Monitoring and early warning method and device for hidden danger related to electricity of electric power system, terminal equipment and storage medium

Abstract

The invention discloses a monitoring and early warning method and device for hidden dangers related to electricity of an electric power system, terminal equipment and a storage medium, and belongs to the technical field of electric power systems.The method comprises the steps that initial electric power inspection data and initial electric power inspection verification data of a to-be-detected electricity utilization area in a plurality of modes of the electric power system in the same preset time period are obtained; the early warning judgment operation is repeatedly executed until errors between the current electric power inspection data and the current electric power inspection verification data are smaller than corresponding preset error threshold values, and the current weight serves as a target weight; according to the current electric power inspection data and the target weight, a target early warning judgment value is obtained through calculation, and when the target early warning judgment value is not smaller than a preset early warning judgment threshold value, electric related hidden danger early warning is carried out on the to-be-detected electricity utilization area; and otherwise, not carrying out the electricity-related hidden danger early warning. According to the invention, the method can solve a problem that the early warning result is not accurate because the obtained power data is directly used for early warning in the prior art.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a method, device, terminal equipment and storage medium for monitoring and early warning of potential electrical hazards in power systems. Background Technology

[0002] A power system is a system for the production and consumption of electrical energy, consisting of power plants, transmission and transformation lines, substations, and consumers. Its function is to convert primary energy sources from nature into electrical energy through power generation devices, and then supply this energy to users through transmission, transformation, and distribution. Therefore, for the safe use of the power system, it is necessary to investigate potential electrical hazards. These hazards refer to various safety risks related to electricity, which may exist in multiple stages of power production, transmission, and use. If not addressed promptly, they may lead to accidents such as electric shock, fire, equipment damage, or even personal injury.

[0003] In existing technologies, power-related hazard warnings are directly based on the acquired power data. However, in reality, the acquired power data may exhibit abnormal fluctuations that differ from the actual data, leading to inaccurate warning results. Summary of the Invention

[0004] This invention provides a method, device, terminal equipment, and storage medium for monitoring and early warning of potential electrical hazards in power systems, which can solve the problem of inaccurate early warning results caused by directly using acquired power data for early warning in the prior art.

[0005] An embodiment of the present invention provides a method for monitoring and early warning of potential electrical hazards in a power system, comprising:

[0006] Acquire initial power inspection data of the power consumption area to be tested under several modes of the power system within the same preset time period, as well as several initial power inspection verification data for data verification.

[0007] Repeat the early warning judgment operation until the current error between the current power inspection data and the corresponding current power inspection verification data in each mode is less than the corresponding preset error threshold, and then use the current weight as the target weight.

[0008] Based on the current power inspection data and target weights under each mode, the target early warning judgment value is calculated. If the target early warning judgment value is not less than the preset early warning judgment threshold, an early warning for potential electrical hazards is issued for the power consumption area to be inspected; otherwise, no early warning for potential electrical hazards is issued for the power consumption area to be inspected.

[0009] The aforementioned early warning judgment operation is as follows:

[0010] Obtain the current weight, the current power inspection data under each mode, and the current power inspection verification data; wherein, the initial weight is the preset initial weight, the initial power inspection data is the aforementioned initial power inspection data, and the initial power inspection verification data is the aforementioned initial power inspection verification data.

[0011] The current warning judgment value is calculated based on the current weight and the current power inspection data. If the current warning judgment value is not less than the above-mentioned preset warning judgment threshold, the current error between the current power inspection data and the current power inspection verification data under each mode is calculated.

[0012] If any current error exceeds the corresponding preset error threshold, the current weight corresponding to the current error exceeding the preset error threshold is updated, and the power inspection data and power inspection verification data for the next preset time period are obtained.

[0013] Furthermore, the current early warning judgment value calculated based on the current weights and current power inspection data includes:

[0014] By using a preset feature extraction model, features are extracted from the current power inspection data under each mode to obtain the dimension values ​​of the power inspection data under each mode.

[0015] The current warning judgment value is calculated based on the values ​​of each dimension and their corresponding current weights.

[0016] Furthermore, the training of the aforementioned preset feature extraction model includes:

[0017] Obtain several historical power inspection data with real labels; wherein, the real labels are used to represent the real dimension values ​​of each historical power inspection data.

[0018] The above-mentioned historical power inspection data is input into the feature extraction model to be trained for iterative training until the loss function converges, and the trained preset historical power inspection data is obtained.

