A method and system for monitoring power transmission and transformation network based on Internet of Things

Abstract

The present invention relates to the field of smart grid technology, and in particular to a transmission and transformation grid monitoring method and system based on the Internet of Things. The method characterizes the noise complexity of an initial data matrix formed by monitoring data vectors through reconstruction difference evaluation, thereby determining the number of denoising iterations. Then, iterative denoising is performed through a denoising model, and the expected denoising effect can be achieved with as few denoising iterations as possible. Therefore, while ensuring the reliability of the target data matrix, that is, while ensuring the accuracy of grid monitoring, the time consumption of the denoising process is reduced. In addition, fault identification is performed in matrix form, and grid early warning information is determined based on the obtained fault identification probability vector. Compared with the prior art method of iterative prediction using a time series prediction model, the method does not require the introduction of time-consuming iterative prediction, thereby effectively improving the timeliness of grid monitoring.

Description

Technical Field

[0001] The present invention relates to the technical field of smart grids, and in particular to a method and system for monitoring a transmission and transformation grid based on the Internet of Things. Background Art

[0002] In the operation of modern power transmission and transformation grid distribution stations, the stability and safety of power grid equipment are crucial to the reliable operation of the power system. With the continuous expansion of the power grid scale and the increase in equipment complexity, real-time monitoring and fault warning of equipment have become increasingly important. The rapid development of Internet of Things technology has provided new possibilities for health monitoring of power grid equipment. Through real-time monitoring of equipment operating status and environmental conditions through a variety of sensors, comprehensive data support is provided for the safe operation of the power grid.

[0003] Traditional monitoring methods for distribution stations in power transmission and transformation networks typically rely on a single data source or simple rule-based judgments to implement equipment status monitoring, making it difficult to effectively integrate multiple data types. Traditional monitoring methods are unable to accurately determine equipment health status when faced with complex equipment failures and changing environmental conditions, resulting in delayed fault identification or misjudgment.

[0004] Based on the above situation, the existing technology has proposed a data monitoring method based on an artificial intelligence model, which usually predicts equipment data based on a time series prediction model, and then realizes the early prediction of faults. However, the time series prediction model usually adopts an iterative prediction method. In the scenario of multiple power grid equipment, the increase in data volume will lead to an increase in prediction time, which makes it difficult to meet the timeliness requirements of power grid monitoring.

[0005] In addition, the data collected by the sensor may contain various types of noise. Under the premise of unknown noise type, it is difficult to denoise the data through an effective denoising algorithm, which will reduce the reliability of the data collected by the sensor and further reduce the accuracy of power grid monitoring.

[0006] Therefore, how to improve the accuracy and timeliness of power grid monitoring has become an urgent problem to be solved. Summary of the Invention

[0007] In order to solve the above technical problems, the technical solution adopted by the present invention is a transmission and transformation network monitoring method based on the Internet of Things, which includes the following steps:

[0008] S101 , forming an initial data matrix from monitoring data vectors respectively acquired at M preset time points, where M is a positive integer.

[0009] S102, performing mask processing on the initial data matrix to obtain a masked data matrix.

[0010] S103: Input the mask data matrix into the trained reconstruction model to obtain a reconstructed data matrix.

[0011] S104: Determine a reconstruction difference evaluation value according to the initial data matrix and the reconstructed data matrix.

[0012] S105 : Mapping the reconstruction difference evaluation value to the number of denoising iterations according to a preset first mapping function.

[0013] S106 , performing iterative denoising on the initial data matrix using a trained denoising model according to the number of denoising iterations to obtain a target data matrix.

[0014] S107: Input the target data matrix into the trained fault recognition model to obtain a fault recognition probability vector.

[0015] S108 : When the fault identification probability vector meets a preset condition, power grid early warning information is generated.

[0016] The present invention also provides a power transmission and transformation network monitoring system based on the Internet of Things, the power transmission and transformation network monitoring system based on the Internet of Things comprising:

[0017] The matrix forming module is used to form an initial data matrix from the monitoring data vectors respectively obtained at M preset time points, where M is a positive integer.

[0018] The matrix mask module is used to perform mask processing on the initial data matrix to obtain a masked data matrix.

[0019] The matrix reconstruction module is used to input the mask data matrix into the trained reconstruction model to obtain a reconstructed data matrix.

