Roof separation instrument alarm device and coal mine roof monitoring system

Abstract

The invention relates to the technical field of coal mine safety, in particular to a roof separation instrument alarm device and a coal mine roof monitoring system.The roof separation instrument alarm device comprises a data acquisition module, a topological feature extraction module, a dynamic feature analysis module, a group theory analysis module and an alarm module, the data acquisition module is in communication connection with the dynamic feature analysis module, the chaotic feature extraction module is in communication connection with the group theory analysis module, and the comprehensive early warning module is in communication connection with the chaotic feature extraction module and the data acquisition module. When the comprehensive early warning index exceeds a preset threshold value, an alarm signal is triggered, meanwhile, displacement and stress data of the roof are monitored, a topological feature extraction technology is introduced, and the system can comprehensively grasp deformation features and mechanical behaviors of the roof. The multi-dimensional analysis method not only improves the monitoring comprehensiveness, but also can effectively reduce misjudgment caused by local abnormality, and greatly improves the early warning accuracy of the system.

Description

Technical Field

[0001] The present invention relates to the technical field of coal mine safety, and more specifically to a roof separation instrument alarm device and a coal mine roof monitoring system. Background Art

[0002] With the continuous increase in the depth of coal mining, the safety problem of coal mine roof has become increasingly prominent and has become one of the key factors affecting coal mine safety production. Roof separation accidents not only threaten the lives of miners, but may also cause huge economic losses. Therefore, the development of an efficient and reliable roof monitoring system is of great significance to improving the level of coal mine safety production.

[0003] At present, coal mine roof monitoring technology mainly includes stress monitoring, displacement monitoring and acoustic emission monitoring. The traditional stress monitoring system installs stress sensors on the roof to monitor the stress changes of the roof in real time. This method is simple to operate, but it is difficult to reflect the overall deformation state of the roof, and it is easily affected by local stress concentration, resulting in false alarms or missed reports. The displacement monitoring system evaluates the stability of the roof by measuring the displacement changes of the roof. Although this method can intuitively reflect the deformation of the roof, it is often difficult to capture small deformation trends in time, thus affecting the timeliness of the early warning. Acoustic emission monitoring technology judges the state of the roof by analyzing the sound wave signal generated when the roof rock layer breaks. It has a certain early warning capability, but in a complex mine environment, the signal is easily interfered with, affecting the accuracy of monitoring.

[0004] In addition, the existing roof monitoring systems generally have the following problems: First, most systems rely on only a single physical quantity for monitoring, which makes it difficult to fully reflect the complex mechanical behavior of the roof. Second, the data analysis method is relatively simple, usually using threshold judgment or simple statistical analysis, which cannot deeply mine the rich information contained in the data. Furthermore, the early warning model lacks dynamic adaptability and is difficult to cope with changes in different geological conditions and mining stages. Finally, the reliability and anti-interference ability of the system need to be improved, and false alarms or missed alarms are prone to occur in complex underground environments. Summary of the invention

[0005] In view of the above problems, the present invention proposes an innovative roof separation instrument alarm device and a coal mine roof monitoring system. The system aims to achieve accurate monitoring and reliable early warning of the roof state through multi-dimensional data collection, advanced mathematical models and intelligent algorithms, thereby effectively improving the safety level of coal mine production.

[0006] The present invention provides a method for intelligent procurement and inventory early warning of catering raw materials, comprising:

[0007] Data acquisition module for:

[0008] Collect roof displacement data;

[0009] Obtain top plate stress data;

[0010] A topological feature extraction module is connected to the data acquisition module for:

[0011] Receiving top plate displacement data sent by the data acquisition module;

[0012] Based on the roof displacement data, construct a topological structure of the roof displacement field;

[0013] Performing spectral decomposition on the top plate displacement field topology to obtain Laplace matrix eigenvalues;

[0014] The dynamic feature analysis module is in communication with the topological feature extraction module and is used to:

[0015] Receiving the Laplace matrix eigenvalues ​​sent by the topological feature extraction module;

[0016] Based on the Laplace matrix eigenvalues, construct a dynamic feature matrix;

[0017] The group theory analysis module is connected to the dynamic feature analysis module for:

[0018] Receiving the dynamic feature matrix sent by the dynamic feature analysis module;

[0019] The dynamic characteristic matrix is ​​regarded as an element of a permutation group, and its group properties are analyzed;

[0020] Compute invariant subgroups;

[0021] The chaos feature extraction module is connected to the group theory analysis module for:

[0022] Receiving the invariant subgroup information sent by the group theory analysis module;

[0023] Based on the invariant subgroup information, construct a Lyapunov exponent spectrum;

[0024] The comprehensive early warning module is connected to the chaos feature extraction module and the data acquisition module for:

[0025] Receiving the Lyapunov exponent spectrum sent by the chaos feature extraction module;

[0026] Receiving the top plate stress data sent by the data acquisition module;

[0027] Calculating a comprehensive early warning index based on the Lyapunov exponent spectrum, the invariant subgroup information and the roof stress data;

[0028] When the comprehensive warning index exceeds a preset threshold, an alarm signal is triggered.

[0029] Preferably, the topological feature extraction module comprises:

[0030] Topology building blocks for:

[0031] Based on the roof displacement data, a roof displacement field topology structure including a set of measuring points, connection relationships between measuring points and edge weight functions is constructed;

[0032] A Laplace matrix generating unit is communicatively connected with the topology building unit and is used for:

[0033] Based on the topology of the top plate displacement field, generating a corresponding Laplace matrix;

[0034] An eigenvalue calculation unit is communicatively connected to the Laplace matrix generation unit and is used for:

[0035] Perform eigenvalue decomposition on the Laplace matrix to obtain eigenvalues ​​of the Laplace matrix.

[0036] Preferably, the dynamic feature analysis module comprises:

[0037] Time series building blocks for:

[0038] Construct time series based on the eigenvalues ​​of the Laplace matrix at different times;

[0039] A matrix generating unit is communicatively connected with the time series building unit and is used for:

[0040] The time series are organized into a dynamic feature matrix.

[0041] Preferably, the group theory analysis module includes:

[0042] Permutation group building blocks for:

[0043] Treating the dynamic characteristic matrix as an element of a permutation group, defining group operations;

[0044] an invariant subgroup calculation unit, communicatively connected to the permutation group construction unit, and configured to:

[0045] Based on the permutation group, an invariant subgroup is computed.

[0046] Preferably, the chaos feature extraction module comprises:

[0047] Perturbation vector generation unit, used to:

[0048] Based on the invariant subgroup information, generating a disturbance vector;

[0049] A Lyapunov exponent calculation unit is communicatively connected with the disturbance vector generation unit and is used for:

[0050] Computes the Lyapunov exponent spectrum.