[0019] In each iteration of training, the current prediction dimension value is obtained based on the current historical power inspection data; the current loss function is calculated based on the current prediction dimension value and the corresponding true label; and it is determined whether the current loss function has converged. If it has converged, the current feature extraction model is used as the above-mentioned preset feature extraction model; otherwise, the model parameters in the current feature extraction model are adjusted, and training continues.

[0020] Furthermore, before obtaining the dimensional values ​​of the power inspection data for each mode, the process also includes: data cleaning of the current power inspection data for each mode.

[0021] Furthermore, it also includes:

[0022] The current warning judgment value is calculated based on the current weight and the current power inspection data. If the current warning judgment value is less than the above-mentioned preset warning judgment threshold, no power-related hazard warning will be issued for the power consumption area to be inspected.

[0023] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments;

[0024] This invention provides a monitoring and early warning device for potential electrical hazards in a power system, comprising:

[0025] The module includes a data acquisition module, a target weight determination module, and a monitoring and early warning module.

[0026] The aforementioned data acquisition module is used to acquire initial power inspection data of the power system under several modes within the same preset time period of the power consumption area to be tested, as well as several initial power inspection verification data for data verification.

[0027] The aforementioned target weight determination module is used to repeatedly execute the early warning judgment operation until the current error between the current power inspection data and the corresponding current power inspection verification data in each mode is less than the corresponding preset error threshold, and the current weight is used as the target weight; wherein, the aforementioned early warning judgment operation is as follows:

[0028] Obtain the current weight, the current power inspection data under each mode, and the current power inspection verification data; wherein, the initial weight is the preset initial weight, the initial power inspection data is the aforementioned initial power inspection data, and the initial power inspection verification data is the aforementioned initial power inspection verification data.

[0029] The current warning judgment value is calculated based on the current weight and the current power inspection data. If the current warning judgment value is not less than the above-mentioned preset warning judgment threshold, the current error between the current power inspection data and the current power inspection verification data under each mode is calculated.

[0030] If any current error is greater than the corresponding preset error threshold, the current weight corresponding to the current error that is greater than the preset error threshold is updated, and the power inspection data and power inspection verification data in the next preset time period are obtained.

[0031] The aforementioned monitoring and early warning module is used to calculate the target early warning judgment value based on the current power inspection data and target weight under each mode, and to issue an early warning for potential electrical hazards in the power consumption area to be inspected if the target early warning judgment value is not less than the preset early warning judgment threshold; otherwise, it will not issue an early warning for potential electrical hazards in the power consumption area to be inspected.

[0032] Furthermore, the aforementioned target weight determination module includes:

[0033] Feature extraction unit and judgment value calculation unit;

[0034] The aforementioned feature extraction unit is used to extract features from the current power inspection data under each modality using a preset feature extraction model, thereby obtaining the dimension values ​​of the power inspection data under each modality.

[0035] The aforementioned judgment value calculation unit is used to calculate the current warning judgment value based on the values ​​of each dimension and the corresponding current weights.

[0036] Furthermore, the aforementioned feature extraction unit includes: a model training subunit;

[0037] The aforementioned model training subunit is used to acquire several historical power grid inspection data with real labels; wherein, the real labels are used to represent the real dimension values ​​of each historical power grid inspection data; the aforementioned historical power grid inspection data is input into the feature extraction model to be trained for iterative training until the loss function converges, resulting in trained preset historical power grid inspection data; wherein, in each iteration of training, the current predicted dimension value is obtained based on the current historical power grid inspection data; the current loss function is calculated based on the current predicted dimension value and the corresponding real label; and it is determined whether the current loss function has converged; if it has converged, the current feature extraction model is used as the aforementioned preset feature extraction model; otherwise, the model parameters in the current feature extraction model are adjusted, and training continues.

[0038] Based on the above method embodiments, the present invention provides a corresponding terminal device embodiment;

[0039] The present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the power system electrical hazard monitoring and early warning method described in any embodiment of the present invention.

[0040] Based on the above method embodiments, the present invention provides a corresponding storage medium embodiment;

[0041] The present invention provides a storage medium including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the monitoring and early warning method for electrical hazards in a power system as described in any embodiment of the present invention.

