Coal mine underground navigation control system and device based on big data

Abstract

The invention discloses a coal mine underground navigation control system and device based on big data, and particularly relates to the technical field of underground navigation control. The underground roadway waveguide effect grade is judged by monitoring underground radio signals in real time and recognizing a signal reflection mode by combining wavelet transform with a convolutional neural network; meanwhile, three-dimensional point cloud data of the underground roadway are processed through multi-scale curvature analysis and a DBSCAN clustering algorithm, and division of a structure mutation region is achieved; determining a navigation error level according to the waveguide effect level and a structure mutation result, and when the navigation error level reaches a second level or a third level, analyzing historical navigation trajectory data by using a dynamic time warping algorithm, and evaluating a current navigation path deviation degree; comprehensive analysis is carried out through the deep belief network, the reliability of the navigation control system is evaluated, whether a path re-planning instruction is triggered or not is judged, and the underground navigation precision and safety are improved.

Description

Technical Field

[0001] The present invention relates to the field of underground navigation control technology, and more specifically, to a coal mine underground navigation control system and device based on big data. Background Art

[0002] As the underground working environment of coal mines becomes increasingly complex and safety standards continue to improve, traditional underground navigation systems mostly use inertial navigation, radio ranging and other methods. However, in actual applications, due to factors such as tunnel curvature, metal support and geological structure, radio signals are often affected by multipath interference and waveguide effects, resulting in increased positioning errors. At the same time, there are structural mutations and local geological anomalies in the underground environment. The traditional single data source method is difficult to reflect environmental changes in real time, resulting in obvious path planning deviations.

[0003] In order to solve the above problems, a coal mine underground navigation control system and device based on big data is provided. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a coal mine underground navigation control system and device based on big data to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions: A coal mine underground navigation control system based on big data, comprising a waveguide effect judgment module, a structure mutation division module, an error level determination module, a deviation degree evaluation module and a reliability evaluation module; The waveguide effect judgment module monitors the propagation characteristics of radio signals in underground tunnels in real time, uses wavelet transform combined with convolutional neural network to identify the reflection mode of radio signals, and judges the waveguide effect level of underground tunnels; The structural mutation division module performs multi-scale curvature analysis on the three-dimensional point cloud data of the underground tunnel and divides the structural mutation area of ​​the underground tunnel based on the DBSCAN clustering algorithm; The error level determination module determines the navigation error level of the underground tunnel based on the waveguide effect level of the underground tunnel and the division result of the underground tunnel structure mutation area; When the navigation error level reaches level 2 or 3, the deviation degree assessment module analyzes the historical navigation trajectory data set through the dynamic time warping algorithm to assess the deviation degree of the current navigation path; The reliability assessment module inputs the waveguide effect level of the underground tunnel, the division results of the underground tunnel structure mutation area, and the deviation degree of the current navigation path into the deep belief network for comprehensive analysis, evaluates the reliability of the navigation control system, and determines whether to trigger the path replanning instruction.

[0006] In a preferred embodiment, by real-time monitoring of radio signal propagation characteristics in underground tunnels, wavelet transform combined with convolutional neural network is used to identify radio signal reflection patterns, and the waveguide effect level of the current underground tunnel is determined, specifically: Collect underground radio signals in real time to obtain amplitude, phase and delay data; Preprocess the collected data to reduce noise; Use wavelet transform to decompose radio signals in time and frequency and extract multi-scale features; Construct a multi-dimensional tensor and input the decomposition results into the convolutional neural network model; Use convolutional neural networks to identify signal reflection patterns and interference features; The level of waveguide effect in the underground tunnel is determined based on the identification results.

[0007] In a preferred embodiment, the level of waveguide effect in the underground tunnel is determined according to the recognition result, specifically: The waveguide effect level is determined based on the preset threshold contained in the decision function: When the arithmetic mean of the key characteristic values ​​is greater than or equal to the preset threshold, the waveguide effect level of the underground tunnel is determined to be a low waveguide effect level; When the arithmetic mean of the key characteristic values ​​is less than a preset threshold, the waveguide effect level of the underground tunnel is determined to be a high waveguide effect level.

[0008] In a preferred embodiment, a multi-scale curvature analysis is performed on the three-dimensional point cloud data of the underground tunnel, and the underground tunnel structure mutation area is divided based on the DBSCAN clustering algorithm, specifically: Collect and analyze 3D point cloud data; Select multiple layers of scales and calculate local curvature values; Extracting super-threshold curvature regions at each scale; Constructing multi-scale curvature features as clustering input vectors; The DBSCAN clustering algorithm was used to identify the structural mutation regions; Output the boundary of the underground tunnel structure mutation area; Provide risk warning for areas with sudden structural changes in underground tunnels.