[0051] Preferably, the comprehensive early warning module comprises:

[0052] Characteristic evaluation unit for:

[0053] The Laplace matrix eigenvalues, invariant subgroup information and Lyapunov index spectrum are evaluated respectively;

[0054] Weight distribution unit, used to:

[0055] Assigning weight coefficients to each evaluation result;

[0056] An index calculation unit, which is in communication with the feature evaluation unit and the weight allocation unit, is used to:

[0057] Based on each assessment result and the corresponding weight coefficient, a comprehensive early warning index is calculated.

[0058] Preferably, the data acquisition module, the topological feature extraction module, the dynamic feature analysis module, the group theory analysis module, the chaos feature extraction module and the comprehensive early warning module exchange data through an Ethernet switch, and the Ethernet switch adopts a redundant link topology structure to improve the reliability and fault tolerance of the system.

[0059] A coal mine roof monitoring system, comprising the roof separation instrument alarm device, further comprising:

[0060] The data storage module is connected to the data acquisition module of the roof separation instrument alarm device for:

[0061] Receiving and storing the top plate displacement data and top plate stress data sent by the data acquisition module;

[0062] The visualization module is communicatively connected with the comprehensive early warning module of the roof separation instrument alarm device, and is used for:

[0063] Receiving the comprehensive warning index sent by the comprehensive warning module;

[0064] Generate a visual interface for roof status;

[0065] The remote control module is connected to the integrated early warning module of the roof separation instrument alarm device for:

[0066] Receiving the alarm signal sent by the comprehensive early warning module;

[0067] Send control instructions to the coal mine safety management system.

[0068] Preferably, the visualization module comprises:

[0069] 3D modeling unit for:

[0070] Construct a three-dimensional tunnel model based on the coal mine tunnel structure data;

[0071] A state mapping unit is connected to the three-dimensional modeling unit for:

[0072] Mapping the comprehensive early warning index onto the three-dimensional lane model;

[0073] An interface rendering unit, which is in communication with the state mapping unit, is used to:

[0074] Rendering top plate status visualization interface.

[0075] Preferably, the remote control module comprises:

[0076] Instruction generation unit, used to:

[0077] Based on the alarm signal, generating corresponding control instructions;

[0078] A communication interface unit, communicatively connected to the instruction generation unit, is used to:

[0079] The control instruction is sent to the coal mine safety management system.

[0080] The beneficial effects of the present invention are mainly reflected in the following aspects:

[0081] The roof separation instrument alarm device and coal mine roof monitoring system of the present invention have achieved significant improvements in many aspects through innovative technical solutions, and have effectively solved the problems existing in the prior art, which are specifically manifested in the following aspects:

[0082] First, the present invention adopts a multi-dimensional and multi-scale data collection and analysis method. By simultaneously monitoring the displacement and stress data of the roof and introducing topological feature extraction technology, the system can fully grasp the deformation characteristics and mechanical behavior of the roof. This multi-dimensional analysis method not only improves the comprehensiveness of monitoring, but also effectively reduces the misjudgment caused by local anomalies, greatly improving the system's early warning accuracy.

[0083] Secondly, the present invention innovatively introduces group theory and chaos theory into the field of roof monitoring. By constructing a dynamic feature matrix and performing group theory analysis, the system can deeply explore the inherent laws of roof deformation. At the same time, the chaos feature extraction technology enables the system to capture slight changes in the roof state, providing the possibility for early warning. The application of these advanced mathematical tools has significantly improved the timeliness and accuracy of the system in terms of warning.

[0084] Furthermore, the present invention designs a comprehensive early warning module, which realizes intelligent early warning decision-making by integrating multiple characteristic indicators and dynamically adjusting weights. This method not only improves the reliability of early warning, but also gives the system a certain degree of adaptive ability, which can better cope with changes in different geological conditions and mining stages.

[0085] In addition, the present invention is also innovative in system architecture design. The use of distributed network structure and redundant link topology greatly improves the reliability and anti-interference ability of the system. At the same time, through optimized parallel computing and network transmission strategies, the real-time performance of complex mathematical models in practical applications is ensured.

[0086] Finally, the invention also integrates 3D visualization and remote control functions, greatly improving the operability and practicality of the system. The intuitive 3D display enables workers to quickly identify potential danger areas, while the remote control function provides technical support for rapid response in emergency situations.

[0087] In summary, the present invention has achieved a qualitative leap in the accuracy, timeliness, reliability and practicality of roof monitoring through a number of technological innovations. These improvements can not only better protect the lives of miners, but also create economic value for coal mining enterprises by reducing false alarms and improving production efficiency. In particular, the significant improvement in the early warning time provides a critical time window for preventing roof accidents, which has important practical significance and application value in actual coal mine safety production. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] Figure 1 The present invention is a logic block diagram of the roof separation instrument alarm device.

[0089] Figure 2 It is a logic block diagram of the coal mine roof monitoring system of the present invention.

[0090] Figure 3 This is a graph showing how the topological features of the present invention change over time.

[0091] Figure 4 This is a graph showing how the group invariance of the present invention changes over time.

[0092] Figure 5 is the Lyapunov index spectrum of the present invention.

[0093] Figure 6 It is the comprehensive early warning index heat map of the present invention. DETAILED DESCRIPTION

[0094] Please refer to Figure 1-2The present invention provides a roof separation instrument alarm device and a coal mine roof monitoring system, which realizes accurate monitoring and early warning of the coal mine roof state through innovative mathematical models and algorithms. The present invention will be described in detail below in conjunction with specific implementation methods.

[0095] The roof separation instrument alarm device of the present invention comprises a data acquisition module 1, a topological feature extraction module 2, a dynamic feature analysis module 3, a group theory analysis module 4, a chaos feature extraction module 5 and a comprehensive early warning module 6. These modules work together to form a complete roof monitoring and early warning system.

[0096] The data acquisition module 1 is responsible for collecting the roof displacement data and the roof stress data. Preferably, the present invention uses a fiber Bragg grating sensor network to collect these data. The fiber Bragg grating sensor has the advantages of high precision and anti-electromagnetic interference, and is particularly suitable for use in complex coal mine environments. In one embodiment, the sampling frequency of the sensor can be set to 10 Hz, so that it can capture the slight changes in the roof without generating too much redundant data.

[0097] The topological feature extraction module 2 receives the top plate displacement data sent by the data acquisition module 1, and constructs the topological structure of the top plate displacement field based on these data. The core of this step is to convert the discrete displacement data into a topological structure with a spatial relationship. Specifically, the present invention adopts the following method to construct the topological structure:

[0098]

[0099] in, is the topological structure of the roof displacement field, which is used to represent the spatial distribution of roof deformation. V is the set of measuring points, which is used to store the position information of each measuring point. E is the connection relationship between measuring points, which is used to describe the adjacency relationship between measuring points. ω is the edge weight function, which is defined as the displacement difference between adjacent measuring points, which is used to quantify the relative movement between measuring points and reflect the local characteristics of roof deformation.