[0042] The embodiments of the present invention have the following beneficial effects:

[0043] This invention provides a method, device, terminal equipment, and storage medium for monitoring and early warning of potential electrical hazards in a power system. The method includes: acquiring initial power inspection data and initial power inspection verification data for data verification under several modes of the power system in the power consumption area to be monitored within the same preset time period; repeatedly performing early warning judgment operations until the current error between the current power inspection data and the corresponding current power inspection verification data under each mode is less than the corresponding preset error threshold, and using the current weight as the target weight; calculating the target early warning judgment value based on the current power inspection data and the target weight, and issuing an early warning for potential electrical hazards to the power consumption area to be monitored if the target early warning judgment value is not less than the preset early warning judgment threshold; otherwise, not issuing an early warning for potential electrical hazards to the power consumption area to be monitored. The aforementioned early warning judgment operation is as follows: Obtain the current weight, current power inspection data for each mode, and current power inspection verification data; wherein, the initial weight is a preset initial weight, the initial power inspection data is the aforementioned initial power inspection data, and the initial power inspection verification data is the aforementioned initial power inspection verification data; calculate the current early warning judgment value based on the current weight and the current power inspection data; if the current early warning judgment value is not less than the aforementioned preset early warning judgment threshold, calculate the current error between the current power inspection data and the current power inspection verification data for each mode; if any current error is greater than the corresponding preset error threshold, update the current weight corresponding to the current error greater than the preset error threshold, and obtain the power inspection data and power inspection verification data for the next preset time period. Therefore, in this invention, when a warning judgment value is obtained based on the acquired power inspection data, the error between the power inspection verification data and the power inspection data is calculated to determine whether the current warning judgment value is correct. If the error is large, it indicates that at least one power inspection data has abnormal fluctuations. Therefore, by adjusting and updating the corresponding weights, the sensitivity of the power inspection data to this item can be adjusted when calculating the warning judgment value, thereby reducing the impact of abnormal fluctuations on the warning judgment result and obtaining a more accurate warning result. Attached Figure Description

[0044] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0045] Figure 1 This is a flowchart illustrating a method for monitoring and early warning of potential electrical hazards in a power system, provided by an embodiment of the present invention.

[0046] Figure 2This is a schematic diagram of the structure of a monitoring and early warning device for potential electrical hazards in a power system, provided in an embodiment of the present invention. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0048] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0049] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0050] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0051] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0052] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0053] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0054] See Figure 1 To address the problem of inaccurate early warning results caused by directly using acquired power data in existing technologies, an embodiment of the present invention provides a method for monitoring and early warning of potential electrical hazards in power systems, comprising:

[0055] Step S101: Obtain initial power inspection data of the power consumption area to be tested under several modes of the power system within the same preset time period, as well as several initial power inspection verification data for data verification.

[0056] Specifically, power inspection data includes several multimodal data types, namely: image modal data, such as images of the appearance of power equipment, such as transformers, lines, and distribution boxes, collected by surveillance cameras, used to identify physical anomalies, such as line breaks, foreign object connections, and equipment corrosion; sound modal data, such as the operating sounds of equipment collected by microphones, such as the low-frequency hum of normal operation, abnormal spark discharge sounds, and abnormal noises from equipment; and sensor modal data, such as temperature data, current data, voltage data, and humidity data, collected by set temperature sensors, current sensors, voltage sensors, and humidity sensors to reflect the operating status of equipment, such as a sudden increase in current due to overload or a sudden rise in temperature before a short circuit.

[0057] Specifically, the equipment used to acquire power inspection verification data is set up near the equipment used to acquire power inspection data. During data acquisition, the two sets of equipment collect relevant power data of the same object at the same time for subsequent verification operations.

[0058] Step S102: Repeat the early warning judgment operation until the current error between the current power inspection data and the corresponding current power inspection verification data in each mode is less than the corresponding preset error threshold, and use the current weight as the target weight.

[0059] The aforementioned early warning judgment operation is as follows:

[0060] Obtain the current weight, the current power inspection data under each mode, and the current power inspection verification data; wherein, the initial weight is the preset initial weight, the initial power inspection data is the aforementioned initial power inspection data, and the initial power inspection verification data is the aforementioned initial power inspection verification data.

[0061] The current warning judgment value is calculated based on the current weight and the current power inspection data. If the current warning judgment value is not less than the above-mentioned preset warning judgment threshold, the current error between the current power inspection data and the current power inspection verification data under each mode is calculated.

[0062] If any current error exceeds the corresponding preset error threshold, the current weight corresponding to the current error exceeding the preset error threshold is updated, and the power inspection data and power inspection verification data for the next preset time period are obtained.