[0020] The difference evaluation module is used to determine a reconstruction difference evaluation value according to the initial data matrix and the reconstructed data matrix.

[0021] The number mapping module is used to map the reconstruction difference evaluation value to the number of denoising iterations according to a preset first mapping function.

[0022] The matrix denoising module is used to iteratively denoise the initial data matrix using a trained denoising model according to the number of denoising iterations to obtain a target data matrix.

[0023] The fault identification module is used to input the target data matrix into the trained fault identification model to obtain a fault identification probability vector.

[0024] The fault warning module is used to generate power grid warning information when the fault identification probability vector meets a preset condition.

[0025] The present invention has at least the following beneficial effects: by evaluating the reconstruction differences of the initial data matrix formed by monitoring data vectors, the noise complexity of the initial data matrix is ​​characterized, thereby determining the number of denoising iterations, and then iterative denoising is performed by the denoising model, so that the expected denoising effect can be achieved with as few denoising iterations as possible, thereby reducing the time consumption of the denoising process while ensuring the reliability of the target data matrix, that is, while ensuring the accuracy of power grid monitoring. In addition, fault identification is performed in matrix form, and power grid early warning information is determined based on the obtained fault identification probability vector. Compared with the prior art method of iterative prediction using a time series prediction model, there is no need to introduce the time consumption of iterative prediction, which effectively improves the timeliness of power grid monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0027] Figure 1 A flowchart of a method for monitoring a power transmission and transformation network based on the Internet of Things provided in the first embodiment of the present invention;

[0028] Figure 2 A schematic diagram of a transmission and transformation grid monitoring system based on the Internet of Things provided in Example 2 of the present invention. DETAILED DESCRIPTION

[0029] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0030] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It is understood that, where appropriate, the above-mentioned terms used to distinguish similar objects can be interchanged so that the present invention can also implement other embodiments other than the above-mentioned illustrated embodiments or described embodiments. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0031] Example 1

[0032] This embodiment provides a method for monitoring a power transmission and transformation network based on the Internet of Things. Figure 1 FIG. 1 is a flowchart of a method for monitoring a power transmission and transformation network based on the Internet of Things according to a first embodiment of the present invention. The method for monitoring a power transmission and transformation network based on the Internet of Things includes the following steps:

[0033] S101, forming an initial data matrix from monitoring data vectors acquired at M preset time points, where M is a positive integer;

[0034] S102, performing mask processing on the initial data matrix to obtain a masked data matrix;

[0035] S103, inputting the mask data matrix into the trained reconstruction model to obtain a reconstructed data matrix;

[0036] S104, determining a reconstruction difference evaluation value according to the initial data matrix and the reconstructed data matrix;

[0037] S105, mapping the reconstruction difference evaluation value to the number of denoising iterations according to a preset first mapping function;

[0038] S106, performing iterative denoising on the initial data matrix using the trained denoising model according to the number of denoising iterations to obtain a target data matrix;

[0039] S107, inputting the target data matrix into the trained fault recognition model to obtain a fault recognition probability vector;

[0040] S108 : When the fault identification probability vector meets a preset condition, power grid early warning information is generated.

[0041] Among them, the preset time point can refer to the time point for collecting the status of the power grid equipment, and the time interval between adjacent preset time points can be a fixed value, which can be set by the implementer, but it must be greater than or equal to the time taken from obtaining the monitoring data vector at the Mth preset time point to determining whether the fault identification probability vector meets the preset conditions.

[0042] The monitoring data vector can be obtained by splicing the monitoring data collected by each sensor, and the initial data matrix can be obtained by splicing M monitoring data vectors.

[0043] The reconstruction difference evaluation value can be used to characterize the noise complexity of the initial data matrix. The prior condition of this embodiment is that the higher the noise complexity, the more difficult it is to accurately reconstruct the masked data matrix after the masking of the initial data matrix through the reconstruction model.

