An Electric Safety Supervision Method and System

Abstract

The present invention relates to the technical field of power consumption monitoring, and discloses a power consumption safety supervision method and system, including obtaining multi-dimensional information of a plurality of data terminals and K data transfer terminals in a target area at fixed time intervals within a first preset time period; dividing the plurality of data terminals into M wireless terminals and N wired terminals based on the connection method, and collecting fine-grained transmission data of data packets of each wireless terminal at fixed time intervals; obtaining the position coordinates of a plurality of base stations in the target area, and analyzing each wireless terminal based on the position coordinates of the base stations to obtain the transmission path of each wireless terminal; processing the transmission paths and fine-grained transmission data of the M wireless terminals to obtain sample data and sample labels, and training to obtain a delay time prediction model; based on the delay time prediction model, dynamically generating a data transmission scheme for the M wireless terminals in the target area to minimize the time window.

Description

Technical Field

[0001] The present invention relates to the technical field of power consumption monitoring, and more specifically, to a power consumption safety supervision method and system. Background Art

[0002] In modern power monitoring systems, data collection usually adopts wired or wireless transmission methods to obtain real-time monitoring data from smart meters. After preprocessing through a data transfer area, the data is transmitted to a monitoring center. Due to differences in the arrival times of data transmitted from monitoring points, to ensure that data with the same timestamp can be processed synchronously, the system usually adopts a time window determined based on the earliest arrival time and the latest arrival time. However, the setting of the time window directly affects the data sampling frequency. When the sampling interval is less than the time window, data congestion will occur. When the sampling interval is greater than or equal to the time window, spectral leakage may occur, resulting in a situation where the monitored data appears normal but the actual data is abnormal. Therefore, to avoid data acquisition latency in the system and reduce the possibility of spectral leakage, it is necessary to minimize the size of the time window as much as possible. Summary of the Invention

[0003] The present invention provides a power consumption safety supervision system to solve the technical problems proposed in the background art.

[0004] The present invention provides a power consumption safety supervision system, including:

[0005] A data collection module, configured to obtain multi-dimensional information of a plurality of data terminals and K data transfer terminals in a target area at fixed time intervals within a first preset time period, including: connection methods and position coordinates;

[0006] A data processing module, configured to divide a plurality of data terminals into M wireless terminals and N wired terminals based on the connection methods, and collect fine-grained transmission data of data packets of each wireless terminal at fixed time intervals, including: transmission time and environmental data;

[0007] A data analysis module, configured to obtain the position coordinates of a plurality of base stations in the target area, and analyze each wireless terminal based on the position coordinates of the base stations to obtain the transmission path of each wireless terminal;

[0008] A model construction module, configured to process the transmission paths and fine-grained transmission data of M wireless terminals to obtain sample data and sample labels, and train a delay time prediction model;

[0009] A time window optimization module, configured to dynamically generate a data transmission scheme for M wireless terminals in the target area based on the delay time prediction model within a second preset time period to minimize the time window.

[0010] Further, several data terminals are divided into M wireless terminals and N wired terminals based on the connection method, including: The connection method includes: wired connection and wireless connection; Based on the connection method between the data terminal and the data relay terminal, the data terminal is divided into M wireless terminals and N wired terminals.

[0011] Further, each wireless terminal is analyzed based on the position coordinates of the base station, including: Obtaining the data packet of each wireless terminal at the q-th moment, and parsing the data packet to obtain the propagation path and transfer moment of the monitoring data of the corresponding wireless terminal; where the propagation path represents the base station through which the monitoring data of the wireless terminal reaches the corresponding data relay terminal, and the transfer moment represents the moment when the monitoring data of the wireless terminal reaches each passing base station; where 1 ≤ q ≤ T1 / t, T1 represents the first preset time period, and t represents the fixed time interval; The data packet includes: monitoring data, passing base stations, and the moment of passing each base station.

[0012] Further, the fine-grained transmission data is processed, including: within the q-th moment of the first preset time period, calculating the delay time according to the transmission time of the propagation path of the monitoring data of the m-th wireless terminal, including: Obtaining the Euclidean distance between the wireless terminal and the base station, the Euclidean distance between the base stations, and the Euclidean distance between the base station and the data relay terminal in the corresponding propagation path; Based on the propagation speed of electromagnetic waves, obtaining the standard transmission time; Taking the difference time between the standard transmission time and the transmission time as the delay time; where 1 ≤ m ≤ M.

