Community security management monitoring system and monitoring method based on Internet of Things

Abstract

This application specifically discloses a community security management and monitoring system based on the Internet of Things and a monitoring method thereof. The community security management and monitoring system based on the Internet of Things obtains a topological feature map based on the topological structure of manhole covers in the community and uses this to obtain temperature feature vectors and pressure feature vectors, thereby obtaining a high-dimensional feature expression of key parameters containing topological structure information. The system further fuses temperature information, pressure information, and gas concentration information based on the thermodynamic gain relationship of gas molecules, thereby achieving effective joint modeling of various monitoring values ​​of various monitoring sensors in a high-dimensional feature space, thereby improving the accuracy of subsequent classification and achieving effective monitoring.

Description

Technical Field

[0001] The present invention relates to the field of intelligent security management, and more specifically, to a cell security management and monitoring system based on the Internet of Things and a monitoring method thereof. Background Art

[0002] With the continuous and rapid development of my country's urban infrastructure construction, various municipal public facilities such as water supply, drainage, gas, heat, electricity, and communications are increasing in cities. The number of manhole covers for various underground pipeline facilities on urban roads is also increasing accordingly.

[0003] In recent years, poor management of manhole covers in cities has led to a nationwide surge in injuries and vehicle damage. When manhole cover incidents occur in residential communities, the impact and harm are even greater. Therefore, improving urban manhole cover management, particularly in residential areas, has become a difficult and hot-button issue plaguing municipal infrastructure management departments across the country.

[0004] At present, the existing management model monitors the manhole covers in the community by setting various monitoring sensors in the manhole covers. However, this management method can only establish a monitoring model based on the monitoring sensor of a single manhole cover for monitoring, which makes the monitoring results easily affected by random fluctuations in values.

[0005] Therefore, in order to obtain more accurate monitoring results and make the manhole cover management of the community better, a monitoring method of the community security management monitoring system based on the Internet of Things is expected. Summary of the Invention

[0006] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a community security management and monitoring system based on the Internet of Things and a monitoring method thereof, which obtains a topological feature map based on the topological structure of the manhole covers in the community and thereby obtains a temperature feature vector and a pressure feature vector, thereby obtaining a high-dimensional feature expression of key parameters containing topological structure information, and further integrates temperature information, pressure information and gas concentration information based on the thermodynamic gain relationship of gas molecules, which can realize the effective joint modeling of various monitoring values ​​of various monitoring sensors in the high-dimensional feature space, thereby improving the accuracy of subsequent classification to achieve effective monitoring.

[0007] According to one aspect of the present application, a cell security management and monitoring system based on the Internet of Things is provided, which includes:

[0008] a topology data unit, configured to convert the topology data of the manhole covers in the cell to be monitored into a topology matrix, wherein each node in the topology data represents a manhole cover, the distance between nodes represents the distance between two manhole covers, the eigenvalue of each off-diagonal position in the topology matrix is ​​the distance between the corresponding two manhole covers, and the eigenvalue of each diagonal position in the topology matrix is ​​0;

[0009] An implicit spatial encoding unit, configured to obtain a topological feature map from the topological matrix using a convolutional neural network as a feature extractor, wherein the topological feature map is used to represent high-dimensional implicit topological association features;

[0010] a sensor data acquisition unit, configured to obtain the temperature value, pressure value and gas concentration value in each of the manhole covers through a temperature sensor, a pressure sensor and a gas concentration sensor deployed in each of the manhole covers;

[0011] a vector construction unit, configured to construct the temperature value and the pressure value in each manhole cover into a temperature vector and a pressure vector according to the arrangement of the nodes in the topological matrix, and to use an encoder to encode the temperature vector and the pressure vector into a high-dimensional feature space to obtain a temperature encoding vector and a pressure encoding vector;

[0012] a mapping unit, configured to map the temperature encoding vector and the pressure encoding vector into a high-dimensional feature space of the topological feature graph to obtain a temperature feature vector and a pressure feature vector;

[0013] a pressure and temperature fusion unit, configured to fuse the temperature eigenvector and the pressure eigenvector based on a thermodynamic gain relationship of gas molecules to obtain a fused eigenvector, wherein the thermodynamic gain relationship of gas molecules is determined based on eigenvalues ​​of respective positions in the pressure eigenvector and the temperature eigenvector, standard temperature, standard atmospheric pressure, and gas concentration value; and

[0014] The monitoring result generating unit is used to pass the fused feature vector through a classifier to obtain a classification result, and the classification result is used to indicate whether the monitoring result is normal.

[0015] In the above-mentioned Internet of Things-based community security management and monitoring system, the topology data unit is further used to: number the manhole covers in the community to be monitored; arrange the numbered manhole covers according to the rows and columns of the matrix; and fill each non-diagonal position of the matrix with the distance between the corresponding two manhole covers and fill each diagonal position of the matrix with a 0 value to obtain the topology matrix.

[0016] In the above-mentioned cell security management and monitoring system based on the Internet of Things, the convolutional neural network processes the topology matrix using the following formula to obtain the topology feature graph; wherein the formula is:

[0017] f i =active(N i ×f i-1 +B i )

[0018] Among them, f i-1 is the input of the i-th layer convolutional neural network, f i is the output of the i-th layer of convolutional neural network, N i is the filter of the i-th layer convolutional neural network, and B i is the bias vector of the i-th layer convolutional neural network, and active represents the nonlinear activation function.

[0019] In the above-mentioned Internet of Things-based community security management and monitoring system, the vector construction unit is further used to: use the fully connected layer of the encoder to perform full-position encoding on the pressure vector to extract high-dimensional implicit features of the eigenvalues ​​at each position in the pressure vector; use the one-dimensional convolutional layer of the encoder to perform one-dimensional convolutional encoding on the pressure vector to extract high-dimensional correlation features of the association between the eigenvalues ​​at adjacent positions in the pressure vector to obtain the pressure encoding vector; use the fully connected layer of the encoder to perform full-position encoding on the temperature vector to extract high-dimensional implicit features of the eigenvalues ​​at each position in the temperature vector; and, use the one-dimensional convolutional layer of the encoder to perform one-dimensional convolutional encoding on the temperature vector to extract high-dimensional correlation features of the association between the eigenvalues ​​at adjacent positions in the temperature vector to obtain the temperature encoding vector.

[0020] In the above-mentioned Internet of Things-based cell security management and monitoring system, the mapping unit is further used to: perform matrix multiplication on the topological feature map and the temperature encoding vector to map the temperature encoding vector to the high-dimensional feature space of the topological feature map to obtain the temperature feature vector; and perform matrix multiplication on the topological feature map and the pressure encoding vector to map the pressure encoding vector to the high-dimensional feature space of the topological feature map to obtain the pressure feature vector.