[0044] Specifically, the denoising model can be implemented using a diffusion generation model. The architecture of the diffusion generation model can include only a denoising neural network during the use phase. The architecture of the denoising neural network can adopt a U-Net architecture. The training process of the diffusion generation model will not be described in detail here. It should be noted that, in the prior art, the diffusion generation model usually requires multiple iterations to recover the denoised data from the initial data matrix that may contain noise of various types of noise. Obviously, multiple iterations will introduce higher denoising time. This embodiment obtains the number of denoising iterations based on the reconstruction difference evaluation value mapping, aiming to limit the number of denoising iterations from being too high, so that the number of denoising iterations can be reduced as much as possible while satisfying the denoising effect, thereby improving the timeliness of the overall power grid monitoring.

[0045] If the number of denoising iterations obtained still cannot meet the time requirement, the implementer can perform knowledge distillation based on the trained diffusion generation model to reduce the number of denoising iterations. If the existing technology usually uses P iterative denoising when inputting the initial sample matrix into the trained diffusion generation model for denoising, then the result of the (R-1)×P / Q iterative denoising is used as the sample, and the result of the R×P / Q iterative denoising is used as the label to train a new diffusion generation model. P, Q and R are all positive integers, and the diffusion generation model is updated with the new diffusion generation model. In particular, when R=1, the initial sample matrix is ​​used as the sample. In this way, the number of denoising iterations can be effectively reduced, but a certain amount of denoising accuracy will be lost accordingly. Therefore, knowledge distillation is performed to update the diffusion generation model only when the number of denoising iterations obtained by reconstructing the difference evaluation value mapping cannot meet the time requirement.

[0046] In a specific embodiment, the monitoring data vector includes monitoring values ​​corresponding to N monitoring data types, and the monitoring data type belongs to a monitoring data type set, and the monitoring data type set includes at least current, voltage, frequency, vibration and temperature.

[0047] The monitoring data type set may also include power for power grid equipment monitoring, and humidity, wind speed, air pressure, air quality, etc. for environmental monitoring.

[0048] In a specific embodiment, the monitoring data vector is a column vector;

[0049] The masking process is performed on the initial data matrix to obtain a masked data matrix, comprising:

[0050] According to the preset sampling probabilities corresponding to the M columns in the initial data matrix, the M columns are randomly selected with weights to obtain reference columns;

[0051] The column vector corresponding to the reference column in the initial data matrix is ​​set to a first preset value to obtain the mask data matrix.

[0052] The monitoring data vector is a column vector of size N×1, and each column of the monitoring data vector corresponds to a different monitoring data type.

[0053] The preset sampling probability may refer to a weight when randomly selecting a corresponding column, and the first preset value may be 0.

[0054] In a specific embodiment, the step of obtaining the preset sampling probabilities corresponding to the M columns in the initial data matrix includes:

[0055] For any column in the initial data matrix, mapping the column number corresponding to the column to a reference value corresponding to the column according to a preset second mapping function;

[0056] Normalization is performed on the reference values ​​corresponding to the M columns to obtain the preset sampling probabilities corresponding to the M columns.

[0057] The second mapping function can be w m =m / M, where m is the number of columns, w m is the reference value of the mth column.

[0058] Specifically, the normalized processing result of the reference value of the mth column, that is, the preset sampling probability can be W m =w m / (∑Me=1(w e )).

[0059] In a specific embodiment, the trained reconstruction model includes a trained first encoder and a trained first decoder;

[0060] Inputting the mask data matrix into the trained reconstruction model to obtain the reconstructed data matrix comprises:

[0061] Inputting the mask data matrix into the trained first encoder to perform feature extraction to obtain a first feature vector;

[0062] The first feature vector is input into the trained first decoder to perform feature reconstruction to obtain the reconstructed data matrix.

[0063] Among them, the reconstruction model can also adopt the U-Net architecture. The training samples of the reconstruction model can be the reference data matrix formed by the reference monitoring data vector under theoretical noise-free conditions. The training labels of the reconstruction model are the reference data matrix itself. The training loss of the reconstruction model can be calculated using the mean square error loss function.

[0064] It should be noted that if the reference monitoring data vector under theoretical noise-free conditions is difficult to obtain, the implementer can also input the initial sample matrix into the trained diffusion generation model and perform iterative denoising to obtain the result as the reference data matrix.

[0065] In a specific embodiment, determining the reconstruction difference evaluation value according to the initial data matrix and the reconstructed data matrix includes:

[0066] Calculating the Euclidean distance between the initial data matrix and the reconstructed data matrix to obtain a difference distance;

[0067] The difference distance is normalized according to a second preset value to obtain the reconstruction difference evaluation value.