[0100] This formula defines a topological structure of the displacement field of the top plate Where V is the set of measuring points, E is the connection relationship between the measuring points, and ω is the edge weight function. The core idea of ​​the topological structure is to transform discrete displacement data into a graph structure with spatial relationships, so as to capture the spatial distribution characteristics of roof deformation. By constructing a topological structure, the system can more intuitively understand the deformation of the roof at different locations. In particular, the edge weight function ω is defined as the displacement difference between adjacent measuring points, which enables the system to quantify the degree of deformation in local areas and help identify potential dangerous areas. For example, if the displacement difference between two measuring points suddenly increases, it may mean that there is a large stress concentration or crack expansion in the roof of the area, and timely warning is required.

[0101] The data acquisition module 1 collects the roof displacement data and stress data in real time through the fiber grating sensor network. These sensors are distributed in different locations in the coal mine tunnels to form a dense monitoring network. The displacement data recorded by each sensor is transmitted to the topological feature extraction module 2. The topological feature extraction module 2 first pre-processes the collected displacement data to remove noise and outliers. Then, based on the Delaunay triangulation algorithm, the connection relationship E between the measuring points is determined, and the displacement difference between adjacent measuring points is calculated as the edge weight ω. Finally, a complete topological structure of the roof displacement field is constructed.

[0102] Next, the topological feature extraction module 2 performs spectral decomposition on the topological structure of the top plate displacement field to obtain the Laplace matrix eigenvalues. The calculation method is as follows:

[0103]

[0104] Where D is the degree matrix, which is used to record the number of connections of each measuring point. A is the adjacency matrix, which is used to represent the connection strength between measuring points. By solving the characteristic equation We can obtain the eigenvalues ​​λ of the Laplace matrix. These eigenvalues ​​contain important topological information of the top plate displacement field and reflect the stability and deformation mode of the system.

[0105] Laplacian Matrix From the topology It is an important matrix derived from the Laplacian matrix, which consists of the degree matrix D and the adjacency matrix A. The diagonal elements of the degree matrix D represent the number of connections of each measurement point, while the adjacency matrix A represents the connection strength between the measurement points. The eigenvalues ​​and eigenvectors of the Laplacian matrix can reflect the stability, connectivity and local structural characteristics of the system.

[0106] The eigenvalue analysis of the Laplace matrix can reveal the intrinsic topological information of the roof displacement field. For example, smaller eigenvalues ​​usually correspond to the global mode of the system, while larger eigenvalues ​​reflect local detail changes. By analyzing the distribution of eigenvalues, the system can identify stable and unstable areas of the roof, thereby providing a basis for early warning. In addition, the spectral decomposition of the Laplace matrix can also be used for dimensionality reduction and data compression to reduce the complexity of subsequent calculations.

[0107] After the topological structure is constructed, the Laplace matrix generation unit 22 calculates the adjacency matrix A and the degree matrix D according to the connection relationship E and the edge weight ω in the topological structure. The elements of these two matrices come directly from the measurement point connection information and displacement difference in the topological structure. The Laplace matrix generation unit 22 first constructs the adjacency matrix A, where A ijrepresents the connection strength between measurement point i and measurement point j (i.e., edge weight ω). Then, the degree matrix D is calculated, and its diagonal element D ii is equal to the number of connections of measuring point i. Finally, through Calculate the Laplacian matrix.

[0108] The dynamic feature analysis module 3 receives the Laplace matrix eigenvalues ​​sent by the topological feature extraction module 2, and constructs a dynamic feature matrix based on these eigenvalues. Specifically, the construction method of the dynamic feature matrix D is as follows:

[0109]

[0110] Among them, λ i (t j ) is the time t j The ith eigenvalue of is used to record the change of eigenvalues ​​over time. m is the number of eigenvalues, indicating the number of different modes in the system. n is the length of the time series, which is used to record the time evolution of the eigenvalues. In practical applications, n = 100 can be selected, that is, the data of the most recent 100 time points can be retained, which can reflect the dynamic changes of the roof state without occupying too much computing resources.

[0111] The dynamic feature matrix D is a time series matrix, in which each row represents the time evolution of an eigenvalue and the columns represent different time points. By recording the changes in the eigenvalues ​​of the Laplace matrix over time, the system can capture the dynamic characteristics of roof deformation. This time series analysis can help identify long-term trends and short-term fluctuations in roof deformation. The construction of the dynamic feature matrix enables the system to track the evolution of the roof state and identify potential danger signals. For example, if there is a clear upward trend in the time series of a certain eigenvalue, it may mean that the stability of the roof is gradually deteriorating and early warning is needed. In addition, the dynamic feature matrix can also be used to detect periodic changes in roof deformation and help predict future deformation trends.

[0112] The dynamic feature analysis module 3 receives the Laplace matrix eigenvalues ​​from the topological feature extraction module 2 and arranges them in chronological order. The time series data of each eigenvalue comes from continuous monitoring at multiple time points, ensuring the integrity and continuity of the data. The time series construction unit 31 uses a sliding window technology to record the changes of each eigenvalue at continuous time points to form a time series. For example, a 60-second sliding window can be selected to ensure that the rapid changes in the top plate state can be captured without affecting the calculation efficiency due to the large amount of data. Then, the matrix generation unit 32 organizes these time series into a dynamic feature matrix D.

[0113] The group theory analysis module 4 receives the dynamic feature matrix sent by the dynamic feature analysis module 3, and regards it as an element of the permutation group to analyze its group properties. The present invention innovatively introduces group theory analysis, which enables us to study the inherent law of top plate deformation from an algebraic perspective. Specifically, we define the permutation group G as follows:

[0114]

[0115] G is a permutation group, which is used to represent the algebraic structure of the dynamic characteristic matrix. D is a dynamic characteristic matrix, which is used to store the time series of eigenvalues. is a matrix multiplication operation, which is used to define the operation rules in the group. The core task of group theory analysis module 4 is to calculate the invariant subgroup H:

[0116]

[0117] H is the invariant subgroup, which represents the set of elements that remain unchanged under the action of the group, reflecting the stability characteristics of the system. g is any element in the group G, representing a state of the system. ghg -1 is a conjugate operation used to test the invariance of element h under the action of the group. The structure of the invariant subgroup H reflects the stability characteristics of the top plate displacement field. For example, if the order of H is small, this may mean that the top plate state is unstable and needs close attention.

[0118] Invariant subgroup H is an important concept in group theory, which refers to a set of elements that remain unchanged under the action of a group. In the present invention, group G is a permutation group consisting of a dynamic characteristic matrix D, and the group operation ° is defined as matrix multiplication. The calculation of the invariant subgroup is performed by the conjugate operation ghg -1 to check whether the element h remains unchanged under the action of the group. The existence of the invariant subgroup reflects the symmetry and stability of the system. The calculation of the invariant subgroup can help the system identify the inherent laws of roof deformation. For example, if the order of the invariant subgroup is small, it may mean that the state of the roof is unstable and there is a greater risk of deformation. By analyzing the structure of the invariant subgroup, the system can better understand the deformation pattern of the roof, predict future deformation trends, and provide a basis for early warning.