[0021] In the above-mentioned IoT-based cell security management and monitoring system, the pressure and temperature fusion unit is further configured to fuse the temperature eigenvector and the pressure eigenvector to obtain a fused eigenvector based on the thermodynamic gain relationship of gas molecules using the following formula:

[0022]

[0023]

[0024] Where i and g refer to the isotopes of the gas, T i and p i is the eigenvalue of the corresponding position of the temperature eigenvector and the pressure eigenvector, T0 and p0 are the standard temperature and standard atmospheric pressure, Q i,g and σ i,g It represents the total number of molecules per unit volume and the heat absorption cross section of a given gas mixture, which is obtained based on the gas concentration values ​​of various gases obtained by the gas concentration sensor.

[0025] In the above-mentioned Internet of Things-based community security management and monitoring system, the monitoring result generation unit is further used to calculate the Softmax classification function value of each position in the fused feature vector as the probability value of whether the manhole cover corresponding to each position is abnormal, and the Softmax classification function value is the weighted sum of the natural exponential function value of the eigenvalue of each position raised to the power divided by the natural exponential function value of the eigenvalue of each position raised to the power; and based on the comparison between the Softmax classification function value and the preset threshold, it is determined whether the manhole cover corresponding to each position is normal.

[0026] According to another aspect of the present application, a monitoring method of a cell security management and monitoring system based on the Internet of Things includes:

[0027] Convert the topological structure data of the manhole covers in the cell to be monitored into a topological matrix, wherein each node in the topological structure data represents a manhole cover, the distance between nodes represents the distance between two manhole covers, the eigenvalue of each non-diagonal position in the topological matrix is ​​the distance between the corresponding two manhole covers, and the eigenvalue of each diagonal position in the topological matrix is ​​0;

[0028] A convolutional neural network is used as a feature extractor to obtain a topological feature map from the topological matrix, wherein the topological feature map is used to represent high-dimensional implicit topological association features;

[0029] Obtaining the temperature value, pressure value and gas concentration value in each of the manhole covers by means of a temperature sensor, a pressure sensor and a gas concentration sensor deployed in each of the manhole covers;

[0030] Constructing the temperature value and the pressure value in each manhole cover into a temperature vector and a pressure vector according to the arrangement of the nodes in the topological matrix, and encoding the temperature vector and the pressure vector into a high-dimensional feature space using an encoder to obtain a temperature encoding vector and a pressure encoding vector;

[0031] Mapping the temperature encoding vector and the pressure encoding vector into the high-dimensional feature space of the topological feature graph to obtain a temperature feature vector and a pressure feature vector;

[0032] fusing the temperature eigenvector and the pressure eigenvector based on a thermodynamic gain relationship of gas molecules to obtain a fused eigenvector, wherein the thermodynamic gain relationship of gas molecules is determined based on eigenvalues ​​of respective positions in the pressure eigenvector and the temperature eigenvector, standard temperature, standard atmospheric pressure, and gas concentration value; and

[0033] The fused feature vector is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether the monitoring result is normal.

[0034] In the above-mentioned monitoring method of the community security management and monitoring system based on the Internet of Things, the topological structure data of the manhole covers in the community to be monitored is converted into a topological matrix, including: numbering the manhole covers in the community to be monitored; arranging the numbered manhole covers according to the rows and columns of the matrix; and filling each non-diagonal position of the matrix with the distance between the corresponding two manhole covers and filling each diagonal position of the matrix with a 0 value to obtain the topological matrix.

[0035] In the monitoring method of the cell security management and monitoring system based on the Internet of Things, the convolutional neural network processes the topology matrix using the following formula to obtain the topology feature graph;

[0036] Wherein, the formula is:

[0037] f i =active(N i ×f i-1 +B i )

[0038] Among them, f i-1 is the input of the i-th layer convolutional neural network, f i is the output of the i-th layer of convolutional neural network, N i is the filter of the i-th layer convolutional neural network, and B i is the bias vector of the i-th layer convolutional neural network, and active represents the nonlinear activation function.

[0039] In the above-mentioned monitoring method of the community security management and monitoring system based on the Internet of Things, the temperature value and pressure value in each manhole cover are constructed into a temperature vector and a pressure vector according to the arrangement of the nodes in the topological matrix, and the temperature vector and the pressure vector are encoded into a high-dimensional feature space using an encoder to obtain a temperature encoding vector and a pressure encoding vector, including: using the fully connected layer of the encoder to perform full-position encoding on the pressure vector to extract the high-dimensional implicit features of the eigenvalues ​​of each position in the pressure vector; using the one-dimensional convolutional layer of the encoder to perform one-dimensional convolutional encoding on the pressure vector to extract the high-dimensional correlation features of the association between the eigenvalues ​​of adjacent positions in the pressure vector to obtain the pressure encoding vector; using the fully connected layer of the encoder to perform full-position encoding on the temperature vector to extract the high-dimensional implicit features of the eigenvalues ​​of each position in the temperature vector; and, using the one-dimensional convolutional layer of the encoder to perform one-dimensional convolutional encoding on the temperature vector to extract the high-dimensional correlation features of the association between the eigenvalues ​​of adjacent positions in the temperature vector to obtain the temperature encoding vector.

[0040] In the above-mentioned monitoring method of the cell security management and monitoring system based on the Internet of Things, the temperature encoding vector and the pressure encoding vector are mapped to the high-dimensional feature space of the topological feature map to obtain the temperature feature vector and the pressure feature vector, including: matrix multiplying the topological feature map and the temperature encoding vector to map the temperature encoding vector to the high-dimensional feature space of the topological feature map to obtain the temperature feature vector; and matrix multiplying the topological feature map and the pressure encoding vector to map the pressure encoding vector to the high-dimensional feature space of the topological feature map to obtain the pressure feature vector.

[0041] In the monitoring method of the cell security management and monitoring system based on the Internet of Things, the fusing of the temperature eigenvector and the pressure eigenvector based on the thermodynamic gain relationship of gas molecules to obtain a fused eigenvector includes: fusing the temperature eigenvector and the pressure eigenvector based on the thermodynamic gain relationship of gas molecules to obtain a fused eigenvector using the following formula; the formula is:

[0042]

[0043]

[0044] Where i and g refer to the isotopes of the gas, T i and p i is the eigenvalue of the corresponding position of the temperature eigenvector and the pressure eigenvector, T0 and p0 are the standard temperature and standard atmospheric pressure, Q i,g and σ i,gIt represents the total number of molecules per unit volume and the heat absorption cross section of a given gas mixture, which is obtained based on the gas concentration values ​​of various gases obtained by the gas concentration sensor.

[0045] In the above-mentioned monitoring method of the community security management and monitoring system based on the Internet of Things, the fused feature vector is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether the monitoring result is normal, including: calculating the Softmax classification function value of each position in the fused feature vector as the probability value of whether the manhole cover corresponding to each position is abnormal, the Softmax classification function value is the weighted sum of the natural exponential function value of the eigenvalue of each position raised to the power divided by the natural exponential function value of the eigenvalue of each position raised to the power; and, based on the comparison between the Softmax classification function value and a preset threshold, determining whether the manhole cover corresponding to each position is normal.

