is used to represent the risk evolution state and distribution of each spatial unit in the mine.

[0118] The embodiment also provides a computer device suitable for the digital mine inspection data analysis method, which comprises a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to realize the digital mine inspection data analysis method proposed in the above embodiment.

[0119] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected by a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is configured to perform wired or wireless communication with an external terminal. The wireless communication can be achieved by WIFI, an operator network, NFC (Near Field Communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0120] The embodiment also provides a storage medium having a computer program stored thereon, the program being executed by a processor to implement the method for analyzing digitalized mine inspection data according to the above embodiment. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic storage, a flash memory, a magnetic disk or an optical disk.

[0121] In summary, the present application achieves the precise alignment and structured expression of multi-source inspection data in a unified coordinate by constructing a three-dimensional digital twin model corresponding to the mine spatial coordinate system, so that the mine spatial structure and the equipment operation state have computability and traceability. By constructing a semantic layered tensor and fusing the coordinate information of the three-dimensional digital twin model, a mine situation field capable of representing structure safety, equipment operation, environment safety and personnel safety is formed, the unified modeling and dynamic mapping of multi-semantic and multi-spatial-temporal data are achieved, and the risk evolution analysis and trend prediction are supported, so that the mine safety monitoring is converted from static display to dynamic evaluation.

[0122] It should be noted that the above examples are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced, without departing from the spirit and scope of the technical solutions of the present application, which should be covered in the scope of the claims of the present application.

Claims

1. A digital mine inspection data analysis method, characterized in that: include, Construct a three-dimensional digital twin model corresponding to the spatial coordinate system of the mine, collect multi-source mine inspection data, and write semantic tags for structural safety, equipment operation, environmental safety and personnel safety according to the attributes of the collected objects to generate a set of inspection data with semantic tags. A semantically hierarchical tensor is constructed based on the inspection data set with semantic tags, and after being stacked according to spatial units, it is fused with the coordinate information of the corresponding spatial units in the three-dimensional digital twin model to form a mine situation field. Based on the time series data of each spatial unit in the mine situation field, the spatiotemporal characteristic drift index within the continuous inspection cycle is calculated, and the persistence parameters and spatial diffusion trend parameters of the abnormal state are extracted to form a risk evolution characteristic set. Based on the risk evolution feature set, the anomaly persistence parameter and spatial diffusion trend parameter are used as weighting factors to calculate the risk evolution score of each spatial unit. The risk evolution score and corresponding spatial coordinates are then written into a three-dimensional digital twin model to generate a risk evolution analysis layer.

2. The digital mine inspection data analysis method as described in claim 1, characterized in that: The specific steps for constructing a three-dimensional digital twin model corresponding to the mine's spatial coordinate system are as follows. Collect lidar scanning data and oblique photogrammetry data of the mining survey area, and perform three-dimensional reconstruction according to a unified spatial coordinate benchmark to generate a mining point cloud dataset. Spatial rasterization is performed on the mine point cloud dataset, and spatial cells are divided. Point cloud density, normal vector direction, average reflection intensity and average texture color are calculated to generate a spatial feature table. A three-dimensional mesh is generated using the geometric and topological relationships in the spatial feature table, and a three-dimensional digital twin model of the corresponding mine spatial coordinate system is constructed using a three-dimensional mesh reconstruction algorithm.

3. The digital mine inspection data analysis method as described in claim 2, characterized in that: The specific steps for generating the semantically labeled inspection data set are as follows: Based on timestamps and spatial coordinates, multi-source mine inspection data are matched with corresponding spatial units in the three-dimensional digital twin model to form a multi-source inspection data table. Based on the attributes of the collected objects, semantic tags for structural safety, equipment operation, environmental safety, and personnel safety are written into the multi-source mine inspection data table, and format verification and field alignment are performed to generate an inspection data set with semantic tags.

4. The digital mine inspection data analysis method as described in claim 3, characterized in that: The specific steps for constructing a semantically hierarchical tensor based on the semantically labeled inspection dataset are as follows. The inspection data set with semantic tags is sorted in three ways according to spatial units, timestamps and semantic tags, and a tensor structure index framework is established. Extract data fields from the inspection data set with semantic tags and perform normalization processing to generate normalized inspection data; Normalized inspection data is written into the index framework of a tensor structure to form a semantically hierarchical tensor.

5. The digital mine inspection data analysis method as described in claim 4, characterized in that: The specific steps for forming the mine's geological field are as follows. Based on the semantic hierarchical tensor, a mapping relationship is established between tensor units and corresponding spatial units in the three-dimensional digital twin model to generate a spatially aligned semantic hierarchical tensor. The semantic hierarchical tensor of spatial alignment is subjected to integrity verification and fused with the coordinate information of the corresponding spatial unit in the three-dimensional digital twin model to generate the mine situation field.

6. The digital mine inspection data analysis method as described in claim 5, characterized in that: The specific steps for calculating the spatiotemporal characteristic drift index within a continuous inspection cycle based on the time series data of each spatial unit in the mine situation field are as follows. Extract the time series data corresponding to each spatial unit from the mine situation field according to the spatial unit number, and generate a spatial unit time series dataset; Based on the spatial unit time series dataset, the numerical difference of two adjacent inspection cycles on the same index is calculated, and the ratio of the numerical difference to the time interval of the inspection cycle is used as the time change rate sequence. Based on the spatial positional relationship of each spatial unit in the three-dimensional digital twin model, the set of adjacent spatial units of each spatial unit is obtained, and the spatial diffusion trend parameters are calculated. The time rate of change sequence and the spatial diffusion trend parameter are merged according to the spatial unit number to generate a spatiotemporal characteristic drift index.

7. The digital mine inspection data analysis method as described in claim 6, characterized in that: The specific steps for forming the risk evolution feature set are as follows: Based on the time change rate series and combined with indicator thresholds, the parameters of abnormal status persistence are statistically analyzed in the order of the inspection cycle. The spatiotemporal characteristic drift index and the abnormal state persistence parameter are merged according to the spatial unit number to form a risk evolution characteristic set.

8. The digital mine inspection data analysis method as described in claim 7, characterized in that: The specific steps for generating the risk evolution analysis layer are as follows: According to the spatial unit number, fixed weighting factors are assigned to the abnormal state persistence parameters and spatial diffusion trend parameters in the risk evolution feature set to generate a weighting factor set; For each spatial unit in the risk evolution feature set, the product of the abnormal state persistence parameter and the weighting factor corresponding to the abnormal state persistence parameter is summed to obtain the persistence weighted component. For each spatial unit in the risk evolution feature set, the product of the spatial diffusion trend parameter and the weighting factor corresponding to the spatial diffusion trend parameter is summed to obtain the diffusion weighted component. The risk evolution score of the corresponding spatial unit is obtained by summing the persistent weighted component and the diffuse weighted component, and a set of spatial unit risk evolution scores is generated. The risk evolution score of each spatial unit in the spatial unit risk evolution score set is bound to the coordinate information of the corresponding spatial unit in the three-dimensional digital twin model to generate a spatial unit risk mapping table. The risk evolution score in the spatial unit risk mapping table is written into the three-dimensional digital twin model according to the spatial unit number to generate the risk evolution analysis layer.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the digital mine inspection data analysis method according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the digital mine inspection data analysis method according to any one of claims 1 to 8.

Images