[0009] In a preferred embodiment, risk warning is performed on the area with sudden structural changes in the underground tunnel, specifically: Preset low risk threshold and high risk threshold to compare with the warning index: When the warning index is less than or equal to the low risk threshold, the area with sudden changes in the underground tunnel structure is defined as a low risk area; When the warning index is greater than the low risk threshold and less than the high risk threshold, the area with sudden changes in the underground tunnel structure is defined as a medium risk area; When the warning index is greater than or equal to the high-risk threshold, the area with sudden structural changes in the underground tunnel is defined as a high-risk area.

[0010] In a preferred embodiment, based on the waveguide effect level of the underground tunnel and the result of the division of the underground tunnel structure mutation area, the underground tunnel navigation error level is determined, specifically: When the waveguide effect level of the underground tunnel is a low waveguide effect level, and the warning index is less than or equal to the low risk threshold, the underground tunnel navigation error level is determined to be level one; When the waveguide effect level of the underground tunnel is a high waveguide effect level, and the warning index is greater than the low risk threshold and less than the high risk threshold, the underground tunnel navigation error level is determined to be level 2; When the waveguide effect level of the underground tunnel is a high waveguide effect level and the warning index is greater than or equal to the high risk threshold, the underground tunnel navigation error level is determined to be level three.

[0011] In a preferred embodiment, the historical navigation trajectory data set is analyzed by a dynamic time warping algorithm to evaluate the deviation degree of the current navigation path, specifically: Standardized historical navigation trajectory data set and real-time navigation trajectory data set; Construct multi-dimensional feature sequences; Choose Euclidean distance as the sequence metric; Use dynamic time warping algorithm to optimally align trajectory sequences; Calculate the trajectory alignment error and output the dynamic time warping matching score; Compare the matching score with the matching score threshold to quantify the degree of deviation of the current navigation path.

[0012] In a preferred embodiment, the matching score is compared with the matching score threshold to quantify the degree of deviation of the current navigation path, specifically: Preset match score threshold, compare the dynamic time warping match score with the preset match score threshold: When the dynamic time warping matching score is greater than or equal to the preset matching score threshold, it indicates that the deviation degree of the current navigation path is high; When the dynamic time warping matching score is greater than or equal to the preset matching score threshold, it indicates that the deviation degree of the current navigation path is low.

[0013] In a preferred embodiment, the waveguide effect level of the underground tunnel, the division result of the underground tunnel structure mutation area, and the deviation degree of the current navigation path are input into the deep belief network for comprehensive analysis to evaluate the reliability of the navigation control system and determine whether to trigger the path replanning instruction, specifically: Integrate the waveguide effect level of the underground tunnel, the results of the underground tunnel structure mutation area division and the dynamic time warping matching score; Construct multi-channel input and feed it into deep belief network learning; Extract multi-source features in the hidden layer and identify potential correlations; Through network weight iteration, a reliability score of the navigation control system is generated.

[0014] In a preferred implementation, determining whether to trigger a path replanning instruction is specifically as follows: Preset the reliability score threshold and compare the reliability score with the reliability score threshold: When the reliability score is greater than the reliability score threshold, it indicates that the reliability of the navigation control system is high, and it is determined that the path replanning instruction is not triggered; When the reliability score is less than or equal to the reliability score threshold, it indicates that the reliability of the navigation control system is poor, and a path replanning instruction is triggered.

[0015] On the other hand, the present invention provides a coal mine underground navigation control device based on big data, comprising: one or more processors; A storage device for storing one or more programs; When one or more programs are executed by one or more processors, the one or more processors implement a coal mine underground navigation control system based on big data.