[0046] Compared with the existing technology, the Internet of Things-based community security management and monitoring system and its monitoring method provided by the present application obtain a topological feature map based on the topological structure of the manhole covers in the community and obtain temperature feature vectors and pressure feature vectors based on this, thereby obtaining a high-dimensional feature expression of key parameters containing topological structure information, and further integrating temperature information, pressure information and gas concentration information based on the thermodynamic gain relationship of gas molecules, which can realize the effective joint modeling of various monitoring values ​​of various monitoring sensors in high-dimensional feature space, thereby improving the accuracy of subsequent classification to achieve effective monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0048] Figure 1 This is an application scenario diagram of a cell security management and monitoring system based on the Internet of Things according to an embodiment of the present application.

[0049] Figure 2 4 is a block diagram of a cell security management and monitoring system based on the Internet of Things according to an embodiment of the present application.

[0050] Figure 3 This is a block diagram of a monitoring result generation unit in an IoT-based cell security management and monitoring system according to an embodiment of the present application.

[0051] Figure 4Flowchart of a monitoring method of a cell security management and monitoring system based on the Internet of Things according to an embodiment of the present application.

[0052] Figure 5 Schematic diagram of the architecture of a monitoring method of a cell security management and monitoring system based on the Internet of Things according to an embodiment of the present application. DETAILED DESCRIPTION

[0053] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described herein.

[0054] Scenario Overview

[0055] As previously mentioned, in recent years, poor management of manhole covers in cities has led to a nationwide increase in injuries and vehicle damage. If manhole cover incidents occur in residential communities, the impact and harm they cause are even greater. Therefore, improving urban manhole cover management, particularly in residential areas, has become a difficult and hot-button issue plaguing municipal infrastructure management departments across the country.

[0056] Currently, the existing management model monitors a residential community's manhole covers by installing various monitoring sensors in the covers. However, this management method can only establish a monitoring model based on the monitoring sensor of a single manhole cover, making the monitoring results easily affected by random fluctuations in the value. Therefore, in order to obtain more accurate monitoring results and achieve better management of residential manhole covers, a monitoring method for a residential security management and monitoring system based on the Internet of Things is desired.

[0057] Therefore, in this application, it is expected to jointly monitor the monitoring values ​​of the monitoring sensors of all manhole covers in the community. Therefore, how to jointly model the various monitoring values ​​of various monitoring sensors of all manhole covers in high-dimensional feature space becomes a problem that needs to be solved.

[0058] Based on this, the technical solution of this application first requires a mathematical representation of the topological structure of manhole covers in a residential area. This can be done by applying an adjacency matrix, where the rows and columns of the matrix represent each node (i.e., a manhole cover), and the corresponding positions in the matrix represent the distances between nodes, thereby obtaining a topological matrix. This topological matrix is ​​then input into a convolutional neural network to obtain a topological feature graph, thereby mining high-dimensional implicit topological association features.

[0059] Since current monitoring sensors primarily consist of temperature sensors, pressure sensors, and gas concentration sensors, and their values ​​are crucial for determining the physical condition of manhole covers, we transform the values ​​of the temperature and pressure sensors into temperature and pressure vectors, respectively, according to the order of the nodes in the topological matrix. These vectors are then encoded into a high-dimensional space using an encoder to generate temperature and pressure encoding vectors. These vectors are then multiplied by the topological feature graph to generate temperature and pressure feature vectors.

[0060] Furthermore, considering that the temperature eigenvector and the pressure eigenvector are only the mapping vectors of the temperature and pressure values ​​in the high-dimensional feature space of the topological information, the two need to be further fused. Therefore, based on the thermodynamic gain relationship of gas molecules, the fused eigenvector is obtained from the temperature eigenvector and the pressure eigenvector, which is expressed as:

[0061]

[0062]

[0063] Where i and g refer to the isotopes of the gas, T i and p i is the eigenvalue of the corresponding position of the temperature eigenvector and the pressure eigenvector, T0 and p0 are the standard temperature and standard atmospheric pressure, Q i,g and σ i,g It represents the total number of molecules per unit volume and the heat absorption cross section of a given gas mixture, which can be obtained based on the gas concentration values ​​of various gases obtained by the gas concentration sensor.

[0064] In this way, after obtaining the fused feature vector, it can be input into the classifier to obtain a classification result on whether the monitoring result is normal. Furthermore, since the feature value of each position in the fused feature vector corresponds to a manhole cover, it is also possible to determine whether a manhole cover is abnormal based on the Softmax classification function value of each position relative to the whole.

[0065] Based on this, the present application proposes a community security management and monitoring system based on the Internet of Things, which includes: a topological data unit, which is used to convert the topological structure data of the manhole covers in the community to be monitored into a topological matrix, wherein each node in the topological structure data represents a manhole cover, and the distance between the nodes represents the distance between the two manhole covers. The eigenvalues ​​of each non-diagonal position in the topological matrix are the distances between the corresponding two manhole covers, and the eigenvalues ​​of each diagonal position in the topological matrix are 0; an implicit spatial encoding unit, which is used to obtain a topological feature map from the topological matrix using a convolutional neural network as a feature extractor, and the topological feature map is used to represent high-dimensional implicit topological association features; a sensor data acquisition unit, which is used to obtain the temperature value, pressure value and gas concentration value in each manhole cover through the temperature sensors, pressure sensors and gas concentration sensors deployed in each manhole cover; a vector construction unit, which is used to convert each manhole cover into a vector. The temperature value and pressure value inside the manhole cover are constructed as a temperature vector and a pressure vector according to the arrangement of the nodes in the topological matrix, and an encoder is used to encode the temperature vector and the pressure vector into a high-dimensional feature space to obtain a temperature encoding vector and a pressure encoding vector; a mapping unit is used to map the temperature encoding vector and the pressure encoding vector into the high-dimensional feature space of the topological feature graph to obtain a temperature feature vector and a pressure feature vector; a pressure and temperature fusion unit is used to fuse the temperature feature vector and the pressure feature vector based on the thermodynamic gain relationship of gas molecules to obtain a fused feature vector, wherein the thermodynamic gain relationship of the gas molecules is determined based on the feature values ​​of each position in the pressure feature vector and the temperature feature vector, the standard temperature and the standard atmospheric pressure and the gas concentration value; and a monitoring result generation unit is used to pass the fused feature vector through a classifier to obtain a classification result, and the classification result is used to indicate whether the monitoring result is normal.