[0013] Further, sample data and sample labels, including: Normalizing the environmental data of each wireless terminal, base station, and data relay terminal in the propagation path of the monitoring data of the m-th wireless terminal at the q-th moment of the first preset time period to obtain sample data; where the environmental data includes several types of weather, and real number encoding processing is performed on each type of weather respectively; Normalizing the delay time of the propagation path of the monitoring data of the m-th wireless terminal to obtain sample labels.

[0014] Further, the delay time prediction model is constructed based on a one-dimensional convolutional neural network and converges through the mean square error loss function.

[0015] Further, a data transmission scheme is dynamically generated for the M wireless terminals in the target area, including:

[0016] Step 71, retaining the pairing relationship between the N wired terminals and the corresponding data relay terminals, and formatting the pairing relationship between the M wireless terminals and the corresponding data relay terminals; where the pairing relationship represents the data connection relationship between the wireless terminal and the data relay terminal.

[0017] Step 72, at the s-th moment of the second preset time period, obtain the environmental data of each wireless terminal, base station, and data relay terminal at the s-th moment; where 1 ≤ s ≤ T2 / t, and T2 represents the second preset time period;

[0018] Step 73, generate R re-pairing schemes for M wireless terminals and K data relay terminals that meet the constraint conditions. Each re-pairing scheme represents the pairing relationship generated between M wireless terminals and K data relay terminals;

[0019] The constraint conditions include: the number L of data terminals connected to each data relay terminal, where Lmin ≤ L ≤ Lmax; where Lmin represents the minimum number of data terminals connected to each data relay terminal, and Lmax represents the maximum number of data terminals connected to each data relay terminal; for the wireless terminals having a pairing relationship with the data relay terminal, the corresponding Euclidean distance is less than or equal to the preset distance threshold D;

[0020] The fitness function includes: obtaining U predicted propagation paths for the monitoring data of the m-th wireless terminal in the r-th re-pairing scheme to reach the n-th data relay terminal; where 1 ≤ r ≤ R and 1 ≤ n ≤ N;

[0021] According to the environmental data of each wireless terminal, base station, and data relay terminal in each predicted propagation path, calculate the predicted delay time of each predicted propagation path through the delay time prediction model;

[0022] Based on the sum value time of the standard transmission time and the predicted delay time of each predicted propagation path, select the minimum sum value time as the transmission time of the m-th wireless terminal in the r-th re-pairing scheme;

[0023] Determine the maximum transmission time Q1 and the minimum transmission time W1 of the M wireless terminals in the r-th re-pairing scheme, and the maximum transmission time Q2 and the minimum transmission time W2 of the N wired terminals;

[0024] ; where represents the time window of the j-th re-pairing scheme, represents the maximum selection operation, represents the minimum selection operation;

[0025] Step 74, take as the fitness function value of the r-th re-pairing scheme, and sort the R re-pairing schemes from smallest to largest based on the fitness function value to obtain a characteristic sorting;

[0026] Step 75, retain a preset number of re-pairing schemes from front to back in the characteristic sorting, and perform crossover and mutation operations on the remaining re-pairing schemes based on the genetic algorithm to update the re-pairing schemes;

[0027] Step 76, repeat Step 74 and Step 75 for a preset number of times to obtain the final feature ranking; use the re - pairing scheme ranked 1 in the feature ranking as the data transmission scheme at the s - th moment in the second preset time period.

[0028] Further, obtaining U predicted propagation paths for the monitoring data of the m - th wireless terminal in the r - th re - pairing scheme to reach the n - th data transfer node includes:

[0029] Establish a transmission topology, where the transmission topology includes: topology nodes and topology edges; map each wireless terminal, data transfer node, and base station to topology nodes; based on the position coordinates, calculate the Euclidean distance between any two topology nodes; if the Euclidean distance is less than or equal to the preset distance threshold D, construct a topology edge between the corresponding topology nodes to obtain the transmission topology;

[0030] Based on the transmission topology, obtain the U paths from the topology node corresponding to the m - th wireless terminal to the topology node corresponding to the n - th data transfer node as the predicted propagation paths.