[0016] The technical effects and advantages of the coal mine underground navigation control system and device based on big data of the present invention are as follows: Through multi-source data fusion and intelligent processing, high-precision positioning and risk warning of the underground navigation control system are achieved. By using the waveguide effect of real-time monitoring of underground radio signals and multi-scale curvature analysis of three-dimensional point cloud data, the navigation control system can accurately identify signal interference and structural mutations caused by metal support, tunnel bending and geological anomalies in the environment, thereby accurately determining the level of navigation error. When the navigation error reaches level 2 or 3, the dynamic time warping algorithm evaluates the current navigation path deviation in real time, and combines the deep confidence network to comprehensively analyze the waveguide effect level, the structural mutation area division results and the deviation degree of the current navigation path, generate a reliability score, and then determine whether to trigger the path replanning instruction. Compared with the traditional single data source method, it greatly improves the adaptability to complex underground environments and decision-making accuracy, reduces positioning errors and safety risks, provides reliable technical guarantees and risk prevention and control measures for underground coal mine operations, and effectively improves navigation accuracy and the overall reliability of the navigation control system. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1This is a structural schematic diagram of a coal mine underground navigation control system based on big data according to the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Embodiment 1: Figure 1 The invention provides a coal mine underground navigation control system based on big data, including a waveguide effect judgment module, a structure mutation division module, an error level determination module, a deviation degree evaluation module and a reliability evaluation module; The waveguide effect judgment module monitors the propagation characteristics of radio signals in underground tunnels in real time, uses wavelet transform combined with convolutional neural network to identify the reflection mode of radio signals, and judges the waveguide effect level of underground tunnels; The structural mutation division module performs multi-scale curvature analysis on the three-dimensional point cloud data of the underground tunnel and divides the structural mutation area of ​​the underground tunnel based on the DBSCAN clustering algorithm; The error level determination module determines the navigation error level of the underground tunnel based on the waveguide effect level of the underground tunnel and the division result of the underground tunnel structure mutation area; When the navigation error level reaches level 2 or 3, the deviation degree assessment module analyzes the historical navigation trajectory data set through the dynamic time warping algorithm to assess the deviation degree of the current navigation path; The reliability assessment module inputs the waveguide effect level of the underground tunnel, the division results of the underground tunnel structure mutation area, and the deviation degree of the current navigation path into the deep belief network for comprehensive analysis, evaluates the reliability of the navigation control system, and determines whether to trigger the path replanning instruction.

[0020] Specifically, by real-time monitoring of the radio signal propagation characteristics in the underground tunnel, the wavelet transform combined with the convolutional neural network is used to identify the radio signal reflection mode and determine the waveguide effect level of the current underground tunnel, including: Real-time collection of underground radio signals to obtain amplitude, phase and delay data: multiple fixed radio signal transmitting and receiving devices are deployed in the underground tunnels, and these radio signal transmitting and receiving devices are used to collect radio signals in real time at different measuring points. The radio signal transmitting device periodically transmits electromagnetic waves at a set power and frequency, and the radio signal receiving device captures the radio signals reflected or scattered from the tunnel walls, metal supports, branches and other structures. The collected data includes the amplitude, phase information and propagation delay of the radio signal. The radio signal can be described by the following formula: ;in, Indicates the time variable The corresponding radio signal value; represents the time variable; Indicates the amplitude of a radio signal; Indicates the frequency of a radio signal; Indicates the initial phase of the signal.

[0021] In addition, the delay value is obtained by measuring the time interval between the transmission and reception of the radio signal, which is recorded as , which represents the time difference between the transmission and reception of a radio signal.

[0022] Preprocess the collected data to reduce noise: The collected raw signal data often contains environmental noise and redundant information, and needs to be preprocessed to improve data quality. In the preprocessing process, a digital filtering algorithm is used to filter the raw radio signal. For noise reduction, a Butterworth low-pass filter is usually used. Its expression is: ;in, represents the radio signal after filtering; Represents the overall function of the digital filtering operation.

[0023] Use wavelet transform to decompose the signal in time and frequency and extract multi-scale features: radio signal after preprocessing The time-frequency decomposition will be performed through continuous wavelet transform. The wavelet transform formula is defined as follows: ;in, Represents the wavelet coefficient, indicating the scale and pan The local energy of the radio signal under represents the mother wavelet function, and its complex conjugate is denoted by ; It represents the scale parameter. The value range is determined according to the characteristics of the radio signal. The smaller the scale, the more high-frequency information can be captured. The larger the scale, the more obvious the low-frequency trend is. Represents the translation parameter, which controls the position of the wavelet function on the time axis.

[0024] Construct a multidimensional tensor and input the decomposition result into the convolutional neural network model: transform the wavelet coefficients According to time, frequency and characteristic channels, it is organized into a multidimensional tensor, denoted as , ;in, Indicates the number of time periods; It represents the number of frequency scales, that is, the number of different scale components selected from the wavelet transform; Indicates the number of feature channels.

[0025] Each element in the tensor Indicated in time period, frequency scale and The eigenvalues ​​under the feature channels.

[0026] The tensor data structure retains the time-frequency feature information extracted by wavelet decomposition, providing high-dimensional input for convolutional neural networks, facilitating deep feature extraction and pattern recognition.