[0066] Figure 1 The figure shows an application scenario diagram of a cell security management and monitoring system based on the Internet of Things according to an embodiment of the present application. Figure 1 As shown, in this application scenario, first, the cell to be monitored is obtained (for example, Figure 1 R) in the manhole cover (for example, as shown in Figure 1 The topological structure of B) is shown in FIG, wherein each node in the topological structure data represents a manhole cover, the distance between nodes represents the distance between two manhole covers, and the temperature sensor (for example, Figure 1 T shown in FIG), a pressure sensor (e.g., Figure 1 P) and a gas concentration sensor (e.g., as shown in Figure 1 G) shown in the figure obtains the temperature value, pressure value and gas concentration value inside each of the manhole covers.

[0067] Then, the obtained topological structure of the manhole covers in the cell to be monitored and the temperature value, pressure value and gas concentration value in each manhole cover are input to a server (for example, Figure 1 In the cloud server S) shown in the figure, the server is capable of processing the topological structure of the manhole covers in the community to be monitored and the temperature values, pressure values ​​and gas concentration values ​​in each of the manhole covers using a community security management and monitoring algorithm based on the Internet of Things to generate a classification result indicating whether the monitoring result is normal.

[0068] After introducing the basic principles of the present application, various non-limiting embodiments of the present application will be described in detail with reference to the accompanying drawings.

[0069] Exemplary Systems

[0070] Figure 2 FIG is a block diagram of a cell security management and monitoring system based on the Internet of Things according to an embodiment of the present application. Figure 2As shown, according to the embodiment of the present application, the cell security management and monitoring system 200 based on the Internet of Things includes: a topological data unit 210, which is used to convert the topological structure data of the manhole covers in the cell to be monitored into a topological matrix, wherein each node in the topological structure data represents a manhole cover, the distance between the nodes represents the distance between the two manhole covers, the eigenvalue of each non-diagonal position in the topological matrix is ​​the distance between the corresponding two manhole covers, and the eigenvalue of each diagonal position in the topological matrix is ​​0; an implicit spatial encoding unit 220, which is used to obtain a topological feature map from the topological matrix using a convolutional neural network as a feature extractor, and the topological feature map is used to represent high-dimensional implicit topological association features; a sensor data acquisition unit 230, which is used to obtain the temperature value, pressure value and gas concentration value in each manhole cover by means of temperature sensors, pressure sensors and gas concentration sensors deployed in each manhole cover; a vector construction unit 240, which is used to convert each The temperature value and pressure value in the manhole cover are constructed as a temperature vector and a pressure vector according to the arrangement of the nodes in the topological matrix, and an encoder is used to encode the temperature vector and the pressure vector into a high-dimensional feature space to obtain a temperature encoding vector and a pressure encoding vector; a mapping unit 250 is used to map the temperature encoding vector and the pressure encoding vector into the high-dimensional feature space of the topological feature graph to obtain a temperature feature vector and a pressure feature vector; a pressure and temperature fusion unit 260 is used to fuse the temperature feature vector and the pressure feature vector based on the thermodynamic gain relationship of gas molecules to obtain a fused feature vector, wherein the thermodynamic gain relationship of the gas molecules is determined based on the feature values ​​of each position in the pressure feature vector and the temperature feature vector, the standard temperature and the standard atmospheric pressure and the gas concentration value; and a monitoring result generation unit 270 is used to pass the fused feature vector through a classifier to obtain a classification result, and the classification result is used to indicate whether the monitoring result is normal.

[0071] Specifically, in an embodiment of the present application, the topological data unit 210 and the implicit spatial coding unit 220 are used to convert the topological structure data of the manhole covers in the cell to be monitored into a topological matrix, and use a convolutional neural network as a feature extractor to obtain a topological feature map from the topological matrix, and the topological feature map is used to represent high-dimensional implicit topological association features. As mentioned above, although various types of monitoring sensors are currently installed in manhole covers, if a monitoring model is established based only on a single monitoring sensor of the manhole cover for monitoring, it is easily affected by random fluctuations in values. Therefore, in the technical solution of the present application, it is expected to jointly monitor the monitoring values ​​of the monitoring sensors of all manhole covers in the cell.

[0072] That is, first, it is necessary to obtain the topological structure of the manhole covers in the cell to be monitored, wherein each node in the topological structure data represents a manhole cover, and the distance between nodes represents the distance between two manhole covers. Then, the topological structure is converted into a topological matrix, wherein the eigenvalue of each non-diagonal position in the topological matrix is ​​the distance between the corresponding two manhole covers, and the eigenvalue of each diagonal position in the topological matrix is ​​0, that is, the topological structure of the manhole covers in the cell is mathematically expressed to facilitate subsequent processing. In a specific example, an adjacency matrix can be applied, that is, the rows and columns of the matrix represent the nodes, that is, manhole covers, and the corresponding positions in the matrix represent the distance between the nodes, so that a topological matrix can be obtained. Finally, the topological matrix is ​​input into a convolutional neural network as a feature extractor for processing to mine the high-dimensional implicit topological association features between the manhole covers, thereby obtaining a topological feature map.

[0073] Accordingly, in a specific example, the topological data unit includes: first, numbering the manhole covers in the cell to be monitored; then, arranging the numbered manhole covers according to the rows and columns of the matrix; finally, filling each non-diagonal position of the matrix with the distance between the corresponding two manhole covers and filling each diagonal position of the matrix with a 0 value to obtain the topological matrix.

[0074] More specifically, in an embodiment of the present application, the convolutional neural network processes the topological matrix using the following formula to obtain the topological feature map;

[0075] Wherein, the formula is:

[0076] f i =active(N i ×f i-1 +B i )

[0077] Among them, f i-1 is the input of the i-th layer convolutional neural network, f i is the output of the i-th layer of convolutional neural network, N i is the filter of the i-th layer convolutional neural network, and B i is the bias vector of the i-th layer convolutional neural network, and active represents the nonlinear activation function.

[0078] Specifically, in the embodiment of the present application, the sensor data acquisition unit 230 and the vector construction unit 240 are used to obtain the temperature value, pressure value and gas concentration value in each manhole cover through the temperature sensor, pressure sensor and gas concentration sensor deployed in each manhole cover, and construct the temperature value and pressure value in each manhole cover into a temperature vector and a pressure vector according to the arrangement of the nodes in the topological matrix, and use an encoder to encode the temperature vector and the pressure vector into a high-dimensional feature space to obtain a temperature encoding vector and a pressure encoding vector. It should be understood that since the current monitoring sensors are mainly temperature sensors, pressure sensors and gas concentration sensors, and the values ​​of the temperature sensors and the pressure sensors are most important for judging the physical state of the manhole cover. Therefore, in the technical solution of the present application, first, the temperature value and pressure value in each manhole cover are collected by the temperature sensor and the pressure sensor deployed in each manhole cover, and considering the subsequent need to effectively fuse the temperature information and pressure information, the gas concentration value is also collected by the gas concentration sensor deployed in each manhole cover. Then, the temperature and pressure values ​​within each manhole cover are constructed into temperature vectors and pressure vectors according to the arrangement of nodes in the topological matrix, so as to facilitate subsequent processing by the encoder. Finally, the temperature vectors and pressure vectors are encoded by the encoder to encode them into a high-dimensional feature space, thereby obtaining temperature encoding vectors and pressure encoding vectors.