[0031] In a second aspect, an electricity - using safety supervision method is applied to any one of the electricity - using safety supervision systems described above, and includes:

[0032] Step 81, within a first preset time period, obtain the multi - dimensional information of several data terminals and K data transfer nodes in the target area at fixed time intervals, including: connection methods and position coordinates;

[0033] Step 82, divide several data terminals into M wireless terminals and N wired terminals based on the connection methods, and collect the fine - grained transmission data of the data packets of each wireless terminal at fixed time intervals, including: transmission time and environmental data;

[0034] Step 83, obtain the position coordinates of several base stations in the target area, and analyze each wireless terminal based on the position coordinates of the base stations to obtain the transmission path of each wireless terminal;

[0035] Step 84, process the transmission paths and fine - grained transmission data of the M wireless terminals to obtain sample data and sample labels, and train a delay time prediction model;

[0036] Step 85, within a second preset time period, based on the delay time prediction model, dynamically generate a data transmission scheme for the M wireless terminals in the target area to minimize the time window.

[0037] The beneficial effects of the present invention are as follows: By integrating fine data collection, delay prediction, and dynamic transmission scheme optimization technologies, real-time collection of wireless and wired terminal data is carried out. A delay prediction model is constructed using a one-dimensional convolutional neural network, and then the genetic algorithm is combined to optimize data transfer pairing, achieving the minimization of the transmission time window. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 is a module diagram of an electricity safety supervision system of the present invention;

[0039] Figure 2 is a flowchart of an electricity safety supervision method of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] Now, the subject matter described herein will be discussed with reference to exemplary embodiments. It should be understood that discussing these embodiments is only to enable those skilled in the art to better understand and thus implement the subject matter described herein. Without departing from the scope of protection of the content of this specification, changes can be made to the functions and arrangements of the elements discussed. Each example can omit, substitute, or add various processes or components as needed. Additionally, the features described relative to some examples can also be combined in other examples.

[0041] As Figure 1 shown, an electricity safety supervision system includes:

[0042] A data collection module, configured to obtain multi-dimensional information of a plurality of data terminals and K data transfer terminals in a target area at fixed time intervals within a first preset time period, including: connection mode and position coordinates;

[0043] A data processing module, configured to divide a plurality of data terminals into M wireless terminals and N wired terminals based on the connection mode, and collect fine-grained transmission data of each wireless terminal's data packet at fixed time intervals, including: transmission time and environmental data;

[0044] A data analysis module, configured to obtain the position coordinates of a plurality of base stations in the target area, and analyze each wireless terminal based on the position coordinates of the base stations to obtain the transmission path of each wireless terminal;

[0045] A model construction module, configured to process the transmission paths and fine-grained transmission data of M wireless terminals to obtain sample data and sample labels, and train a delay time prediction model;

[0046] A time window optimization module, configured to dynamically generate a data transmission scheme for M wireless terminals in the target area based on the delay time prediction model within a second preset time period to minimize the time window.

[0047] In an embodiment of the present invention, a number of data terminals are divided into M wireless terminals and N wired terminals based on the connection method, including: The connection method includes: wired connection and wireless connection; based on the connection method between the data terminal and the data transfer terminal, the data terminal is divided into M wireless terminals and N wired terminals.

[0048] Specifically, the system divides the terminals into two categories according to the connection method between the data terminal and the data transfer terminal: wired terminals and wireless terminals. Specifically, the connection between the wired terminal and the data transfer terminal is realized through a physical connection. This connection method itself is relatively stable and not easily affected by the external environment. At the same time, any modification to this connection method requires a high cost; while the wireless terminal relies on wireless connection, which has stronger flexibility, and the connection status can be adjusted according to needs without actual physical transformation.