[0119] The group theory analysis module 4 receives the dynamic feature matrix D from the dynamic feature analysis module 3 and regards it as an element of the permutation group G. Each matrix D corresponds to a system state and describes the deformation characteristics of the top plate at a certain moment. The invariant subgroup calculation unit 42 first defines the group operation ° as matrix multiplication, and then uses the conjugate operation ghg -1 Check whether each element h remains unchanged under the group action. Finally, calculate the invariant subgroup H and analyze its structure and properties.

[0120] The chaos feature extraction module 5 receives the invariant subgroup information sent by the group theory analysis module 4, and constructs the Lyapunov index spectrum based on it. The Lyapunov index is an important indicator for describing the degree of chaos in a dynamic system. The present invention calculates the Lyapunov index using the following method:

[0121]

[0122] L i is the i-th Lyapunov index, which is used to quantify the degree of chaos in the system. i (t) is the disturbance vector at time t, which represents the state deviation of the system at that time. i (0) is the initial perturbation vector, which represents the initial state deviation of the system. In practical applications, we can choose to calculate the first three largest Lyapunov exponents, which are usually sufficient to reflect the dynamic characteristics of the system.

[0123] Lyapunov exponent L i It is an important indicator to measure the degree of chaos in a dynamic system, indicating the separation speed of the system in the phase space. i (0) represents a small deviation of the system at the initial moment, while δz i (t) represents the evolution of the deviation over time t. The larger the value of the Lyapunov exponent, the higher the degree of chaos in the system, indicating that the long-term behavior of the system is difficult to predict. The calculation of the Lyapunov exponent can help the system identify the nonlinear dynamic characteristics of the roof deformation. For example, if some Lyapunov exponents are large, it may mean that the deformation of the roof is highly chaotic and unpredictable. By analyzing the distribution of the Lyapunov exponent, the system can identify the stable and unstable areas of the roof and help predict future deformation trends.

[0124] The chaos feature extraction module 5 receives the invariant subgroup information from the group theory analysis module 4 and generates the initial disturbance vector δz based on it. i (0). The perturbation vector is generated by δz i (0) = ∈v i , where ∈ is a small positive constant representing the amplitude of the disturbance, v i is the i-th eigenvector of the invariant subgroup H. The Lyapunov exponent calculation unit 52 first generates an initial perturbation vector δz i (0), and then calculate the disturbance vector δz through numerical simulation i (t) evolves with time t. Finally, through the formula Calculate the Lyapunov exponent L i .

[0125] Finally, the comprehensive warning module 6 receives the Lyapunov exponent spectrum sent by the chaos feature extraction module 5 and the roof stress data sent by the data acquisition module 1, and calculates the comprehensive warning index based on this information. The calculation method of the comprehensive warning index is as follows:

[0126] W=α 1 f 1 (λ)+α 2 f 2 (H)+α 3 f 3 (Li)+α 4 f 4 (σ)

[0127] W is a comprehensive early warning index, which is used to evaluate the safety status of the roof. 1 (λ) is the evaluation function of the Laplace matrix eigenvalue, which is used to quantify the topological structure change of the system. 2 (H) is the evaluation function of the invariant subgroup, which is used to quantify the stability of the system. 3 (L i ) is the evaluation function of the Lyapunov index, which is used to quantify the degree of chaos in the system. 4 (σ) is the evaluation function of the top plate stress, which is used to quantify the mechanical state of the top plate. 1 ,α 2 ,α 3 ,α 4 is the weight coefficient, which is used to adjust the contribution of each evaluation function. σ is the top plate stress, which is used to directly reflect the stress condition of the top plate.

[0128] The comprehensive early warning index W is a weighted summation result, combining multiple evaluation functions f 1 ,f 2 ,f 3 ,f 4 The output of each evaluation function corresponds to different features: Laplace matrix eigenvalue λ, invariant subgroup H, Lyapunov index L i and top plate stress σ. Weight coefficient α 1 ,α 2 ,α 3 ,α 4 Used to adjust the importance of each feature. The calculation of the comprehensive early warning index combines multi-dimensional monitoring data to form a comprehensive early warning index. By setting the weight coefficient reasonably, the system can highlight the importance of certain features according to actual conditions. For example, when the roof stress suddenly increases, α can be appropriately increased. 4 , in order to improve the system's sensitivity to emergencies. The calculation of the comprehensive early warning index provides a scientific basis for the system's decision-making, helping miners take timely measures to avoid safety accidents.

[0129] The comprehensive warning module 6 receives the Lyapunov index spectrum from the chaos feature extraction module 5, the invariant subgroup information from the group theory analysis module 4, the Laplace matrix eigenvalues ​​from the dynamic feature analysis module 3, and the roof stress data from the data acquisition module 1. After these data are processed by their respective evaluation functions, corresponding evaluation results are generated. The feature evaluation unit 61 first evaluates the Laplace matrix eigenvalues, invariant subgroups, Lyapunov indexes, and roof stress, and generates an evaluation function f 1 ,f 2 ,f 3 ,f 4 Then, the weight allocation unit 62 dynamically adjusts the weight coefficient α according to the expert experience and the real-time status of the system. 1 ,α 2 ,α 3 ,α 4 Finally, the index calculation unit 63 performs weighted summation of the evaluation results to calculate the comprehensive warning index W.

[0130] The choice of weight coefficient can be adjusted according to the actual situation. For example, α can be set 1 =0.3,α 2 =0.2,α 3 =0.3,α 4 =0.2. When W exceeds the preset threshold, the comprehensive warning module 6 will trigger an alarm signal. According to experience, the warning threshold can be set to 0.8, which shows good sensitivity and reliability in practical applications.

[0131] The topological feature extraction module 2 of the present invention comprises a topological construction unit 21 , a Laplacian matrix generation unit 22 and an eigenvalue calculation unit 23 .

[0132] The topology construction unit 21 constructs a topology structure of the top plate displacement field including a set of measuring points, connection relationships between measuring points, and edge weight functions based on the top plate displacement data. In practical applications, the Delaunay triangulation algorithm can be used to determine the connection relationship between measuring points. This method can effectively reflect the local characteristics of the top plate deformation.

[0133] The Laplace matrix generation unit 22 is in communication connection with the topology construction unit 21, and generates a corresponding Laplace matrix based on the topology structure of the top plate displacement field. The generation process of the Laplace matrix is ​​as described above, and will not be repeated here.

[0134] The eigenvalue calculation unit 23 is in communication connection with the Laplace matrix generation unit 22, and performs eigenvalue decomposition on the Laplace matrix to obtain the eigenvalue of the Laplace matrix. Preferably, the Lanczos algorithm can be used for eigenvalue calculation, which has good calculation efficiency for large-scale sparse matrices.