[0027] Using convolutional neural networks to identify signal reflection patterns and interference features: Using multi-layer convolutional neural networks to construct multidimensional tensors Processing is performed to identify reflection patterns and interference features in radio signals. Multi-layer convolutional neural networks mainly use convolutional layers, pooling layers, and fully connected layers. The core convolution operation can be described as: ;in, Indicates that after the convolution operation, The obtained feature map value; represents the activation function; Respectively represent the size of the convolution kernel in the time, frequency and feature channel directions; represents the convolution kernel The weight of each element; Represents the output multidimensional tensor The elements at the corresponding offset positions; Represents the bias term, which is a constant compensation term in the convolution operation.

[0028] After being trained with a large amount of labeled data, the convolutional neural network can learn the characteristics corresponding to different reflection and interference patterns, thereby accurately extracting the implicit waveguide effect information in the radio signal.

[0029] Determine and output the waveguide effect level of the underground tunnel based on the recognition results: After obtaining the feature mapping output of the convolutional neural network, the waveguide effect of the current underground tunnel is graded and judged based on the output results. The obtained feature mapping values ​​are screened, aggregated and statistically processed to obtain the key feature values. Introduce the decision function The key eigenvalues ​​are integrated and calculated, and the expression of the decision function is: ;in, Indicates the determined waveguide effect level; Represents the arithmetic mean of the selected key characteristic values, calculated as .in, Indicates Key feature values; Indicates the total number of key eigenvalues.

[0030] According to the preset threshold contained in the decision function Determine the level of waveguide effect: When the arithmetic mean of the key eigenvalues Greater than or equal to the preset threshold When , the waveguide effect level of the underground tunnel is determined to be a low waveguide effect level; When the arithmetic mean of the key eigenvalues Less than the preset threshold When , the waveguide effect level of the underground tunnel is determined to be a high waveguide effect level.

[0031] Specifically, multi-scale curvature analysis is performed on the three-dimensional point cloud data of the underground tunnel, and the underground tunnel structure mutation area is divided based on the DBSCAN clustering algorithm, including: Collect 3D point cloud data: Collect 3D point cloud data of underground tunnels through laser scanning or structured light sensors. Define the entire 3D point cloud data set as: ;in, Represents the collected 3D point cloud data set; Represents any single sampling point in a 3D point cloud data set; Indicates sampling point The coordinate information of . Respectively represent the coordinate values ​​of the sampling points in three orthogonal directions (such as X-axis, Y-axis, and Z-axis); Represents three-dimensional real number space.

[0032] Select multiple scales and calculate local curvature values: In order to reveal the local changes in the geometry of the underground tunnel, a multi-scale analysis method is used to calculate the local curvature of each pre-processed point. At a certain scale, the point set in its local neighborhood is recorded as: ;in, Indicated by point Centered on scale The neighborhood set under contains all points Distance not exceeding point; Represents a 3D point cloud data set Any point of The local neighborhood of Indicate point With point The Euclidean distance between Indicates the current scale The neighborhood radius defined below.

[0033] In the neighborhood collection By calculating the covariance matrix of the neighborhood and solving its eigenvalues, we can get three eigenvalues: , and define . Calculate the local curvature value, the calculation formula is: ;in, Indicated in scale Next, click The local curvature value of Indicated in scale Next, by calculating the point Neighborhood Set The minimum eigenvalue obtained after the covariance matrix of ; Indicated in scale Next, by calculating the point Neighborhood Set The intermediate eigenvalues ​​obtained after the covariance matrix of Indicated in scale Next, by calculating the point Neighborhood Set The maximum eigenvalue is obtained after resolving the covariance matrix of .

[0034] For multiple scales, define the scale set as , then for each point A range of local curvature values ​​can be obtained (in ), reflection point Local deformation properties at different scales. Indicates the total number of scales selected.

[0035] Extracting super-threshold curvature regions at each scale: At each scale, the calculated local curvature value is compared with the pre-set local curvature threshold, and the region with a local curvature value higher than the local curvature threshold is extracted. The local curvature threshold is defined as , whose value is determined by calibration based on experimental data.

[0036] For scale , if point Satisfy the local curvature value Greater than the local curvature threshold , then it is considered that in scale Next, click There is a local geometric mutation phenomenon, which is classified into the super-threshold curvature region and is recorded as .in, Indicated in scale The set of all points whose local curvature values ​​exceed the local curvature threshold.

[0037] Construct the local curvature values ​​at multiple scales as clustering input vectors: Construct a feature vector. Let the feature vector be ;in, Indicate point The multi-scale curvature eigenvector of ; Respectively indicate points At different scales The local curvature value is calculated below.