[0079] More specifically, in an embodiment of the present application, the vector construction unit includes: first, using the fully connected layer of the encoder to perform full-position encoding on the pressure vector to extract high-dimensional implicit features of the eigenvalues ​​of each position in the pressure vector; then, using the one-dimensional convolutional layer of the encoder to perform one-dimensional convolutional encoding on the pressure vector to extract high-dimensional correlation features of the association between the eigenvalues ​​of adjacent positions in the pressure vector to obtain the pressure encoding vector; then, using the fully connected layer of the encoder to perform full-position encoding on the temperature vector to extract high-dimensional implicit features of the eigenvalues ​​of each position in the temperature vector; finally, using the one-dimensional convolutional layer of the encoder to perform one-dimensional convolutional encoding on the temperature vector to extract high-dimensional correlation features of the association between the eigenvalues ​​of adjacent positions in the temperature vector to obtain the temperature encoding vector.

[0080] Specifically, in this embodiment of the present application, the mapping unit 250 is configured to map the temperature encoding vector and the pressure encoding vector into the high-dimensional feature space of the topological feature map to obtain a temperature feature vector and a pressure feature vector. That is, the topological feature map is multiplied by the temperature encoding vector and the pressure encoding vector, respectively, to map the temperature encoding vector and the pressure encoding vector into the high-dimensional feature space of the topological feature map, thereby obtaining a temperature feature vector and a pressure feature vector that represent high-dimensional features of key parameters containing topological structure information.

[0081] More specifically, in an embodiment of the present application, the mapping unit includes: first, performing matrix multiplication on the topological feature map and the temperature encoding vector to map the temperature encoding vector to the high-dimensional feature space of the topological feature map to obtain the temperature feature vector; then, performing matrix multiplication on the topological feature map and the pressure encoding vector to map the pressure encoding vector to the high-dimensional feature space of the topological feature map to obtain the pressure feature vector.

[0082] Specifically, in an embodiment of the present application, the pressure and temperature fusion unit 260 is used to fuse the temperature feature vector and the pressure feature vector based on the thermodynamic gain relationship of gas molecules to obtain a fused feature vector. It should be understood that, considering that the temperature feature vector and the pressure feature vector are only mapping vectors of temperature values ​​and pressure values ​​in the high-dimensional feature space of topological information, the two need to be further fused. Therefore, in the technical solution of the present application, a fused feature vector is obtained from the temperature feature vector and the pressure feature vector based on the thermodynamic gain relationship of gas molecules. That is, the temperature information, the pressure information and the gas concentration information are further fused based on the thermodynamic gain relationship of gas molecules to achieve effective joint modeling of various monitoring values ​​of various monitoring sensors in the high-dimensional feature space, thereby improving the accuracy of subsequent classification to achieve effective monitoring. It is worth mentioning that here, the thermodynamic gain relationship of the gas molecules is determined based on the characteristic values ​​of each position in the pressure feature vector and the temperature feature vector, the standard temperature and standard atmospheric pressure and the gas concentration value.

[0083] More specifically, in the embodiment of the present application, the pressure and temperature fusion unit is further configured to fuse the temperature feature vector and the pressure feature vector to obtain a fused feature vector using the following formula based on the thermodynamic gain relationship of gas molecules;

[0084] The formula is:

[0085]

[0086]

[0087] Where i and g refer to the isotopes of the gas, T i and p i is the eigenvalue of the corresponding position of the temperature eigenvector and the pressure eigenvector, T0 and p0 are the standard temperature and standard atmospheric pressure, Q i,g and σ i,g It represents the total number of molecules per unit volume and the heat absorption cross section of a given gas mixture, which is obtained based on the gas concentration values ​​of various gases obtained by the gas concentration sensor.

[0088] Specifically, in an embodiment of the present application, the monitoring result generating unit 270 is used to pass the fused feature vector through a classifier to obtain a classification result, and the classification result is used to indicate whether the monitoring result is normal. That is, after obtaining the fused feature vector, the fused feature vector can be input into the classifier to obtain a classification result indicating whether the monitoring result is normal. In particular, since the eigenvalue of each position of the fused feature vector corresponds to a manhole cover, it is also possible to determine whether a manhole cover has an abnormality based on the Softmax classification function value of each position in the fused feature vector relative to the whole.

[0089] More specifically, in an embodiment of the present application, the monitoring result generating unit includes: first, calculating the Softmax classification function value of each position in the fused feature vector as the probability value of whether the manhole cover corresponding to each position is abnormal, and the Softmax classification function value is the weighted sum of the natural exponential function value of the eigenvalue of each position raised to the power divided by the natural exponential function value of the eigenvalue of each position raised to the power, that is, the formula is: pi = exp(-xi) / ∑ i exp(-xi), where xi represents the eigenvalue of each position in the fused feature vector. Then, based on the comparison between the Softmax classification function value and a preset threshold, it is determined whether the manhole cover corresponding to each position is normal. Specifically, when the Softmax classification function value is greater than the preset threshold, the classification result is that the manhole cover corresponding to the position is abnormal; when the Softmax classification function value is less than the preset threshold, the classification result is that the manhole cover corresponding to the position is normal.

[0090] Figure 3 FIG2 is a block diagram of a monitoring result generating unit in a cell security management and monitoring system based on the Internet of Things according to an embodiment of the present application. Figure 3As shown, the monitoring result generating unit 270 includes: a probability value calculating subunit 271, used to calculate the Softmax classification function value of each position in the fused feature vector as the probability value of whether the manhole cover corresponding to each position is abnormal, and the Softmax classification function value is a weighted sum of the natural exponential function value of the eigenvalue of each position raised to a power divided by the natural exponential function value of the eigenvalue of each position raised to a power; a comparing subunit 272, based on the comparison between the Softmax classification function value obtained by the probability value calculating subunit 271 and a preset threshold, determines whether the manhole cover corresponding to each position is normal.

[0091] In summary, the IoT-based community security management and monitoring system 200 according to the embodiment of the present application is explained, which obtains a topological feature map based on the topological structure of the manhole covers in the community and obtains temperature feature vectors and pressure feature vectors based on this, thereby obtaining a high-dimensional feature expression of key parameters containing topological structure information, and further integrates temperature information, pressure information and gas concentration information based on the thermodynamic gain relationship of gas molecules, which can realize effective joint modeling of various monitoring values ​​of various monitoring sensors in high-dimensional feature space, thereby improving the accuracy of subsequent classification to achieve effective monitoring.