[0135] Through the above detailed description, the working principle and specific implementation method of the roof separation instrument alarm device and the coal mine roof monitoring system of the present invention have been fully revealed. The system realizes accurate monitoring and early warning of the roof state through innovative mathematical models and algorithms, which can effectively improve the safety production level of coal mines.

[0136] The dynamic feature analysis module 3 of the present invention comprises a time series construction unit 31 and a matrix generation unit 32. These two units work together to convert discrete feature value data into a matrix reflecting the dynamic characteristics of the system.

[0137] The main function of the time series construction unit 31 is to construct a time series based on the eigenvalues ​​of the Laplace matrix at different times. Specifically, for each eigenvalue, the time series construction unit 31 will record its changes at consecutive time points. Preferably, the present invention adopts a sliding window technology to construct a time series, and the window size can be adjusted according to actual needs. For example, in one embodiment, a 60-second sliding window can be selected, which can capture the rapid changes in the top plate state without affecting the calculation efficiency due to the large amount of data.

[0138] The matrix generation unit 32 is connected to the time series construction unit 31 in communication, and its main task is to organize the time series into a dynamic feature matrix. This process can be expressed as:

[0139] D=[λ 1 ,λ 2 ,...,λ m ] T ,

[0140] Where D is a dynamic feature matrix, which is used to store the time series of eigenvalues. i is the time series of the i-th eigenvalue, indicating the change of the eigenvalue over time. m is the number of eigenvalues, indicating the number of different modes in the system. In this way, we obtain a two-dimensional matrix containing time and space information, which lays the foundation for subsequent group theory analysis.

[0141] It is worth noting that in the process of constructing the dynamic feature matrix D, data missing or abnormal situations may be encountered. In order to deal with these problems, in a preferred embodiment of the present invention, the matrix generation unit 32 also includes a data preprocessing function. For example, interpolation can be used to fill in missing data, or methods such as median filtering can be used to remove abnormal values ​​to ensure the quality and reliability of the generated dynamic feature matrix.

[0142] The group theory analysis module 4 of the present invention includes a permutation group construction unit 41 and an invariant subgroup calculation unit 42. The introduction of these two units enables the present invention to deeply analyze the internal law of top plate deformation from an algebraic perspective, which is a major innovation of the present invention.

[0143] The core task of the permutation group construction unit 41 is to regard the dynamic characteristic matrix as an element of the permutation group and define the group operation. In one embodiment of the present invention, the group operation is defined as matrix multiplication, that is:

[0144]

[0145] Among them, g 1 and g 2 is an element in group G, representing the state of the system at different time points. It is a group operation, defined as matrix multiplication, which is used to describe the transformation relationship between system states. The advantage of this definition is that it maintains the time series relationship in the dynamic characteristic matrix and introduces an algebraic structure, allowing us to use the powerful tools of group theory to analyze system characteristics.

[0146] The invariant subgroup calculation unit 42 is communicatively connected to the permutation group construction unit 41, and its main function is to calculate the invariant subgroup based on the constructed permutation group.

[0147] In practical applications, the calculation of invariant subgroups may involve a large number of matrix operations. In order to improve the computational efficiency, a parallel computing technique is used in the preferred embodiment of the present invention. Specifically, the group elements can be distributed to multiple processing units and the invariance test can be performed simultaneously, which can significantly reduce the computational time, especially when processing large-scale data.

[0148] The chaos feature extraction module 5 of the present invention comprises a disturbance vector generation unit 51 and a Lyapunov exponent calculation unit 52. The design of these two units aims to capture the nonlinear dynamic characteristics of the roof system and provide an important basis for early warning.

[0149] The main task of the disturbance vector generation unit 51 is to generate a disturbance vector based on the invariant subgroup information. In one embodiment of the present invention, the method for generating the disturbance vector is as follows:

[0150] δz i (0) = ∈v i ,

[0151] Among them, δz i (0) is the initial perturbation vector, which indicates a small deviation of the system in the initial state. ∈ is a small positive constant (e.g. 10 -6 ), used to control the amplitude of the disturbance. iis the i-th eigenvector of the invariant subgroup H, representing the invariant mode of the system. The advantage of this method is that it takes into account the invariant characteristics of the system, making the generated perturbation vector more representative.

[0152] The Lyapunov exponent calculation unit 52 is connected to the disturbance vector generation unit 51 in communication, and its core function is to calculate the Lyapunov exponent spectrum. The calculation process of the Lyapunov exponent is as described above and will not be repeated here. It is worth mentioning that in practical applications, we usually only need to calculate the largest Lyapunov exponents. For example, in a preferred embodiment of the present invention, the first three largest Lyapunov exponents can be calculated, which is sufficient to reflect the main dynamic characteristics of the system.

[0153] The comprehensive early warning module 6 of the present invention comprises a feature evaluation unit 61, a weight distribution unit 62 and an index calculation unit 63. These three units work together to convert multi-dimensional monitoring data into intuitive early warning indexes, which is the decision-making core of the system of the present invention.

[0154] The main function of the feature evaluation unit 61 is to evaluate the Laplace matrix eigenvalues, the invariant subgroup information and the Lyapunov index spectrum respectively. In one embodiment of the present invention, the evaluation function can be designed as follows:

[0155]

[0156] Among them, λ 0 , H 0 and L i0 They are the reference values ​​under normal conditions, used to compare the difference between the current state and the normal state. 1 , β 2 and β 3 It is a tuning parameter used to adjust the sensitivity of the evaluation function. ||λ-λ 0 || 2 Represents the square difference between the eigenvalue and the reference value, used to quantify the change in the eigenvalue. |H|-|H 0 | represents the difference between the order of the invariant subgroup and the reference value, which is used to quantify the change in stability of the system. It represents the sum of the absolute differences between the first three Lyapunov exponents and the reference value, and is used to quantify the change in the degree of chaos of the system. The design of these evaluation functions takes into account the physical meaning of each feature and can effectively quantify the abnormal degree of the roof state.

[0157] The role of the weight allocation unit 62 is to allocate weight coefficients to each evaluation result. The allocation of weights can be based on expert experience or can be automatically optimized by a machine learning method. In a preferred embodiment of the present invention, an adaptive weight allocation strategy can be adopted, i.e., the weights are dynamically adjusted according to the real-time state of the system. For example, when a sudden increase in the Lyapunov exponent is detected, the corresponding weight can be appropriately increased to improve the sensitivity of the system to emergencies.

[0158] The index calculation unit 63 is in communication connection with the feature evaluation unit 61 and the weight allocation unit 62, and its main task is to calculate the comprehensive warning index based on each evaluation result and the corresponding weight coefficient. The calculation formula of the comprehensive warning index is as described above and will not be repeated here. It should be emphasized that the setting of the warning threshold is crucial to the actual effect of the system. In the implementation process of the present invention, it is recommended to determine the optimal threshold through analysis of a large amount of historical data and field testing. For example, the threshold can be initially set to 0.7, and then fine-tuned according to the false alarm rate and missed alarm rate of the system, and finally a balance point is found.