[0038] Application of DBSCAN clustering algorithm to identify structural mutation regions: using point The multi-scale curvature feature vector of point cloud data is used for cluster analysis. The multi-scale curvature feature vector of is defined as: ;in, Indicate point Multi-scale curvature eigenvector The neighborhood set centered on Representation and point The multi-scale curvature eigenvectors of the points to be compared The multi-scale curvature eigenvector of ; Represents the multi-scale curvature eigenvector and multi-scale curvature eigenvectors The Euclidean distance between .

[0039] If satisfied , then the definition point is the core point; Represents a neighborhood set The number of elements in It is the distance radius threshold in the DBSCAN clustering algorithm.

[0040] According to the clustering concept of DBSCAN, all core points and their neighboring points that are mutually density-reachable will be classified into the same cluster, thereby dividing the area with local structural mutations. For each type of area obtained by clustering, it is recorded as ,in The clustering index.

[0041] Output the boundary of the underground tunnel structure mutation area: For each cluster area , using the convex hull algorithm to calculate the outer boundary of the clustering area. Definition ;in, Represents cluster area The boundary set of is determined by the convex hull algorithm; Represents the convex hull solving function, whose output is the set of points that constitute the minimum convex polygon of the region.

[0042] Risk warning for underground tunnel structure mutation areas: underground tunnel structure mutation areas are divided into cluster areas The region boundaries corresponding to the cluster regions Define the warning index. Each point within Calculate its multi-scale average local curvature value , and its calculation formula is: ;in, Indicate point The average local curvature value at all scales; Indicates the total number of scales selected.

[0043] Clustering area Warning index It is defined as the mean of the average curvature values ​​of all points in the region, that is ;in, Indicates the warning index.

[0044] Specifically, based on the waveguide effect level of the underground tunnel and the division result of the underground tunnel structure mutation area, the underground tunnel navigation error level is determined, including: Preset low risk threshold and high risk threshold and compare with the warning index: When the warning index is less than or equal to the low risk threshold, the area with sudden changes in the underground tunnel structure is defined as a low risk area; When the warning index is greater than the low risk threshold and less than the high risk threshold, the underground tunnel structure mutation area is defined as a medium risk area; When the warning index is greater than or equal to the high-risk threshold, the area with sudden structural changes in the underground tunnel is defined as a high-risk area.

[0045] The low-risk and high-risk thresholds are determined by combining historical data from underground structure maintenance and navigation systems, as well as the experience of safety experts and industry standards.

[0046] When the waveguide effect level of the underground tunnel is a low waveguide effect level, and the warning index is less than or equal to the low risk threshold, the underground tunnel navigation error level is determined to be level one; When the waveguide effect level of the underground tunnel is a high waveguide effect level, and the warning index is greater than the low risk threshold and less than the high risk threshold, the underground tunnel navigation error level is determined to be level 2; When the waveguide effect level of the underground tunnel is a high waveguide effect level and the warning index is greater than or equal to the high risk threshold, the underground tunnel navigation error level is determined to be level three.

[0047] Specifically, the historical navigation trajectory data set is analyzed through the dynamic time warping algorithm to evaluate the deviation degree of the current navigation path, including: Standardized historical navigation trajectory data set and real-time navigation trajectory data set: The historical trajectory data set and the real-time trajectory data set are preprocessed, and the historical navigation trajectory data set is recorded as: ;in, Represents a collection of historical navigation trajectory data; Represents the first The original data vector of sampling points; Represents the dimension of the original data vector (e.g. contains spatial coordinate information); Indicates the total number of sampling points in the historical navigation trajectory data set.

[0048] Similarly, the real-time navigation trajectory data set is recorded as: ;in, Represents a real-time navigation trajectory data set; Represents the first The original data vector of sampling points; The total number of sampling points in the real-time navigation trajectory data set.

[0049] Standardize the historical navigation trajectory data set and define the standardization function , and get the normalized vector ;in, Represents the first The original data vector of the sampling points is normalized. Similarly, the real-time navigation trajectory data set is normalized to obtain the normalized vector ;in, Represents the first The original data vector after normalization of the sampling points.

[0050] Constructing a multi-dimensional feature sequence: standardizing the historical navigation trajectory data set Middle sampling points, whose timestamps are , and calculate the velocity vectors between adjacent points: ;in, Indicates The instantaneous velocity vector at each sampling point.

[0051] The historical navigation trajectory feature vector is constructed as: ;in, Represents the first The feature vector of sampling points; the vector dimension is , which is the original position dimension plus the velocity dimension and a single time quantity.