[0063] Specifically, in this invention, power inspection data has multiple modalities. For image modal data and sound module data, the feature vector of the current image (or sound) corresponding to the current power inspection data and the feature vector of the image (or sound) corresponding to the current power inspection verification data in the same dimension can be extracted. The Euclidean distance between the two images (or sounds) is calculated based on the feature vectors between the two images (or sounds) to obtain the similarity between the two images (or sounds). The current error between the current power inspection data and the current power inspection verification data is measured based on the above similarity. For example, the similarity corresponding to the preset error threshold can be set to 80%. When the similarity between the two images (or sounds) is greater than 80%, it can be determined that the current error is greater than the corresponding preset error threshold.

[0064] For sensor modal data, the difference between the sensor value corresponding to the current power inspection data and the sensor value corresponding to the current power inspection verification data is directly calculated to obtain the current error.

[0065] Specifically, if any current error exceeds the corresponding preset error threshold, it indicates that the power inspection data corresponding to this current error shows abnormal fluctuations, and the existing weights are too sensitive to it, leading to erroneous warning results. Therefore, the corresponding weights need to be reduced. To illustrate, taking a three-stage weight adjustment process as an example, in the first weight adjustment, the initial weight is subtracted by a multiple of a first preset percentage (e.g., 10%) from the initial weight to obtain the first updated weight. Then, the updated weight is combined with the unupdated weight (i.e., the weight corresponding to the current error not exceeding the corresponding preset error threshold) and used as the weight for the next adjustment.

[0066] Specifically, if the second judgment finds that the weight still needs to be adjusted, the value of the initial weight is obtained by subtracting a second preset percentage (e.g., 20%) from the initial weight. The updated weight is then combined with the unupdated weight (i.e., the weight corresponding to the current error that is not greater than the corresponding preset error threshold) and used as the weight for the next time.

[0067] Specifically, if the third judgment finds that the weight still needs adjustment, the initial weight is reduced by a third preset percentage (e.g., 30%) to obtain the updated weight. This updated weight is then combined with the unupdated weight (i.e., the weight corresponding to the current error that is not greater than the corresponding preset error threshold) and used as the weight for the next adjustment. Generally, after three adjustments, all current errors will be no greater than the corresponding preset error threshold.

[0068] To illustrate, for the first adjustment, taking the weights corresponding to the dimension values ​​of the image modality data as an example, the formula for calculating the weights after the first update is:

[0069] α1 = α - α * 10%

[0070] In the formula, α1 represents the updated weight corresponding to the dimension value of the image modality data after the first adjustment, and α represents the initial weight corresponding to the dimension value of the image modality data.

[0071] For the second adjustment, still using image modal data as an example, the updated weight calculation formula is as follows:

[0072] α2=α-α*20%

[0073] In the formula, α2 represents the updated weights corresponding to the dimension values ​​of the image modality data after the second adjustment. For the third adjustment, still using image modality data as an example, the formula for calculating the updated weights is:

[0074] α3 = α - α * 30%

[0075] In the formula, α3 represents the updated weights corresponding to the dimension values ​​of the image modal data after the third adjustment.

[0076] For illustration, if α is 0.35, then after the first adjustment, α1 will be 0.315. The next power inspection data is then used to calculate the warning judgment value. If the warning judgment value is not less than the preset warning judgment threshold, and the current error corresponding to α is still greater than the corresponding preset error threshold, the second weight adjustment is performed according to the above process to obtain a new weight α2. This process is repeated until the target weights for all dimensions are obtained. For illustration, if the weight adjusted in the first adjustment is α, and the second adjustment requires modifying the weight corresponding to another dimension value, then for the weight that needs adjustment, which is in the first adjustment stage, the adjustment will be performed at the aforementioned 10% margin.

[0077] In a preferred embodiment, the calculation of the current early warning judgment value based on the current weight and the current power inspection data includes:

[0078] By using a preset feature extraction model, features are extracted from the current power inspection data under each mode to obtain the dimension values ​​of the power inspection data under each mode.

[0079] Specifically, the preset feature extraction model is an unsupervised autoencoder neural network model. Power inspection data is input into the preset feature extraction model to extract the features of each power inspection data, and after standardization processing, the aforementioned dimension values ​​can be obtained.

[0080] The current warning judgment value is calculated based on the values ​​of each dimension and their corresponding current weights.

[0081] Specifically, the warning judgment value can be calculated using the following formula:

[0082] X = α*A + β*B + γ*C

[0083] In the formula, X represents the warning judgment value, A represents the dimension value of the image modal data in the power inspection data, α represents the weight corresponding to the image modal data, B represents the dimension value of the sound modal data in the power inspection data, β represents the weight corresponding to the sound modal data, C represents the dimension value of the sensor modal data in the power inspection data, and γ represents the weight corresponding to the sensor modal data.