[0159] The data acquisition module 1, topological feature extraction module 2, dynamic feature analysis module 3, group theory analysis module 4, chaos feature extraction module 5 and comprehensive early warning module 6 in the roof separation instrument alarm device of the present invention exchange data through an Ethernet switch. This network architecture design greatly improves the data transmission efficiency and reliability of the system.

[0160] In a preferred embodiment of the present invention, the Ethernet switch adopts a redundant link topology to improve the reliability and fault tolerance of the system. Specifically, the present invention adopts a dual-ring network topology, which can be expressed as:

[0161] G=(V,E 1 ∪E 2 ),

[0162] Among them, G is the network topology diagram, which is used to represent the network structure of the system. V is the node set (i.e., each functional module), which is used to represent the devices in the network. 1 and E 2 They represent two independent ring links, which are used to provide redundant paths to ensure that even if one of the links fails, the system can still work normally through the other link, thereby greatly improving the reliability of the system.

[0163] In addition, in order to further improve the efficiency and reliability of data transmission, the present invention also adopts the following technologies at the network level: different types of data streams are isolated through VLAN to reduce network broadcast storms and improve network performance. Quality of Service (QoS) strategy assigns different priorities to different types of data streams to ensure that key data (such as alarm signals) can be transmitted in time. Rapid Spanning Tree Protocol (RSTP): used to quickly detect and process network loops to ensure network stability.

[0164] Through the above detailed description, the core modules and working principles of the roof separation instrument alarm device and the coal mine roof monitoring system of the present invention have been fully revealed. The system realizes multi-dimensional analysis and accurate early warning of the roof state through innovative mathematical models and algorithms, providing strong technical support for improving the safety level of coal mine production.

[0165] The present invention further provides a coal mine roof monitoring system, which includes the above-mentioned roof separation instrument alarm device, and also includes a data storage module 7, a visualization module 8 and a remote control module 9.

[0166] The data storage module 7 is connected to the data acquisition module 1 of the roof delamination instrument alarm device for receiving and storing the roof displacement data and roof stress data sent by the data acquisition module 1. Preferably, the data storage module 7 adopts distributed storage technology, which can improve the reliability and access efficiency of data.

[0167] The visualization module 8 is connected to the comprehensive warning module 6 of the roof separation instrument alarm device for receiving the comprehensive warning index sent by the comprehensive warning module 6 and generating a visualization interface of the roof state. In one embodiment, the visualization module 8 can use three-dimensional modeling technology to map the comprehensive warning index to the three-dimensional tunnel model in the form of a heat map, so that the spatial distribution of the roof state can be intuitively displayed.

[0168] The remote control module 9 is connected to the comprehensive warning module 6 of the roof separation alarm device for receiving the alarm signal sent by the comprehensive warning module 6 and sending control instructions to the coal mine safety management system. For example, when a high-level alarm signal is received, the remote control module 9 can automatically send a stop-work evacuation instruction to ensure the safety of the miners.

[0169] The visualization module 8 of the present invention includes a three-dimensional modeling unit 81, a state mapping unit 82 and an interface rendering unit 83. The coordinated work of these three units enables complex roof state information to be presented to the user in an intuitive and understandable manner, greatly improving the practicality and operability of the system.

[0170] The main function of the 3D modeling unit 81 is to construct a 3D tunnel model based on the coal mine tunnel structure data. In a preferred embodiment of the present invention, the 3D modeling unit 81 adopts a voxel-based modeling technology. The advantage of this technology is that it can accurately represent the complex structure of the tunnel and has high computational efficiency. Specifically, the 3D modeling process can be expressed as:

[0171] M={v ijk |i=1,2,...,N x ; j = 1, 2, ..., N y ; k = 1, 2, ..., N z},

[0172] Where M is a three-dimensional tunnel model, which is used to represent the geometric structure of the tunnel. ijk Represents a voxel with coordinates (i, j, k), which is used to represent a minimum unit in the lane. N x 、N y and N z Respectively represent the number of voxels in the x, y, and z directions, and are used to define the resolution of the model. In order to improve the accuracy of the model, the present invention recommends setting the size of the voxel to 0.1 meters, which can well balance the model accuracy and computational complexity.

[0173] The state mapping unit 82 is in communication with the three-dimensional modeling unit 81, and its core task is to map the comprehensive warning index to the three-dimensional lane model. In one embodiment of the present invention, the state mapping process can be expressed as:

[0174] C ijk =f(W ijk ),

[0175] C ijk is the color value of voxel (i,j,k), which is used to indicate the safety status of the location. ijk is the comprehensive warning index of the location, which is used to quantify the safety risk of the location. f is a mapping function, which is used to convert the warning index into a color value. Preferably, the present invention uses a heat map to represent the roof status, that is, mapping the warning index to a color spectrum from green (safe) to red (dangerous). This intuitive visualization method enables staff to quickly identify potential dangerous areas.

[0176] The interface rendering unit 83 is in communication with the state mapping unit 82, and its main function is to render the top board state visualization interface. In a preferred embodiment of the present invention, the interface rendering unit 83 uses a real-time rendering technology based on a GPU. This technology can make full use of the parallel computing capabilities of modern graphics hardware to achieve smooth rendering of large-scale three-dimensional scenes. Specifically, the rendering process can be expressed as:

[0177] I=R(M,C,V),

[0178] Where I is the final rendered image, which is used to display the visualization results of the roof status. R is the rendering function, which is used to combine the 3D model, color mapping and viewpoint parameters into the final image. M is the 3D model, which is used to represent the geometric structure of the tunnel. C is the color mapping, which is used to indicate the safety level of the roof status. V is the viewpoint parameter, which is used to control the user's viewing angle, allowing the user to observe the roof status from different angles.

[0179] The remote control module 9 of the present invention comprises an instruction generation unit 91 and a communication interface unit 92. The design of these two units enables the system to respond to abnormal situations in a timely manner and realize intelligent management of coal mine safety.

[0180] The core task of the instruction generation unit 91 is to generate corresponding control instructions based on the received alarm signal. In one embodiment of the present invention, the generation process of the control instruction can be expressed as:

[0181] A=g(S,E),

[0182] Wherein, A is the generated control instruction, which is used to guide the coal mine to take corresponding safety measures. S is the received alarm signal, which is used to trigger the generation of control instructions. E is the current environmental state, which is used to consider the current operating conditions. g is the instruction generation function, which is used to generate appropriate control instructions according to the alarm signal and the environmental state. Preferably, the present invention adopts a rule-based expert system to implement the instruction generation function. For example, when a high-level alarm signal is received, the system may generate an instruction of "stop operation and evacuate personnel"; while for low-level alarms, it may only need to generate an instruction of "strengthen monitoring".