[0052] Similarly, for the standardized real-time navigation trajectory data set Construct feature vectors. Define the timestamp of each point in the standardized real-time navigation trajectory data set as , and calculate the velocity vector: ;in, Indicates The instantaneous velocity vector at each sampling point.

[0053] Construct the real-time navigation trajectory feature vector as: ;in, Represents the first The feature vector of sampling points; the vector dimension is , which is the original position dimension plus the velocity dimension and a single time quantity.

[0054] Select Euclidean distance as the sequence metric: To quantify the similarity between feature vectors, select Euclidean distance as the metric. Define the Euclidean distance function between the historical feature vector and the real-time feature vector as: ;in, Represents the feature vector With the eigenvector The Euclidean distance between Represents the feature vector In the Component values ​​on the dimension; Represents the feature vector In the Component values ​​on the dimension; represents the total dimension of the feature vector, which is .

[0055] Use the dynamic time warping algorithm to optimally align the trajectory sequence: The dynamic time warping algorithm is used to calculate the optimal alignment path of two trajectory sequences on the time axis. The constructed historical navigation trajectory feature vector sequence is defined as , the real-time navigation trajectory feature vector sequence is . Define the cumulative distance matrix (in , ), its recursive formula is: ;in, Indicates the cumulative distance matrix Row, No. The value of the column, i.e. the value of the The feature vector is the same as the feature vector of the real-time navigation trajectory. The minimum distance accumulated on the feature vector alignment path; represents the minimum cumulative distance selected from three possible paths; Represents the cumulative distance of the adjacent cells above; Represents the cumulative distance of the adjacent cells on the left; Represents the cumulative distance between adjacent cells on the diagonal.

[0056] The boundary conditions are usually , and for or Set to infinity to ensure the correctness of the recursive calculation.

[0057] Through a recursive formula, the dynamic time warping algorithm can find the alignment path that minimizes the cumulative distance, thereby quantifying the similarity between two trajectory sequences.

[0058] Calculate trajectory alignment error and output dynamic time warping matching score: cumulative distance matrix The end point Represents the total error on the optimal alignment path of two trajectory sequences. The length of the optimal alignment path is defined as , then the dynamic time warping matching score Defined as: ;in, represents the dynamic time warping matching score; represents the length of the optimal alignment path; The value representing the end point of the accumulation matrix.

[0059] Compare the dynamic time warping matching score with the matching score threshold to quantify the degree of deviation of the current navigation path: Preset match score threshold, compare the dynamic time warping match score with the preset match score threshold: When the dynamic time warping matching score is greater than or equal to the preset matching score threshold, it indicates that the deviation degree of the current navigation path is high, and the current navigation path deviates significantly from the pre-planned or historical reference path; When the dynamic time warping matching score is greater than or equal to the preset matching score threshold, it indicates that the deviation degree of the current navigation path is low, and the current navigation path is basically consistent with the pre-planned or historical reference path.

[0060] The preset matching score threshold is a positive real number determined by experimental data and expert experience.

[0061] Specifically, the waveguide effect level of the underground tunnel, the division result of the underground tunnel structure mutation area, and the deviation degree of the current navigation path are input into the deep belief network for comprehensive analysis to evaluate the reliability of the navigation control system and determine whether to trigger the path replanning instruction, including: Integrate the waveguide effect level of underground tunnels, the results of the underground tunnel structure mutation area division and the dynamic time warping matching score: Integrate the waveguide effect level of underground tunnels (including low waveguide effect and high waveguide effect), the results of the underground tunnel structure mutation area division (including low risk, medium risk and high risk areas) and the dynamic time warping matching score into a comprehensive input vector. Define the comprehensive input vector as: ;in, represents the comprehensive input vector; Indicates the waveguide effect level of the underground tunnel, and its value is 0 or 1, 0 represents a low waveguide effect level, and 1 represents a high waveguide effect level; It represents the result of the division of underground tunnel structure mutation area, and its value is integer 1, 2 or 3, corresponding to low risk area, medium risk area and high risk area respectively; Represents the dynamic time warping matching score.

[0062] Construct multi-channel input and feed it into deep belief network learning: construct the comprehensive input vector into a multi-channel input matrix. Define the constructed multi-channel input matrix as: ;in, Represents a multi-channel input matrix; Represents the transpose symbol.

[0063] The multi-channel input matrix is ​​used as the input layer data of the deep belief network, and transmits the waveguide effect level information of the underground tunnel, the underground tunnel structure mutation area division result information and the dynamic time warping matching score information through different channels.