[0049] In an embodiment of the present invention, each wireless terminal is analyzed based on the position coordinates of the base station, including: obtaining the data packet of each wireless terminal at the q-th moment, and parsing the data packet to obtain the propagation path and transfer moment of the monitoring data of the corresponding wireless terminal; where the propagation path represents the base station through which the monitoring data of the wireless terminal reaches the corresponding data transfer terminal, and the transfer moment represents the moment when the monitoring data of the wireless terminal reaches each transferred base station; where 1≤q≤T1 / t, T1 represents the first preset time period, and t represents the fixed time interval; the data packet includes: monitoring data, the transferred base station, and the moment of passing through each base station.

[0050] Specifically, the system analyzes the transmission path of the wireless terminal by parsing the data packet transmitted by the wireless terminal. Specifically, during the forwarding process of each data packet, it will automatically record each base station passed through and the moment of reaching each base station. By parsing these data packets, the system can restore the complete path that the wireless terminal data passes from the source to the data transfer terminal and obtain the transfer moments of each transfer base station.

[0051] In an embodiment of the present invention, the fine-grained transmission data is processed, including: within the q-th moment of the first preset time period, calculating the delay time according to the transmission time of the propagation path of the monitoring data of the m-th wireless terminal, including: obtaining the Euclidean distance between the wireless terminal and the base station, the Euclidean distance between the base stations, and the Euclidean distance between the base station and the data transfer terminal in the corresponding propagation path; based on the propagation speed of electromagnetic waves, obtaining the standard transmission time; taking the difference time between the standard transmission time and the transmission time as the delay time; where 1≤m≤M.

[0052] Specifically, the system performs fine-grained processing on the data packets transmitted by the wireless terminal to quantify the delay in the actual transmission process. Specifically, first, based on the Euclidean distance between the wireless terminal and the base station, as well as between the base stations, and using the theory that electromagnetic waves propagate at the speed of light, a theoretically required standard transmission time is calculated (i.e., transmission time = distance / speed of light). However, the actual transmission time is often affected by factors such as environmental interference and signal attenuation, and there is a difference from the standard transmission time. By calculating the difference between the actual transmission time and the standard transmission time, a delay time can be obtained, and this delay time is used to quantify the negative impact of environmental factors on data transmission.

[0053] In an embodiment of the present invention, the sample data and sample labels include: normalizing the environmental data of each wireless terminal, base station, and data transfer terminal in the propagation path of the monitoring data of the m-th wireless terminal at the q-th moment within the first preset time period to obtain sample data; wherein, the environmental data includes several types of weather, and real number coding is performed on each type of weather respectively; normalizing the delay time of the propagation path of the monitoring data of the m-th wireless terminal to obtain sample labels.

[0054] Specifically, the system constructs sample data and sample labels for training the delay prediction model by normalizing the monitoring data of the wireless terminal at a specific moment (the q-th moment) within the first preset time period. Specifically, the sample data comes from the environmental data of all nodes (wireless terminals, base stations, and data transfer terminals) in the propagation path of the monitoring data of the m-th wireless terminal. This environmental data contains several types of weather information. By performing real number coding on each weather data and then normalizing it, the data features are within a unified numerical range; while the sample label is to normalize the calculated delay time in this propagation path, aiming to eliminate the influence of different scales on model training, so as to more effectively reflect the influence of environmental factors on data transmission delay.

[0055] For example, the order of a propagation path is , then the corresponding environmental data is , where weather 2 corresponds to wireless terminal m, weather 3 corresponds to base station 1, weather 5 corresponds to base station 2, and weather 4 corresponds to the data transfer terminal. Then, After normalization, sample data is obtained.

[0056] In an embodiment of the present invention, the delay time prediction model is constructed based on a one-dimensional convolutional neural network and converges through a mean square error loss function.