[0183] The communication interface unit 92 is in communication connection with the instruction generation unit 91, and its main function is to send the generated control instructions to the coal mine safety management system. In a preferred embodiment of the present invention, the communication interface unit 92 adopts a low-latency and high-reliability communication solution based on 5G technology. This solution can ensure that in a complex underground environment, the control instructions can be quickly and reliably transmitted to the relevant execution units. Specifically, the communication process can be expressed as:

[0184] T=h(A,N),

[0185] Wherein, T is the transmission result, which is used to confirm whether the control instruction is successfully sent. A is the control instruction to be transmitted, which is used to guide the coal mine to take corresponding safety measures. N is the current network status, which is used to evaluate the quality of the communication environment. h is a transmission function, which is used to send the control instruction to the target device through the network. In order to improve the reliability of the system, the present invention also designs a communication redundancy mechanism, that is, sending instructions through multiple channels at the same time to cope with possible communication interruptions. In order to improve the reliability of the system, the present invention also designs a communication redundancy mechanism, that is, sending instructions through multiple channels at the same time to cope with possible communication interruptions.

[0186] Through the above detailed description, all the core technical features of the roof separation instrument alarm device and the coal mine roof monitoring system of the present invention have been fully revealed. The system realizes all-round monitoring, analysis and early warning of the coal mine roof status through innovative mathematical models, advanced algorithms and reliable network architecture, and provides strong technical support for improving the safety level of coal mine production. In order to verify the superiority of the present invention, the present invention carried out a series of experimental comparisons. The following will introduce in detail an embodiment of the present invention and two comparative examples, and demonstrate the innovation and practicality of the present invention in the field of coal mine roof monitoring through specific test results.

[0187] Embodiment 1: Roof separation instrument alarm device and coal mine roof monitoring system using the present invention

[0188] The system of the present invention is installed in the coal mining face of a large coal mine, covering an area of ​​about 1,000 square meters. The system includes 100 fiber grating sensors with a sampling frequency of 10 Hz. The topological feature extraction module uses the Delaunay triangulation algorithm to construct the topological structure of the roof displacement field, the dynamic feature analysis module uses a 60-second sliding window to construct the time series, the group theory analysis module calculates the invariant subgroups, and the chaos feature extraction module calculates the first three maximum Lyapunov exponents. The warning threshold of the comprehensive warning module is set to 0.7.

[0189] Comparative Example 1: Traditional Stress Monitoring System

[0190] A traditional stress monitoring system was installed on the same coal mine working face, also using 100 stress sensors with a sampling frequency of 1Hz. The system only monitors roof stress without performing complex mathematical analysis. An alarm is triggered when the stress exceeds a preset threshold.

[0191] Comparative Example 2: Simple Displacement Monitoring System

[0192] A simple displacement monitoring system was installed on the same coal mine working face, using 100 displacement sensors with a sampling frequency of 5Hz. The system monitored the displacement of the roof and analyzed it using simple statistical methods. An alarm was triggered when the displacement exceeded a preset threshold.

[0193] Testing standards and methods:

[0194] 1. Early warning accuracy: The proportion of actual roof abnormalities among the early warning signals issued by the system within 30 days is counted.

[0195] 2. Warning lead time: the average time interval from the system issuing a warning signal to the actual occurrence of roof abnormality.

[0196] 3. False alarm rate: The proportion of roof abnormality events that occurred within 30 days and for which the system failed to issue an early warning.

[0197] 4. System response time: the average time from data collection to generating early warning signals.

[0198] 5. False alarm rate: The proportion of warning signals issued by the system within 30 days that did not actually cause roof abnormalities.

[0199] The test results are shown in the following table:

[0200] Detection indicators Example 1 Comparative Example 1 Comparative Example 2 Early warning accuracy 92 75 80 Warning lead time 2.5 hours 0.5 hours 1 hour False negative rate 3 15 10 System response time 5 seconds 2 seconds 3 seconds False Positive Rate 5 20 15

[0201] It can be seen from the above test results that the system of the present invention (Example 1) is significantly superior to the traditional method (Comparative Example 1 and Comparative Example 2) in all indicators. The specific analysis is as follows:

[0202] 1. Early warning accuracy: The system of the present invention has achieved a high accuracy of 92%, which is much higher than the traditional method. This is mainly due to the multi-dimensional data analysis method adopted by the present invention, especially the introduction of topological feature extraction and group theory analysis, which enables the system to more comprehensively grasp the changes in the roof state.

[0203] 2. Early warning time: The system of the present invention can issue an early warning 2.5 hours in advance on average, which provides sufficient time for coal mine workers to take preventive measures. In contrast, the early warning time of traditional methods is obviously insufficient. This significant improvement is attributed to the chaos feature extraction module of the present invention, which can capture small changes in the system's dynamic behavior, thereby achieving early warning.

[0204] 3. False alarm rate: The system of the present invention reduces the false alarm rate to 3%, greatly improving the safety level. This is due to the comprehensive early warning module of the present invention, which comprehensively considers multiple characteristic indicators and can more comprehensively evaluate the roof status.

[0205] 4. System response time: Although the system of the present invention performs more complex mathematical calculations, the response time is still controlled within 5 seconds through optimized parallel computing and network architecture, meeting the needs of real-time monitoring.

[0206] 5. False alarm rate: The system of the present invention reduces the false alarm rate to 5%, greatly reducing unnecessary downtime and economic losses. This is mainly due to the dynamic feature analysis and group theory analysis of the present invention, which can effectively distinguish between normal roof deformation and potential dangerous situations.

[0207] In summary, the roof separation instrument alarm device and coal mine roof monitoring system of the present invention show significant advantages in terms of early warning accuracy, timeliness and reliability. These improvements can not only better protect the lives of miners, but also create economic value for coal mining enterprises by reducing false alarms and improving production efficiency. In particular, the significant improvement in the early warning lead time provides a critical time window for preventing roof accidents, which has important practical significance in actual coal mine safety production.

[0208] The best embodiment of the present invention is the solution described in Example 1. In practical applications, the system parameters can be fine-tuned according to specific coal mine geological conditions and production requirements, such as adjusting the arrangement density of sensors, optimizing the warning threshold, etc., to achieve the best monitoring effect.

[0209] Figure 3 The dynamic change process of the topological characteristics of the roof displacement field over time is shown. The data is spectrally decomposed into the Laplace matrix and its eigenvalue λ is extracted to reflect the temporal change of the overall topological structure of the roof deformation. When the change of the topological eigenvalue is stable, the system is in a relatively stable state; when the curve shows a sudden change or the fluctuation increases, it may indicate that the roof has entered a potential instability stage. The increase in the fluctuation amplitude in the figure may reflect the increase in the external stress disturbance of the roof, which provides precursor information for the early warning of the roof system, can accurately capture the deformation trend of the roof, and avoid the limitations of relying on single-point data in traditional methods.