[0064] Extract multi-source features in the hidden layer and identify potential associations: The deep belief network has a multi-layer structure, and its hidden layer can learn the potential features and association information in the input data. The hidden layer is used to perform nonlinear transformation on the multi-channel input matrix to extract high-level features. The output of the hidden layer is defined as: ;in, Represents the hidden layer output vector, which is an abstract expression of deep features; Represents the hidden layer weight matrix, which is used to linearly transform the multi-channel input matrix. The size of the matrix is , represents the number of neurons in the hidden layer, Indicates the number of input channels, equal to 3; represents the activation function; Represents the hidden layer bias vector, with dimension .

[0065] Through hidden layer calculation, the deep belief network can comprehensively analyze the intrinsic correlation between the waveguide effect level of the underground tunnel, the results of the underground tunnel structure mutation area division and the dynamic time warping matching score, and capture the interaction effect between different information in the high-dimensional feature space to form a deep feature representation with distinguishing ability.

[0066] Generate the reliability score of the navigation control system through network weight iteration: Utilize the multi-layer learning mechanism of the deep belief network, continuously adjust the network weights through the back propagation algorithm, optimize the mapping relationship between the hidden layer and the output layer, and thus generate the reliability score of the navigation control system. The calculation formula for defining the output layer is: ;in, represents the network layer output, i.e., the reliability score of the navigation control system; represents the output layer weight matrix; Represents the output layer bias term; represents the output activation function.

[0067] Preset the reliability score threshold and compare the reliability score with the reliability score threshold: When the reliability score is greater than the reliability score threshold, it means that the reliability of the navigation control system is high, and it is determined that the path replanning instruction is not triggered; When the reliability score is less than or equal to the reliability score threshold, it indicates that the reliability of the navigation control system is poor, and at this time, it is determined to trigger the path replanning instruction.

[0068] Embodiment 2: The difference between Example 2 of the present invention and Example 1 is that this example introduces a coal mine underground navigation control device based on big data.

[0069] A coal mine underground navigation control device based on big data, comprising one or more processors; further comprising a storage device for storing one or more programs; when the one or more programs are executed by one or more processors, the one or more processors implement a coal mine underground navigation control system based on big data.

[0070] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0071] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or may be transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium may be a solid-state hard disk.

[0072] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0073] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0074] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0075] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0076] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0077] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0078] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0079] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A coal mine underground navigation control system based on big data, characterized in that: It includes a waveguide effect judgment module, a structural mutation division module, an error level determination module, an offset degree assessment module and a reliability assessment module; The waveguide effect judgment module monitors the propagation characteristics of radio signals in underground tunnels in real time, uses wavelet transform combined with convolutional neural network to identify the reflection mode of radio signals, and judges the waveguide effect level of underground tunnels; The structural mutation division module performs multi-scale curvature analysis on the three-dimensional point cloud data of the underground tunnel and divides the structural mutation area of ​​the underground tunnel based on the DBSCAN clustering algorithm; The error level determination module determines the navigation error level of the underground tunnel based on the waveguide effect level of the underground tunnel and the result of the division of the underground tunnel structure mutation area; When the navigation error level reaches level 2 or 3, the deviation degree assessment module analyzes the historical navigation trajectory data set through the dynamic time warping algorithm to assess the deviation degree of the current navigation path; The reliability assessment module inputs the waveguide effect level of the underground tunnel, the division results of the underground tunnel structure mutation area, and the deviation degree of the current navigation path into the deep belief network for comprehensive analysis, evaluates the reliability of the navigation control system, and determines whether to trigger the path replanning instruction.

2. The coal mine underground navigation control system based on big data according to claim 1, characterized in that: By real-time monitoring of the radio signal propagation characteristics in the underground tunnel, the wavelet transform combined with the convolutional neural network is used to identify the radio signal reflection mode and determine the waveguide effect level of the current underground tunnel, specifically: Collect underground radio signals in real time to obtain amplitude, phase and delay data; Preprocess the collected data to reduce noise; Use wavelet transform to decompose radio signals in time and frequency and extract multi-scale features; Construct a multi-dimensional tensor and input the decomposition results into the convolutional neural network model; Use convolutional neural networks to identify signal reflection patterns and interference features; The level of waveguide effect in the underground tunnel is determined based on the identification results.

3. The coal mine underground navigation control system based on big data according to claim 2 is characterized in that: The waveguide effect level of the underground tunnel is determined based on the identification results, specifically: The waveguide effect level is determined based on the preset threshold contained in the decision function: When the arithmetic mean of the key characteristic values ​​is greater than or equal to the preset threshold, the waveguide effect level of the underground tunnel is determined to be a low waveguide effect level; When the arithmetic mean of the key characteristic values ​​is less than a preset threshold, the waveguide effect level of the underground tunnel is determined to be a high waveguide effect level.