[0092] As described above, the IoT-based cell security management and monitoring system 200 according to the embodiment of the present application can be implemented in various terminal devices, such as a server for the IoT-based cell security management and monitoring algorithm. In one example, the IoT-based cell security management and monitoring system 200 according to the embodiment of the present application can be integrated into the terminal device as a software module and / or hardware module. For example, the IoT-based cell security management and monitoring system 200 can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the IoT-based cell security management and monitoring system 200 can also be one of the many hardware modules of the terminal device.

[0093] Alternatively, in another example, the IoT-based cell security management and monitoring system 200 and the terminal device may also be separate devices, and the IoT-based cell security management and monitoring system 200 may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0094] Exemplary Methods

[0095] Figure 4 The figure shows a flow chart of the monitoring method of the community security management and monitoring system based on the Internet of Things. Figure 4As shown, the monitoring method of the cell security management and monitoring system based on the Internet of Things according to the embodiment of the present application includes the following steps: S110, converting the topological structure data of the manhole covers in the cell to be monitored into a topological matrix, wherein each node in the topological structure data represents a manhole cover, the distance between the nodes represents the distance between the two manhole covers, the eigenvalue of each non-diagonal position in the topological matrix is ​​the distance between the corresponding two manhole covers, and the eigenvalue of each diagonal position in the topological matrix is ​​0; S120, using a convolutional neural network as a feature extractor to obtain a topological feature map from the topological matrix, the topological feature map is used to represent high-dimensional implicit topological association features; S130, obtaining the temperature value, pressure value and gas concentration value in each manhole cover by deploying a temperature sensor, a pressure sensor and a gas concentration sensor in each manhole cover; S140, converting each manhole cover into a topological matrix. The temperature value and pressure value inside the manhole cover are constructed as a temperature vector and a pressure vector according to the arrangement of the nodes in the topological matrix, and an encoder is used to encode the temperature vector and the pressure vector into a high-dimensional feature space to obtain a temperature encoding vector and a pressure encoding vector; S150, the temperature encoding vector and the pressure encoding vector are mapped to the high-dimensional feature space of the topological feature graph to obtain a temperature feature vector and a pressure feature vector; S160, the temperature feature vector and the pressure feature vector are fused based on the thermodynamic gain relationship of gas molecules to obtain a fused feature vector, wherein the thermodynamic gain relationship of the gas molecules is determined based on the feature values ​​of each position in the pressure feature vector and the temperature feature vector, the standard temperature and standard atmospheric pressure and the gas concentration value; and, S170, the fused feature vector is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether the monitoring result is normal.

[0096] Figure 5 The figure shows a schematic diagram of the architecture of the monitoring method of the cell security management and monitoring system based on the Internet of Things according to an embodiment of the present application. Figure 5 As shown, in the network architecture of the monitoring method of the cell security management and monitoring system based on the Internet of Things, first, the topological structure of the manhole cover in the cell to be monitored (for example, Figure 5 The P) data shown in FIG is converted into a topological matrix (e.g., Figure 5 M as shown in FIG); S120, with a convolutional neural network (for example, as Figure 5 ) as a feature extractor to obtain a topological feature map (e.g., as Figure 5 F shown in FIG); S140, the temperature value inside each of the manhole covers (for example, Figure 5 Q1) and pressure values ​​(e.g., Figure 5Q2) is constructed as a temperature vector according to the arrangement of nodes in the topology matrix (for example, Figure 5 V1 as shown in FIG) and the pressure vector (e.g., Figure 5 V2 as shown in ) and use an encoder (e.g. Figure 5 E) encoding the temperature vector and the pressure vector into a high-dimensional feature space to obtain a temperature encoding vector (e.g., as shown in FIG. Figure 5 V3 as shown in FIG) and the pressure encoding vector (e.g., as Figure 5 S150, mapping the temperature encoding vector and the pressure encoding vector to the high-dimensional feature space of the topological feature graph to obtain a temperature feature vector (for example, as shown in FIG. Figure 5 VF1) and the pressure characteristic vector (e.g., as shown in Figure 5 VF2 shown in FIG); S160, based on the thermodynamic gain relationship of the gas molecules, the temperature feature vector and the pressure feature vector are fused to obtain a fused feature vector (for example, Figure 5 VF as shown in FIG), wherein the thermodynamic gain relationship of the gas molecules is based on the characteristic values ​​of each position in the pressure characteristic vector and the temperature characteristic vector, the standard temperature and the standard atmospheric pressure and the obtained gas concentration value (for example, as Figure 5 Q3) shown in FIG; and, S170, the fused feature vector is passed through a classifier (e.g., Figure 5 The circle S shown in FIG) is used to obtain a classification result, which is used to indicate whether the monitoring result is normal.

[0097] More specifically, in steps S110 and S120, the topological structure data of the manhole covers in the cell to be monitored is converted into a topological matrix, wherein each node in the topological structure data represents a manhole cover, the distance between nodes represents the distance between two manhole covers, the eigenvalue of each non-diagonal position in the topological matrix is ​​the distance between the corresponding two manhole covers, and the eigenvalue of each diagonal position in the topological matrix is ​​0, and a convolutional neural network is used as a feature extractor to obtain a topological feature map from the topological matrix, and the topological feature map is used to represent high-dimensional implicit topological association features. That is, first, it is necessary to obtain the topological structure of the manhole covers in the cell to be monitored, wherein each node in the topological structure data represents a manhole cover, and the distance between nodes represents the distance between two manhole covers. Then, the topological structure is converted into a topological matrix, wherein the eigenvalue of each non-diagonal position in the topological matrix is ​​the distance between the corresponding two manhole covers, and the eigenvalue of each diagonal position in the topological matrix is ​​0. That is, the topological structure of the manhole covers in the cell is mathematically expressed to facilitate subsequent processing. In a specific example, an adjacency matrix can be applied, that is, the rows and columns of the matrix represent the various nodes, that is, manhole covers, and the corresponding positions in the matrix represent the distance between the nodes, so that a topological matrix can be obtained. Finally, the topological matrix is ​​input into a convolutional neural network as a feature extractor for processing to mine the high-dimensional implicit topological association features between the various manhole covers, thereby obtaining a topological feature map.

[0098] More specifically, in steps S130 and S140, the temperature value, pressure value, and gas concentration value in each manhole cover are obtained by deploying temperature sensors, pressure sensors, and gas concentration sensors in each manhole cover, and the temperature value and pressure value in each manhole cover are constructed into a temperature vector and a pressure vector according to the arrangement of the nodes in the topological matrix, and the temperature vector and the pressure vector are encoded into a high-dimensional feature space using an encoder to obtain a temperature encoding vector and a pressure encoding vector. That is, first, the temperature value and pressure value in each manhole cover are collected by the temperature sensor and the pressure sensor deployed in each manhole cover, and considering the subsequent need to effectively fuse the temperature information and pressure information, the gas concentration value is also collected by the gas concentration sensor deployed in each manhole cover. Then, the temperature value and pressure value in each manhole cover are constructed into a temperature vector and a pressure vector according to the arrangement of the nodes in the topological matrix, so that they can be processed by the subsequent encoder. Finally, the temperature vector and the pressure vector are encoded by an encoder to encode the temperature vector and the pressure vector into a high-dimensional feature space, thereby obtaining a temperature encoding vector and a pressure encoding vector.