[0068] The second preset value may be obtained by performing Euclidean distance calculation based on the initial data matrix and the zero matrix, and the normalization process may refer to comparing the difference distance with the second preset value.

[0069] Specifically, when calculating the Euclidean distance between the initial data matrix and the reconstructed data matrix, the implementer may only perform Euclidean distance calculation on the column vectors corresponding to the reference column in the initial data matrix and the reconstructed data matrix respectively. Accordingly, the second preset value may be obtained by performing Euclidean distance calculation based on the column vector corresponding to the reference column in the initial data matrix and the zero vector.

[0070] In a specific embodiment, mapping the reconstruction difference evaluation value to the number of denoising iterations according to a preset first mapping function includes:

[0071] mapping the reconstruction difference evaluation value to an intermediate value according to the first mapping function;

[0072] The intermediate value is rounded up to obtain the number of denoising iterations.

[0073] The first mapping function may be y=tx, where t is the maximum number of denoising iterations, x is the reconstruction difference evaluation value, and y is the intermediate value.

[0074] In a specific embodiment, the iterative denoising of the initial data matrix using a trained denoising model according to the number of denoising iterations to obtain a target data matrix includes:

[0075] Initialize the temporary iteration number K=1, and use the initial data matrix as the temporary data matrix;

[0076] Inputting the temporary data matrix into the trained denoising model to obtain a denoised data matrix;

[0077] If K is less than the number of denoising iterations, the denoised data matrix is ​​used as the temporary data matrix, K=K+1 is updated, and the process returns to the step of inputting the temporary data matrix into the trained denoising model;

[0078] If K is equal to the number of denoising iterations, the denoised data matrix finally output by the trained denoising model is used as the target data matrix.

[0079] In a specific embodiment, the fault identification probability vector corresponds to the Mth preset time point;

[0080] When the fault identification probability vector meets a preset condition, generating power grid early warning information includes:

[0081] Determine I probability change vectors based on the fault identification probability vector and the historical probability vectors corresponding to the M-th preset time point to the M-th preset time point, wherein the probability change vector includes probability change values ​​corresponding to J fault type dimensions, where I and J are both positive integers;

[0082] If the probability change values ​​corresponding to any fault type dimension in the I probability change vectors are all greater than the third preset value, it is determined that the fault identification probability vector meets the preset condition, and the power grid warning information is generated according to the fault type dimension.

[0083] Among them, the historical probability vectors corresponding to the Mi-i+1th preset time point and the Mi-th preset time point are subtracted to obtain the probability change vector corresponding to the Mi-i+1th preset time point, i is an integer in the range of [2, I], the third preset value can be set to 0, and the implementer can adjust the third preset value according to actual conditions.

[0084] Specifically, the fault recognition model can include a second encoder and a fully connected layer. The initial parameters of the second encoder can be the same as the parameters of the trained first encoder. Since the trained first encoder is trained based on the reconstruction task of the data matrix, it can be considered that the trained first encoder can learn the partial correlation between different monitoring data types. Making the initial parameters of the second encoder the same as the parameters of the trained first encoder can improve the feature extraction capability of the second encoder, thereby improving the recognition accuracy of the trained fault recognition model compared to training based on randomly initialized second encoder parameters. The fault recognition model can be trained using the training method of a conventional multi-classification model.

[0085] After the monitoring data vector is collected at each preset time point, the fault identification probability vector corresponding to the preset time point is processed. If the fault identification probability vector of a preset time point does not meet the preset conditions, the fault identification probability vector of the preset time point is stored as a historical probability vector.

[0086] It can be seen that compared with the prior art method of using a time series prediction model for fault prediction, this embodiment avoids the time series iterative prediction process, thereby reducing the processing time of power grid monitoring data and improving the timeliness of power grid monitoring. Moreover, when predicting faults, a fault identification probability vector is obtained based on the actual collected monitoring data vector, and then a probability change trend analysis is performed based on the fault identification probability vector and the historical probability vector to perform fault prediction, thereby avoiding the introduction of errors in time series prediction and improving the accuracy of fault prediction. Moreover, the historical probability vector can be stored in advance, further improving the timeliness of power grid monitoring.