4. The coal mine underground navigation control system based on big data according to claim 3 is characterized in that: Multi-scale curvature analysis is performed on the three-dimensional point cloud data of the underground tunnel, and the underground tunnel structure mutation area is divided based on the DBSCAN clustering algorithm, specifically: Collect 3D point cloud data; Select multiple layers of scales and calculate local curvature values; Extracting super-threshold curvature regions at each scale; Constructing multi-scale curvature features as clustering input vectors; The DBSCAN clustering algorithm was used to identify the structural mutation regions; Output the boundary of the underground tunnel structure mutation area; Provide risk warning for areas with sudden changes in underground tunnel structure.

5. The coal mine underground navigation control system based on big data according to claim 4, characterized in that: Provide risk warning for areas with sudden structural changes in underground tunnels, specifically: Preset low risk threshold and high risk threshold to compare with the warning index: When the warning index is less than or equal to the low risk threshold, the area with sudden changes in the underground tunnel structure is defined as a low risk area; When the warning index is greater than the low risk threshold and less than the high risk threshold, the underground tunnel structure mutation area is defined as a medium risk area; When the warning index is greater than or equal to the high-risk threshold, the area with sudden structural changes in the underground tunnel is defined as a high-risk area.

6. The coal mine underground navigation control system based on big data according to claim 5, characterized in that: Based on the waveguide effect level of the underground tunnel and the division results of the underground tunnel structure mutation area, the underground tunnel navigation error level is determined, specifically: When the waveguide effect level of the underground tunnel is a low waveguide effect level, and the warning index is less than or equal to the low risk threshold, the underground tunnel navigation error level is determined to be level one; When the waveguide effect level of the underground tunnel is a high waveguide effect level, and the warning index is greater than the low risk threshold and less than the high risk threshold, the underground tunnel navigation error level is determined to be level 2; When the waveguide effect level of the underground tunnel is a high waveguide effect level and the warning index is greater than or equal to the high risk threshold, the underground tunnel navigation error level is determined to be level three.

7. The coal mine underground navigation control system based on big data according to claim 6, characterized in that: The historical navigation trajectory data set is analyzed through the dynamic time warping algorithm to evaluate the deviation degree of the current navigation path, specifically: Standardized historical navigation trajectory data set and real-time navigation trajectory data set; Construct multi-dimensional feature sequences; Choose Euclidean distance as the sequence metric; Use dynamic time warping algorithm to optimally align trajectory sequences; Calculate the trajectory alignment error and output the dynamic time warping matching score; Compare the matching score with the matching score threshold to quantify the degree of deviation of the current navigation path.

8. The coal mine underground navigation control system based on big data according to claim 7, characterized in that: Compare the matching score with the matching score threshold to quantify the current navigation path deviation degree, specifically: Preset match score threshold, compare the dynamic time warping match score with the preset match score threshold: When the dynamic time warping matching score is greater than or equal to the preset matching score threshold, it indicates that the deviation degree of the current navigation path is high; When the dynamic time warping matching score is greater than or equal to the preset matching score threshold, it indicates that the deviation degree of the current navigation path is low.

9. The coal mine underground navigation control system based on big data according to claim 8, characterized in that: The waveguide effect level of the underground tunnel, the division result of the underground tunnel structure mutation area, and the deviation degree of the current navigation path are input into the deep belief network for comprehensive analysis to evaluate the reliability of the navigation control system and determine whether to trigger the path replanning instruction. Specifically: Integrate the waveguide effect level of the underground tunnel, the results of the underground tunnel structure mutation area division and the dynamic time warping matching score; Construct multi-channel input and feed it into deep belief network learning; Extract multi-source features in the hidden layer and identify potential correlations; Through network weight iteration, a reliability score of the navigation control system is generated.

10. The coal mine underground navigation control system based on big data according to claim 9, characterized in that: Determine whether to trigger the path replanning instruction, specifically: Preset the reliability score threshold and compare the reliability score with the reliability score threshold: When the reliability score is greater than the reliability score threshold, it indicates that the reliability of the navigation control system is high, and it is determined that the path replanning instruction is not triggered; When the reliability score is less than or equal to the reliability score threshold, it indicates that the reliability of the navigation control system is poor, and a path replanning instruction is triggered.

11. A coal mine underground navigation control device based on big data, characterized in that: include: one or more processors; A storage device for storing one or more programs; When one or more programs are executed by one or more processors, the one or more processors implement a coal mine underground navigation control system based on big data as described in any one of claims 1 to 10.

Images