[0057] Specifically, the delay time prediction model is constructed based on a one-dimensional convolutional neural network, and its training process relies on a strictly divided dataset and a scientific verification method. First, all the data collected by wireless terminals, after preprocessing and normalization, constitute the input features of the model, while the corresponding delay time, after normalization, serves as the target output. To ensure that the model has good generalization ability, the dataset is usually divided into a training set, a validation set, and a test set, for example, divided according to the ratio of 70%, 15%, and 15%; at the same time, to further eliminate the contingency of data division and improve data utilization, the method of K-fold cross-validation can also be used, that is, the entire dataset is divided into K subsets, each subset takes turns as the validation set, and the rest is used as the training set, and finally the average is taken to evaluate the model performance. In terms of network structure, the one-dimensional convolutional neural network (1D-CNN) uses the convolutional layer to locally extract the input temporal features, then reduces the dimension and removes noise through the activation function and pooling layer, and finally the fully connected layer maps the extracted features to the delay prediction value. The training objective of the entire model is to minimize the mean square error (MSE) between the predicted value and the actual delay value, which is a commonly used loss function to measure prediction error. The MSE loss function can effectively punish large deviations by calculating the average of the squares of the differences between the predicted value and the true value, thereby guiding the model to correct errors faster. During the training process, the backpropagation algorithm is combined with gradient descent or advanced optimization algorithms such as Adam to gradually update the model parameters, so that the value of the loss function continuously decreases until convergence. With the help of K-fold cross-validation, not only can the stability of the model under different data divisions be verified, but also a reliable basis can be provided for adjusting hyperparameters (such as learning rate, convolutional kernel size, and number of layers, etc.), thereby further improving the prediction accuracy and robustness of the model. Generally speaking, this training strategy enables the delay time prediction model to fully utilize the temporal and environmental information in the data, accurately capture the complex factors affecting data transmission delay, and achieve accurate prediction of transmission delay.

[0058] In one embodiment of the present invention, a data transmission scheme is dynamically generated for M wireless terminals in a target area, including:

[0059] Step 71, retain the pairing relationship between N wired terminals and the corresponding data transfer terminals, and format the pairing relationship between M wireless terminals and the corresponding data transfer terminals; wherein, the pairing relationship represents the data connection relationship between the wireless terminal and the data transfer terminal;

[0060] Step 72, at the s-th moment of the second preset time period, obtain the environmental data of each wireless terminal, base station, and data transfer terminal at the s-th moment; wherein, 1 ≤ s ≤ T2 / t, and T2 represents the second preset time period;

[0061] Step 73: Generate R re - pairing schemes for M wireless terminals and K data relay terminals that meet the constraint conditions. Each re - pairing scheme represents the pairing relationship between M wireless terminals and K data relay terminals;

[0062] The constraint conditions include: the number L of data terminals connected to each data relay terminal, where Lmin ≤ L ≤ Lmax; here, Lmin represents the minimum number of data terminals connected to each data relay terminal, and Lmax represents the maximum number of data terminals connected to each data relay terminal; for the wireless terminals having a pairing relationship with the data relay terminal, the corresponding Euclidean distance is less than or equal to the preset distance threshold D;

[0063] The fitness function includes: obtaining U predicted propagation paths for the monitoring data of the m - th wireless terminal in the r - th re - pairing scheme to reach the n - th data relay terminal; where 1 ≤ r ≤ R, 1 ≤ n ≤ N;

[0064] According to the environmental data of each wireless terminal, base station, and data relay terminal in each predicted propagation path, calculate the predicted delay time of each predicted propagation path through the delay time prediction model;

[0065] Based on the sum value of the standard transmission time and the predicted delay time of each predicted propagation path, select the minimum sum value as the transmission time of the m - th wireless terminal in the r - th re - pairing scheme;

[0066] Determine the maximum transmission time Q1 and the minimum transmission time W1 of the M wireless terminals in the r - th re - pairing scheme, and the maximum transmission time Q2 and the minimum transmission time W2 of the N wired terminals;

[0067] ; where, represents the time window of the j - th re - pairing scheme, represents the maximum selection operation, represents the minimum selection operation;

[0068] Step 74: Take as the fitness function value of the r - th re - pairing scheme, and sort the R re - pairing schemes from small to large based on the fitness function value to obtain a feature sorting;

[0069] Step 75: Retain a preset number of re - pairing schemes from the front to the back of the feature sorting, and perform crossover and mutation operations on the remaining re - pairing schemes based on the genetic algorithm to update the re - pairing schemes;

[0070] Step 76: Repeat Step 74 and Step 75 for a preset number of times to obtain the final feature sorting; take the re - pairing scheme with the first position in the feature sorting as the data transmission scheme at the s - th moment in the second preset time period.