[0099] More specifically, in step S150, the temperature encoding vector and the pressure encoding vector are mapped into the high-dimensional feature space of the topological feature map to obtain a temperature feature vector and a pressure feature vector. That is, the temperature encoding vector and the pressure encoding vector are multiplied by the topological feature map to map the temperature encoding vector and the pressure encoding vector into the high-dimensional feature space of the topological feature map, thereby obtaining a temperature feature vector and a pressure feature vector that represent high-dimensional features of key parameters containing topological structure information.

[0100] More specifically, in step S160, the temperature feature vector and the pressure feature vector are fused based on the thermodynamic gain relationship of gas molecules to obtain a fused feature vector, wherein the thermodynamic gain relationship of the gas molecules is determined based on the characteristic values ​​of each position in the pressure feature vector and the temperature feature vector, the standard temperature and the standard atmospheric pressure and the gas concentration value. It should be understood that, considering that the temperature feature vector and the pressure feature vector are only mapping vectors of the temperature value and the pressure value in the high-dimensional feature space of the topological information, the two need to be further fused. Therefore, in the technical solution of the present application, a fused feature vector is obtained from the temperature feature vector and the pressure feature vector based on the thermodynamic gain relationship of gas molecules. That is, the temperature information, the pressure information and the gas concentration information are further fused based on the thermodynamic gain relationship of gas molecules to achieve effective joint modeling of the various monitoring values ​​of the various monitoring sensors in the high-dimensional feature space, thereby improving the accuracy of subsequent classification to achieve effective monitoring.

[0101] More specifically, in step S170, the fused feature vector is passed through a classifier to obtain a classification result, which is used to indicate whether the monitoring result is normal. That is, after obtaining the fused feature vector, the fused feature vector can be input into the classifier to obtain a classification result indicating whether the monitoring result is normal. In particular, since the eigenvalue of each position in the fused feature vector corresponds to a manhole cover, it is also possible to determine whether a manhole cover has an abnormality based on the Softmax classification function value of each position in the fused feature vector relative to the whole.

[0102] In summary, the monitoring method of the IoT-based community security management and monitoring system based on the embodiment of the present application is explained, which obtains a topological feature map based on the topological structure of the manhole covers in the community and obtains the temperature feature vector and the pressure feature vector thereby, thereby obtaining a high-dimensional feature expression of key parameters containing topological structure information, and further integrates the temperature information, pressure information and gas concentration information based on the thermodynamic gain relationship of gas molecules, which can realize the effective joint modeling of various monitoring values ​​of various monitoring sensors in the high-dimensional feature space, thereby improving the accuracy of subsequent classification to achieve effective monitoring.

[0103] The basic principles of the present application have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, and effects mentioned in this application are merely illustrative and not restrictive, and it should not be assumed that these advantages, strengths, and effects are required of each embodiment of this application. In addition, the specific details disclosed above are merely illustrative and facilitating understanding, and are not restrictive. The above details do not limit this application to necessarily being implemented using the above specific details.

[0104] The block diagrams of the devices, devices, equipment, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As will be appreciated by those skilled in the art, these devices, devices, equipment, and systems can be connected, arranged, or configured in any manner. Words such as "include," "comprise," "have," and the like are open-ended words, meaning "including but not limited to," and can be used interchangeably therewith. The words "or" and "and" used herein refer to the words "and / or" and can be used interchangeably therewith, unless the context clearly indicates otherwise. The word "such as" used herein refers to the phrase "such as but not limited to," and can be used interchangeably therewith.

[0105] It should also be noted that in the apparatus, device, and method of the present application, each component or each step can be decomposed and / or recombined, and such decomposition and / or recombination should be regarded as equivalent solutions of the present application.

[0106] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0107] The above description has been provided for the purpose of illustration and description. Furthermore, this description is not intended to limit the embodiments of the present application to the forms disclosed herein. Although a number of example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. A community security management and monitoring system based on the Internet of Things, characterized in that: include: a topology data unit, configured to convert the topology data of the manhole covers in the cell to be monitored into a topology matrix, wherein each node in the topology data represents a manhole cover, the distance between nodes represents the distance between two manhole covers, the eigenvalue of each off-diagonal position in the topology matrix is ​​the distance between the corresponding two manhole covers, and the eigenvalue of each diagonal position in the topology matrix is ​​0; An implicit spatial encoding unit, configured to obtain a topological feature map from the topological matrix using a convolutional neural network as a feature extractor, wherein the topological feature map is used to represent high-dimensional implicit topological association features; a sensor data acquisition unit, configured to obtain the temperature value, pressure value and gas concentration value in each of the manhole covers through a temperature sensor, a pressure sensor and a gas concentration sensor deployed in each of the manhole covers; a vector construction unit, configured to construct the temperature value and the pressure value in each manhole cover into a temperature vector and a pressure vector according to the arrangement of the nodes in the topological matrix, and to use an encoder to encode the temperature vector and the pressure vector into a high-dimensional feature space to obtain a temperature encoding vector and a pressure encoding vector; a mapping unit, configured to map the temperature encoding vector and the pressure encoding vector into a high-dimensional feature space of the topological feature graph to obtain a temperature feature vector and a pressure feature vector; a pressure and temperature fusion unit, configured to fuse the temperature eigenvector and the pressure eigenvector based on a thermodynamic gain relationship of gas molecules to obtain a fused eigenvector, wherein the thermodynamic gain relationship of gas molecules is determined based on eigenvalues ​​of respective positions in the pressure eigenvector and the temperature eigenvector, standard temperature, standard atmospheric pressure, and gas concentration value; and a monitoring result generating unit, configured to pass the fused feature vector through a classifier to obtain a classification result, wherein the classification result is used to indicate whether the monitoring result is normal; Wherein, the vector construction unit is used to: Performing full-position encoding on the pressure vector using a fully connected layer of the encoder to extract high-dimensional implicit features of eigenvalues ​​at each position in the pressure vector; Performing one-dimensional convolution encoding on the pressure vector using a one-dimensional convolution layer of the encoder to extract high-dimensional correlation features of correlations between feature values ​​at adjacent positions in the pressure vector to obtain the pressure encoding vector; Performing full-position encoding on the temperature vector using the fully connected layer of the encoder to extract high-dimensional implicit features of the eigenvalues ​​at each position in the temperature vector; and Performing one-dimensional convolution encoding on the temperature vector using a one-dimensional convolution layer of the encoder to extract high-dimensional correlation features of correlations between feature values ​​at adjacent positions in the temperature vector to obtain the temperature encoding vector; The mapping unit is configured to: perform matrix multiplication on the topological feature map and the temperature encoding vector to map the temperature encoding vector to a high-dimensional feature space of the topological feature map to obtain the temperature feature vector; and perform matrix multiplication on the topological feature map and the pressure encoding vector to map the pressure encoding vector to the high-dimensional feature space of the topological feature map to obtain the pressure feature vector; The pressure and temperature fusion unit is configured to fuse the temperature characteristic vector and the pressure characteristic vector using the following formula based on the thermodynamic gain relationship of gas molecules to obtain a fused characteristic vector; The formula is: ， ， Where i and g refer to the isotopes of the gas. and is the eigenvalue of the corresponding position of the temperature eigenvector and the pressure eigenvector, and is standard temperature and standard atmospheric pressure, and It represents the total number of molecules per unit volume and the heat absorption cross section of a given gas mixture, which is obtained based on the gas concentration values ​​of various gases obtained by the gas concentration sensor.