[0087] In the first embodiment of the present invention, the reconstruction difference evaluation of the initial data matrix formed by the monitoring data vector is used to characterize the noise complexity of the initial data matrix, thereby determining the number of denoising iterations. Then, the denoising model is used for iterative denoising, and the expected denoising effect can be achieved with as few denoising iterations as possible. This reduces the time consumption of the denoising process while ensuring the reliability of the target data matrix, that is, while ensuring the accuracy of power grid monitoring. In addition, fault identification is performed in matrix form, and power grid early warning information is determined based on the obtained fault identification probability vector. Compared with the prior art method of iterative prediction using a time series prediction model, this method does not require the introduction of time-consuming iterative prediction, thereby effectively improving the timeliness of power grid monitoring.

[0088] Example 2

[0089] This embodiment 2 provides a transmission and transformation grid monitoring system based on the Internet of Things, such as Figure 2 FIG. 1 is a schematic diagram of a transmission and transformation network monitoring system based on the Internet of Things provided in a second embodiment of the present invention. The transmission and transformation network monitoring system based on the Internet of Things includes:

[0090] A matrix forming module 201 is configured to form an initial data matrix from monitoring data vectors acquired at M preset time points, where M is a positive integer;

[0091] A matrix masking module 202 is configured to perform masking on the initial data matrix to obtain a masked data matrix;

[0092] A matrix reconstruction module 203 is used to input the mask data matrix into a trained reconstruction model to obtain a reconstructed data matrix;

[0093] A difference evaluation module 204 is configured to determine a reconstruction difference evaluation value based on the initial data matrix and the reconstructed data matrix;

[0094] A times mapping module 205 is configured to map the reconstruction difference evaluation value to a denoising iteration number according to a preset first mapping function;

[0095] A matrix denoising module 206 is configured to iteratively denoise the initial data matrix using a trained denoising model according to the number of denoising iterations to obtain a target data matrix;

[0096] The fault identification module 207 is used to input the target data matrix into the trained fault identification model to obtain a fault identification probability vector;

[0097] The fault warning module 208 is configured to generate power grid warning information when the fault identification probability vector meets a preset condition.

[0098] It should be noted that the specific limitations of the IoT-based power grid monitoring system can be found in the aforementioned limitations of the IoT-based power grid monitoring method, and will not be further elaborated here. The information interaction and execution process between the aforementioned modules, as well as other aspects, are based on the same concept as the method embodiments of the present invention. Their specific functions and technical effects can be found in the method embodiments and will not be further elaborated here.

[0099] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any form. Although the present invention has been disclosed as above in terms of preferred embodiments, they are not intended to limit the present invention. Any technician familiar with this profession can make some changes or modifications to equivalent embodiments of equivalent changes using the technical contents disclosed above without departing from the scope of the technical solution of the present invention. However, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A method for monitoring a power transmission and transformation network based on the Internet of Things, characterized in that: The Internet of Things-based transmission and transformation grid monitoring method comprises the following steps: S101, forming an initial data matrix from monitoring data vectors acquired at M preset time points, where M is a positive integer; S102, performing mask processing on the initial data matrix to obtain a masked data matrix; S103, inputting the mask data matrix into the trained reconstruction model to obtain a reconstructed data matrix; S104, determining a reconstruction difference evaluation value according to the initial data matrix and the reconstructed data matrix; S105, mapping the reconstruction difference evaluation value to the number of denoising iterations according to a preset first mapping function; S106, performing iterative denoising on the initial data matrix using the trained denoising model according to the number of denoising iterations to obtain a target data matrix; S107, inputting the target data matrix into the trained fault recognition model to obtain a fault recognition probability vector; S108 : When the fault identification probability vector meets a preset condition, power grid early warning information is generated.

2. The method for monitoring a power transmission and transformation network based on the Internet of Things according to claim 1, characterized in that: The monitoring data vector includes monitoring values ​​corresponding to N monitoring data types, respectively. The monitoring data types belong to a monitoring data type set, and the monitoring data type set includes at least current, voltage, frequency, vibration, and temperature.

3. The method for monitoring a power transmission and transformation network based on the Internet of Things according to claim 1, wherein: The monitoring data vector is a column vector; The masking process is performed on the initial data matrix to obtain a masked data matrix, comprising: According to the preset sampling probabilities corresponding to the M columns in the initial data matrix, the M columns are randomly selected with weights to obtain reference columns; The column vector corresponding to the reference column in the initial data matrix is ​​set to a first preset value to obtain the mask data matrix.