[0210] Figure 4 The number of invariant subgroups of the dynamic characteristic matrix in the group theory analysis changes over time in the form of a bar graph, reflecting the change in the symmetry of the system. The number of invariant subgroups indicates the stability of the dynamic behavior of the system. A decrease or disappearance of the number indicates that the symmetry of the system is lost and the risk of potential instability increases significantly. Group invariance analysis reveals the regularity of roof deformation from a mathematical perspective and provides a deeper explanation of physical significance. Through the analysis of symmetry changes, the understanding of the causes of roof instability is enhanced, providing a scientific basis for the formulation of targeted protective measures.

[0211] Figure 5The dynamic evolution of the Lyapunov exponent spectrum is shown. The Lyapunov exponent quantitatively reflects the sensitivity of the system to the initial conditions, and the positive part indicates that the system may enter a chaotic state. The positive value of the Lyapunov exponent indicates that the system is in a chaotic state and the possibility of instability increases. The curve in the figure shows the process of the system transitioning from stability to chaos. When the exponential value grows rapidly, the key point at which the system is about to become unstable can be identified, providing a basis for early warning. The introduction of chaos theory effectively solves the problem that traditional linear methods cannot capture complex nonlinear behaviors and improves the ability to dynamically predict the system.

[0212] Figure 6 The spatial distribution of the comprehensive warning index W in the roof area is displayed in the form of a two-dimensional heat map. The darker the color (red), the higher the risk. The heat map intuitively shows the risk distribution in different areas, which helps to accurately locate high-risk areas and take targeted measures. The map combines multi-dimensional information such as topological eigenvalues, invariant subgroups, and Lyapunov exponents to generate a global risk assessment index that comprehensively reflects the safety status of the roof. It provides intuitive visualization tools to help safety managers quickly understand the risk distribution and make scientific decisions.

[0213] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modification, replacement, and improvement made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A roof separation instrument alarm device, characterized in that: include: Data acquisition module for: Collect roof displacement data; Obtain top plate stress data; A topological feature extraction module is connected to the data acquisition module for: Receiving top plate displacement data sent by the data acquisition module; Based on the roof displacement data, construct a topological structure of the roof displacement field; Performing spectral decomposition on the top plate displacement field topology to obtain Laplace matrix eigenvalues; The dynamic feature analysis module is in communication with the topological feature extraction module and is used to: Receiving the Laplace matrix eigenvalues ​​sent by the topological feature extraction module; Based on the Laplace matrix eigenvalues, construct a dynamic feature matrix; The group theory analysis module is connected to the dynamic feature analysis module for: Receiving the dynamic feature matrix sent by the dynamic feature analysis module; The dynamic characteristic matrix is ​​regarded as an element of a permutation group, and its group properties are analyzed; Compute invariant subgroups; The chaos feature extraction module is connected to the group theory analysis module for: Receiving the invariant subgroup information sent by the group theory analysis module; Based on the invariant subgroup information, construct a Lyapunov exponent spectrum; The comprehensive early warning module is connected to the chaos feature extraction module and the data acquisition module for: Receiving the Lyapunov exponent spectrum sent by the chaos feature extraction module; Receiving the top plate stress data sent by the data acquisition module; Calculating a comprehensive early warning index based on the Lyapunov exponent spectrum, the invariant subgroup information and the roof stress data; When the comprehensive warning index exceeds a preset threshold, an alarm signal is triggered.

2. The roof separation instrument alarm device according to claim 1, characterized in that: The topological feature extraction module comprises: Topology building blocks for: Based on the roof displacement data, a roof displacement field topology structure including a set of measuring points, connection relationships between measuring points and edge weight functions is constructed; A Laplace matrix generating unit is communicatively connected with the topology building unit and is used for: Based on the topology of the top plate displacement field, generating a corresponding Laplace matrix; An eigenvalue calculation unit is communicatively connected to the Laplace matrix generation unit and is used for: Perform eigenvalue decomposition on the Laplace matrix to obtain eigenvalues ​​of the Laplace matrix.

3. The roof separation instrument alarm device according to claim 1, characterized in that: The dynamic feature analysis module comprises: Time series building blocks for: Construct time series based on the eigenvalues ​​of the Laplace matrix at different times; A matrix generating unit is communicatively connected with the time series building unit and is used for: The time series are organized into a dynamic feature matrix.

4. The roof separation instrument alarm device according to claim 1, characterized in that: The group theory analysis module includes: Permutation group building blocks for: Treating the dynamic characteristic matrix as an element of a permutation group, defining group operations; an invariant subgroup calculation unit, communicatively connected to the permutation group construction unit, and configured to: Based on the permutation group, an invariant subgroup is computed.

5. The roof separation instrument alarm device according to claim 1, characterized in that: The chaos feature extraction module comprises: Perturbation vector generation unit, used to: Based on the invariant subgroup information, generating a disturbance vector; A Lyapunov exponent calculation unit is communicatively connected with the disturbance vector generation unit and is used for: Computes the Lyapunov exponent spectrum.

6. The roof separation instrument alarm device according to claim 1, characterized in that: The comprehensive early warning module includes: Characteristic evaluation unit for: The Laplace matrix eigenvalues, invariant subgroup information and Lyapunov index spectrum are evaluated respectively; A weight distribution unit is used to: Assigning weight coefficients to each evaluation result; An index calculation unit, which is in communication with the feature evaluation unit and the weight allocation unit, is used to: Based on each assessment result and the corresponding weight coefficient, a comprehensive early warning index is calculated.

7. The roof separation instrument alarm device according to claim 1, characterized in that: The data acquisition module, the topological feature extraction module, the dynamic feature analysis module, the group theory analysis module, the chaos feature extraction module and the comprehensive early warning module exchange data through an Ethernet switch, and the Ethernet switch adopts a redundant link topology structure to improve the reliability and fault tolerance of the system.

8. A coal mine roof monitoring system, characterized in that: The roof separation instrument alarm device according to any one of claims 1 to 7 further comprises: The data storage module is connected to the data acquisition module of the roof separation instrument alarm device for: Receiving and storing the top plate displacement data and top plate stress data sent by the data acquisition module; The visualization module is communicatively connected with the comprehensive early warning module of the roof separation instrument alarm device, and is used for: Receiving the comprehensive warning index sent by the comprehensive warning module; Generate a visual interface for roof status; The remote control module is connected to the integrated early warning module of the roof separation instrument alarm device for: Receiving the alarm signal sent by the comprehensive early warning module; Send control instructions to the coal mine safety management system.

9. The coal mine roof monitoring system according to claim 7, characterized in that: The visualization module comprises: 3D modeling unit for: Construct a three-dimensional tunnel model based on the coal mine tunnel structure data; A state mapping unit is connected to the three-dimensional modeling unit for: Mapping the comprehensive early warning index onto the three-dimensional lane model; An interface rendering unit, which is in communication with the state mapping unit, is used to: Rendering top plate status visualization interface.

10. The coal mine roof monitoring system according to claim 7, characterized in that: The remote control module comprises: Instruction generation unit, used to: Based on the alarm signal, generating corresponding control instructions; A communication interface unit, communicatively connected to the instruction generation unit, is used to: The control instruction is sent to the coal mine safety management system.

Images