[0071] Specifically, the system retains a stable and fixed pairing relationship between N wired terminals and the data relay terminal. For the connection between M wireless terminals and the data relay terminal, reformatting processing is required (step 71), which lays a data foundation for subsequent dynamic re - pairing. Next, at a certain moment s in the second preset time period (step 72), the system collects the environmental data of each wireless terminal, base station, and data relay terminal at that time, providing real - time information for dynamically adjusting the transmission scheme. In step 73, the system generates R re - pairing schemes that meet the preset constraint conditions based on this real - time data. The constraint conditions include: the number of wireless terminals connected to each data relay terminal must satisfy between Lmin and Lmax. At the same time, the Euclidean distance between the wireless terminal and the data relay terminal must be less than or equal to a preset distance threshold D. For each candidate scheme, the system further obtains U predicted propagation paths from each wireless terminal to the designated data relay terminal, and uses the previously trained delay prediction model to calculate the predicted delay time of each path. By summing the standard transmission time (calculated based on the Euclidean distance and the speed of light) of each propagation path and its predicted delay time, the smallest sum value is selected as the transmission time of the wireless terminal in this scheme, and at the same time, the maximum and minimum transmission times of the wireless terminals and wired terminals in the entire scheme are calculated, thereby obtaining the corresponding transmission time window. In step 74, the system sorts the R re - pairing schemes according to the fitness function value (i.e., the time - window index) calculated above, and selects the optimal candidate scheme. Then in step 75, the system not only retains some of the schemes with higher rankings, but also updates the remaining schemes using the crossover and mutation operations in the genetic algorithm to introduce more diversity and optimization potential. After multiple iterations (step 76), the re - pairing scheme ranked first is finally selected as the best data transmission scheme at this moment. This method not only considers the physical connection constraints and distance limitations between wireless terminals and data relay terminals, but also combines real - time environmental data and the output of the delay prediction model, and is continuously optimized through the genetic algorithm, thus effectively narrowing the transmission time window.

[0072] In an embodiment of the present invention, obtaining U predicted propagation paths for the monitoring data of the m - th wireless terminal in the r - th re - pairing scheme to reach the n - th data relay terminal includes:

[0073] Establish a transmission topology, where the transmission topology includes: topology nodes and topology edges; map each wireless terminal, data relay terminal, and base station as topology nodes; based on the position coordinates, calculate the Euclidean distance between any two topology nodes; if the Euclidean distance is less than or equal to the preset distance threshold D, then construct topology edges between the corresponding topology nodes to obtain the transmission topology;

[0074] Obtain the topological node corresponding to the m-th wireless terminal based on the transmission topology, and use the U paths reaching the topological node corresponding to the n-th data relay as the predicted propagation paths.

[0075] Specifically, in order to predict the propagation path of the data of the m-th wireless terminal from its departure to the n-th data relay, the system first constructs a transmission topology structure. Map all involved data entities (wireless terminals, data relays, and base stations) to topological nodes. Based on the location information of each node, calculate the Euclidean distance between any two topological nodes; if the Euclidean distance between two nodes is less than or equal to the preset distance threshold D, establish a topological edge between them, thus forming a transmission network reflecting the possibility of actual physical connection. Using the constructed transmission topology, the system starts from the topological node corresponding to the m-th wireless terminal and searches for all possible paths to reach the topological node corresponding to the n-th data relay; screen out U of them as the predicted propagation paths to evaluate the delay and reliability during the process of data transmission from the wireless terminal to the data relay.

[0076] An electrical safety supervision method, applied to any one of the described electrical safety supervision systems, includes:

[0077] Step 81, within a first preset time period, obtain the multi-dimensional information of several data terminals and K data relays in the target area at fixed time intervals, including: connection method and position coordinates;

[0078] Step 82, divide several data terminals into M wireless terminals and N wired terminals based on the connection method, and collect the fine-grained transmission data of the data packets of each wireless terminal at fixed time intervals, including: transmission time and environmental data;

[0079] Step 83, obtain the position coordinates of several base stations in the target area, and analyze each wireless terminal based on the position coordinates of the base stations to obtain the transmission path of each wireless terminal;

[0080] Step 84, process the transmission paths and fine-grained transmission data of the M wireless terminals to obtain sample data and sample labels, and train a delay time prediction model;

[0081] Step 85, within a second preset time period, based on the delay time prediction model, dynamically generate a data transmission scheme for the M wireless terminals in the target area to minimize the time window.