4. The method for monitoring a power transmission and transformation network based on the Internet of Things according to claim 3, characterized in that: The step of obtaining the preset sampling probabilities corresponding to the M columns in the initial data matrix includes: For any column in the initial data matrix, mapping the column number corresponding to the column to a reference value corresponding to the column according to a preset second mapping function; Normalization is performed on the reference values ​​corresponding to the M columns to obtain the preset sampling probabilities corresponding to the M columns.

5. The method for monitoring a power transmission and transformation network based on the Internet of Things according to claim 1, wherein: The trained reconstruction model includes a trained first encoder and a trained first decoder; Inputting the mask data matrix into the trained reconstruction model to obtain the reconstructed data matrix comprises: Inputting the mask data matrix into the trained first encoder to perform feature extraction to obtain a first feature vector; The first feature vector is input into the trained first decoder to perform feature reconstruction to obtain the reconstructed data matrix.

6. The method for monitoring a power transmission and transformation network based on the Internet of Things according to claim 1, wherein: The step of determining a reconstruction difference evaluation value based on the initial data matrix and the reconstructed data matrix includes: Calculating the Euclidean distance between the initial data matrix and the reconstructed data matrix to obtain a difference distance; The difference distance is normalized according to a second preset value to obtain the reconstruction difference evaluation value.

7. The method for monitoring a power transmission and transformation network based on the Internet of Things according to claim 1, characterized in that: Mapping the reconstruction difference evaluation value to the number of denoising iterations according to a preset first mapping function includes: mapping the reconstruction difference evaluation value to an intermediate value according to the first mapping function; The intermediate value is rounded up to obtain the number of denoising iterations.

8. The method for monitoring a power transmission and transformation network based on the Internet of Things according to claim 1, wherein: The iterative denoising of the initial data matrix is ​​performed using a trained denoising model according to the number of denoising iterations to obtain a target data matrix, including: Initialize the temporary iteration number K=1, and use the initial data matrix as the temporary data matrix; Inputting the temporary data matrix into the trained denoising model to obtain a denoised data matrix; If K is less than the number of denoising iterations, the denoised data matrix is ​​used as the temporary data matrix, K=K+1 is updated, and the process returns to the step of inputting the temporary data matrix into the trained denoising model; If K is equal to the number of denoising iterations, the denoised data matrix finally output by the trained denoising model is used as the target data matrix.

9. The method for monitoring a power transmission and transformation network based on the Internet of Things according to claim 1, wherein: The fault identification probability vector corresponds to the Mth preset time point; When the fault identification probability vector meets a preset condition, generating power grid early warning information includes: Determine I probability change vectors based on the fault identification probability vector and the historical probability vectors corresponding to the M-th preset time point to the M-th preset time point, wherein the probability change vector includes probability change values ​​corresponding to J fault type dimensions, where I and J are both positive integers; If the probability change values ​​corresponding to any fault type dimension in the I probability change vectors are all greater than the third preset value, it is determined that the fault identification probability vector meets the preset condition, and the power grid warning information is generated according to the fault type dimension.

10. A power transmission and transformation network monitoring system based on the Internet of Things, characterized in that: The Internet of Things-based transmission and transformation grid monitoring system includes: A matrix forming module, configured to form an initial data matrix from monitoring data vectors acquired at M preset time points, where M is a positive integer; a matrix masking module, configured to perform masking processing on the initial data matrix to obtain a masked data matrix; A matrix reconstruction module, used for inputting the mask data matrix into a trained reconstruction model to obtain a reconstructed data matrix; a difference evaluation module, configured to determine a reconstruction difference evaluation value based on the initial data matrix and the reconstructed data matrix; A number mapping module, configured to map the reconstruction difference evaluation value to a denoising iteration number according to a preset first mapping function; A matrix denoising module, configured to iteratively denoise the initial data matrix using a trained denoising model according to the number of denoising iterations to obtain a target data matrix; A fault identification module is used to input the target data matrix into a trained fault identification model to obtain a fault identification probability vector; The fault warning module is used to generate power grid warning information when the fault identification probability vector meets a preset condition.

Images