2. The cell security management and monitoring system based on the Internet of Things according to claim 1, wherein: The topology data unit is further used to: number the manhole covers in the cell to be monitored; arrange the numbered manhole covers according to the rows and columns of the matrix; and fill each non-diagonal position of the matrix with the distance between the corresponding two manhole covers and fill each diagonal position of the matrix with a 0 value to obtain the topology matrix.

3. The cell security management and monitoring system based on the Internet of Things according to claim 2, wherein: The convolutional neural network processes the topology matrix using the following formula to obtain the topological feature map; Wherein, the formula is: ， in, is the input of the i-th layer convolutional neural network, is the output of the i-th layer convolutional neural network, is the filter of the i-th layer convolutional neural network, and is the bias vector of the i-th layer convolutional neural network, represents a non-linear activation function.

4. The cell security management and monitoring system based on the Internet of Things according to claim 1, wherein: The monitoring result generating unit is further used to: calculate the Softmax classification function value of each position in the fused feature vector as the probability value of whether the manhole cover corresponding to each position is abnormal, the Softmax classification function value is the weighted sum of the natural exponential function value of the eigenvalue of each position raised to the power divided by the natural exponential function value of the eigenvalue of each position raised to the power; and based on the comparison between the Softmax classification function value and a preset threshold, determine whether the manhole cover corresponding to each position is normal.

5. A monitoring method for a cell security management and monitoring system based on the Internet of Things, characterized in that: include: Convert the topological structure data of the manhole covers in the cell to be monitored into a topological matrix, wherein each node in the topological structure data represents a manhole cover, the distance between nodes represents the distance between two manhole covers, the eigenvalue of each non-diagonal position in the topological matrix is ​​the distance between the corresponding two manhole covers, and the eigenvalue of each diagonal position in the topological matrix is ​​0; A convolutional neural network is used as a feature extractor to obtain a topological feature map from the topological matrix, wherein the topological feature map is used to represent high-dimensional implicit topological association features; Obtaining the temperature value, pressure value and gas concentration value in each of the manhole covers by means of a temperature sensor, a pressure sensor and a gas concentration sensor deployed in each of the manhole covers; Constructing the temperature value and the pressure value in each manhole cover into a temperature vector and a pressure vector according to the arrangement of the nodes in the topological matrix, and encoding the temperature vector and the pressure vector into a high-dimensional feature space using an encoder to obtain a temperature encoding vector and a pressure encoding vector; Mapping the temperature encoding vector and the pressure encoding vector into the high-dimensional feature space of the topological feature graph to obtain a temperature feature vector and a pressure feature vector; fusing the temperature eigenvector and the pressure eigenvector based on a thermodynamic gain relationship of gas molecules to obtain a fused eigenvector, wherein the thermodynamic gain relationship of gas molecules is determined based on eigenvalues ​​of respective positions in the pressure eigenvector and the temperature eigenvector, standard temperature, standard atmospheric pressure, and gas concentration value; and Passing the fused feature vector through a classifier to obtain a classification result, wherein the classification result is used to indicate whether the monitoring result is normal; The temperature value and pressure value in each manhole cover are constructed into a temperature vector and a pressure vector according to the arrangement of the nodes in the topological matrix, and an encoder is used to encode the temperature vector and the pressure vector into a high-dimensional feature space to obtain a temperature encoding vector and a pressure encoding vector, including: Performing full-position encoding on the pressure vector using a fully connected layer of the encoder to extract high-dimensional implicit features of eigenvalues ​​at each position in the pressure vector; Performing one-dimensional convolution encoding on the pressure vector using a one-dimensional convolution layer of the encoder to extract high-dimensional correlation features of correlations between feature values ​​at adjacent positions in the pressure vector to obtain the pressure encoding vector; Performing full-position encoding on the temperature vector using the fully connected layer of the encoder to extract high-dimensional implicit features of the eigenvalues ​​at each position in the temperature vector; and Performing one-dimensional convolution encoding on the temperature vector using a one-dimensional convolution layer of the encoder to extract high-dimensional correlation features of correlations between feature values ​​at adjacent positions in the temperature vector to obtain the temperature encoding vector; The step of mapping the temperature encoding vector and the pressure encoding vector into the high-dimensional feature space of the topological feature map to obtain the temperature feature vector and the pressure feature vector comprises: performing matrix multiplication on the topological feature map and the temperature encoding vector to map the temperature encoding vector into the high-dimensional feature space of the topological feature map to obtain the temperature feature vector; and performing matrix multiplication on the topological feature map and the pressure encoding vector to map the pressure encoding vector into the high-dimensional feature space of the topological feature map to obtain the pressure feature vector. The step of fusing the temperature characteristic vector and the pressure characteristic vector based on the thermodynamic gain relationship of gas molecules to obtain a fused characteristic vector includes: fusing the temperature characteristic vector and the pressure characteristic vector based on the thermodynamic gain relationship of gas molecules using the following formula to obtain a fused characteristic vector; The formula is: ， ， Where i and g refer to the isotopes of the gas. and is the eigenvalue of the corresponding position of the temperature eigenvector and the pressure eigenvector, and is standard temperature and standard atmospheric pressure, and It represents the total number of molecules per unit volume and the heat absorption cross section of a given gas mixture, which is obtained based on the gas concentration values ​​of various gases obtained by the gas concentration sensor.

6. The monitoring method of the cell security management and monitoring system based on the Internet of Things according to claim 5, wherein: The topological structure data of the manhole covers in the community to be monitored is converted into a topological matrix, including: Numbering the manhole covers in the community to be monitored; Arrange the numbered manhole covers in rows and columns of a matrix; and The topological matrix is ​​obtained by filling each non-diagonal position of the matrix with the distance between two manhole covers and filling each diagonal position of the matrix with a 0 value.

Images