[0082] The above has described the embodiments of this embodiment, but this embodiment is not limited to the above specific implementation manners. The above specific implementation manners are merely illustrative and not restrictive. Under the inspiration of this embodiment, those of ordinary skill in the art can also make many forms, all of which fall within the protection scope of this embodiment.

Claims

1. A power safety supervision system, characterized in that: include: The data acquisition module is used to obtain multi-dimensional information of a plurality of data terminals and K data transfer terminals in the target area at fixed time intervals within a first preset time period, including: connection mode and location coordinates; A data processing module, used to divide a number of data terminals into M wireless terminals and N wired terminals based on the connection mode, and collect fine-grained transmission data of the data packets of each wireless terminal at fixed time intervals, including: transmission time and environment data; A data analysis module is used to obtain the location coordinates of several base stations in the target area, and analyze each wireless terminal based on the location coordinates of the base stations to obtain the propagation path of each wireless terminal; A model building module is used to process the propagation paths and fine-grained transmission data of M wireless terminals to obtain sample data and sample labels, and train a delay time prediction model; The time window optimization module is used to dynamically generate a data transmission plan for M wireless terminals in the target area within a second preset time period based on the delay time prediction model to minimize the time window.

2. The power safety supervision system according to claim 1, characterized in that: Based on the connection mode, a number of data terminals are divided into M wireless terminals and N wired terminals, including: the connection mode includes: wired connection and wireless connection; based on the connection mode between the data terminal and the data transfer terminal, the data terminal is divided into M wireless terminals and N wired terminals.

3. The power safety supervision system according to claim 2, characterized in that: Each wireless terminal is analyzed based on the location coordinates of the base station, including: obtaining the data packet of each wireless terminal at the qth moment, and parsing the data packet to obtain the propagation path and transfer time of the corresponding wireless terminal's monitoring data; wherein the propagation path represents the base station via which the monitoring data of the wireless terminal arrives at the corresponding data transfer terminal, and the transfer time represents the time when the monitoring data of the wireless terminal arrives at each base station via which it is transferred; wherein 1≤q≤T1 / t, T1 represents a first preset time period, and t represents a fixed time interval; the data packet includes: monitoring data, base stations via which it is transferred, and the time at which it is transferred to each base station.

4. The power safety supervision system according to claim 3, characterized in that: Processing fine-grained transmission data includes: within the qth moment of the first preset time period, calculating the delay time according to the transmission time of the propagation path of the monitoring data of the mth wireless terminal, including: obtaining the Euclidean distance between the wireless terminal and the base station, the Euclidean distance between the base stations, and the Euclidean distance between the base station and the data transfer terminal in the corresponding propagation path; obtaining the standard transmission time based on the transmission speed of the electromagnetic wave; and taking the difference time between the standard transmission time and the transmission time as the delay time; wherein 1≤m≤M.

5. The power safety supervision system according to claim 4, characterized in that: Obtaining sample data and sample labels includes: normalizing the environmental data of each wireless terminal, base station and data transfer terminal in the propagation path of the monitoring data of the m-th wireless terminal at the q-th moment in the first preset time period to obtain sample data; wherein the environmental data includes several types of weather, and real number coding is performed on each type of weather; and normalizing the delay time of the propagation path of the monitoring data of the m-th wireless terminal to obtain the sample label.

6. The power safety supervision system according to claim 5, characterized in that: The delay time prediction model is built based on a one-dimensional convolutional neural network and converges through the mean square error loss function.

7. The power safety supervision system according to claim 6, characterized in that: Dynamically generate a data transmission plan for M wireless terminals in the target area, including: Step 71, retaining the pairing relationship between N wired terminals and the corresponding data transfer terminals, and performing formatting processing on the pairing relationship between M wireless terminals and the corresponding data transfer terminals; wherein the pairing relationship represents the data connection relationship between the wireless terminal and the data transfer terminal; Step 72, at the s-th moment in the second preset time period, obtain the environmental data of each wireless terminal, base station and data transfer terminal at the s-th moment; wherein 1≤s≤T2 / t, T2 represents the second preset time period; Step 73, generating R re-pairing schemes of the M wireless terminals and the K data transfer terminals that meet the constraint conditions, each re-pairing scheme represents generating a pairing relationship between the M wireless terminals and the K data transfer terminals; The constraint conditions include: the number of data terminals connected to each data transfer terminal, Lmin≤L≤Lmax; where Lmin represents the minimum number of data terminals connected to each data transfer terminal, and Lmax represents the maximum number of data terminals connected to each data transfer terminal; the Euclidean distance of the wireless terminal that has established a pairing relationship with the data transfer terminal is less than or equal to the preset distance threshold D; The fitness function includes: obtaining U predicted propagation paths for the monitoring data of the mth wireless terminal in the rth re-pairing scheme to reach the nth data transfer terminal; wherein 1≤r≤R, 1≤n≤K; According to the environmental data of each wireless terminal, base station and data transfer terminal in each predicted propagation path, the predicted delay time of each predicted propagation path is calculated by the delay time prediction model; Based on the sum of the standard transmission time and the predicted delay time of each predicted propagation path, selecting the minimum sum time as the transmission time of the mth wireless terminal in the rth re-pairing scheme; Determine the maximum transmission time Q1 and the minimum transmission time W1 of the M wireless terminals in the rth re-pairing scheme, and the maximum transmission time Q2 and the minimum transmission time W2 of the N wired terminals; Time(r)=max(Q1,Q2)-min(W1,W2); where Time(r) represents the time window of the rth re-pairing scheme, max represents the maximum selection operation, and min represents the minimum selection operation; Step 74, taking Time(r) as the fitness function value of the rth re-pairing scheme, and sorting the R re-pairing schemes from small to large based on the fitness function value to obtain a feature sorting; Step 75, sorting the features from front to back to retain a preset number of re-pairing schemes, and performing crossover and mutation operations on the remaining re-pairing schemes based on a genetic algorithm to update the re-pairing schemes; Step 76, repeating steps 74 and 75 for a preset number of times to obtain a final feature ranking; using the re-pairing scheme ranked 1 in the feature ranking as the data transmission scheme at the sth moment in the second preset time period.

8. The power safety supervision system according to claim 7, characterized in that: Obtaining U predicted propagation paths for the monitoring data of the mth wireless terminal in the rth re-pairing scheme to reach the nth data transfer terminal, including: Establish a transmission topology, which includes topological nodes and topological edges; map each wireless terminal, data transfer terminal and base station to a topological node; calculate the Euclidean distance between any two topological nodes based on the location coordinates; if the Euclidean distance is less than or equal to a preset distance threshold D, construct a topological edge between the corresponding topological nodes to obtain a transmission topology; Based on the transmission topology, the topological node corresponding to the m-th wireless terminal is obtained, and U paths to the topological node corresponding to the n-th data transfer terminal are used as the predicted propagation paths.

9. A method for monitoring electricity safety, characterized in that: An electricity safety supervision system applied to any one of claims 1 to 8, comprising: Step 81, within a first preset time period, obtaining multi-dimensional information of a plurality of data terminals and K data transfer terminals in a target area at fixed time intervals, including: connection modes and location coordinates; Step 82, dividing a number of data terminals into M wireless terminals and N wired terminals based on the connection mode, and collecting fine-grained transmission data of the data packets of each wireless terminal at fixed time intervals, including: transmission time and environment data; Step 83, obtaining the position coordinates of several base stations in the target area, and analyzing each wireless terminal based on the position coordinates of the base stations to obtain the propagation path of each wireless terminal; Step 84, processing the propagation paths and fine-grained transmission data of the M wireless terminals to obtain sample data and sample labels, and training to obtain a delay time prediction model; Step 85: Dynamically generate a data transmission plan for the M wireless terminals in the target area within the second preset time period based on the delay time prediction model to minimize the time window.

Images