[0084] Preferably, initially, the weights of the above items are the initial weights, and the calculated warning judgment value is the initial warning judgment value. The initial weight corresponding to the dimension value of the image modal data can be set to 0.35, the initial weight corresponding to the sound modal data can be set to 0.3, and the initial weight corresponding to the sensor modal data can be set to 0.25.

[0085] Preferably, by linking and integrating various power inspection data with different weights, a multi-dimensional monitoring system is formed, thereby reducing the early warning error rate and avoiding the problem of low efficiency of manual inspection.

[0086] In this preferred embodiment, feature extraction is performed on power inspection data to obtain dimension values, and then weighted calculation is performed based on the weights corresponding to each dimension value to obtain the early warning judgment value.

[0087] In another preferred embodiment, the training of the aforementioned preset feature extraction model includes:

[0088] Obtain several historical power inspection data with real labels; wherein, the real labels are used to represent the real dimension values ​​of each historical power inspection data.

[0089] The above-mentioned historical power inspection data is input into the feature extraction model to be trained for iterative training until the loss function converges, and the trained preset historical power inspection data is obtained.

[0090] In each iteration of training, the current prediction dimension value is obtained based on the current historical power inspection data; the current loss function is calculated based on the current prediction dimension value and the corresponding true label; and it is determined whether the current loss function has converged. If it has converged, the current feature extraction model is used as the above-mentioned preset feature extraction model; otherwise, the model parameters in the current feature extraction model are adjusted, and training continues.

[0091] Specifically, by determining whether the loss function obtained in each iteration of training converges, the model parameters of the preset feature extraction model are determined, and then the final preset feature extraction model is obtained after training is completed.

[0092] In this preferred embodiment, the feature extraction model is trained using historical power inspection data to obtain a pre-trained preset feature extraction model.

[0093] In another preferred embodiment, before obtaining the dimension values ​​of the power inspection data for each mode, the method further includes: performing data cleaning on the current power inspection data for each mode.

[0094] Specifically, before extracting the dimensional values ​​of power inspection data, it is preprocessed to remove interfering and abnormal data, and to select high-quality, interference-free data that meets the analysis requirements, thus providing a reliable data foundation for subsequent dimensional value extraction.

[0095] In this preferred embodiment, by cleaning the power inspection data before extracting the dimensional values, higher quality model input data is obtained.

[0096] In another preferred embodiment, it further includes:

[0097] The current warning judgment value is calculated based on the current weight and the current power inspection data. If the current warning judgment value is less than the above-mentioned preset warning judgment threshold, no power-related hazard warning will be issued for the power consumption area to be inspected.

[0098] Specifically, when the current warning judgment value is found to be less than the above-mentioned preset warning judgment threshold, the result that no power-related hazard warning will be issued for the power consumption area to be detected can be obtained directly, and there is no need to perform subsequent related error calculations again.

[0099] Preferably, the preset warning judgment threshold is set to 0.5. For example, assuming that the dimension of the image modal data is 0.6, the dimension of the sound modal data is 0.4, and the sensor modal data has no obvious abnormalities and the dimension is close to 0, the warning judgment factor is obtained by weighted calculation and is 0.3. Then, there is no potential electrical hazard in the power consumption area to be detected.

[0100] In this preferred embodiment, if the current warning judgment value is less than the preset warning judgment threshold, no warning is issued for the power consumption area to be detected.

[0101] Step S103: Calculate the target warning judgment value based on the current power inspection data and target weight under each mode, and issue a power-related hazard warning for the power consumption area to be inspected if the target warning judgment value is not less than the preset warning judgment threshold; otherwise, do not issue a power-related hazard warning for the power consumption area to be inspected.

[0102] Specifically, the current dimension value is first extracted from the current power inspection data under each modality using a preset feature extraction model. Then, the target early warning judgment value is calculated based on the same calculation method as the above early warning judgment value, but the weights in the formula need to be replaced with the target weights.

[0103] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments.

[0104] like Figure 2 As shown, one embodiment of the present invention provides a monitoring and early warning device for potential electrical hazards in a power system, comprising:

[0105] The module includes a data acquisition module, a target weight determination module, and a monitoring and early warning module.

[0106] The aforementioned data acquisition module is used to acquire initial power inspection data of the power system under several modes within the same preset time period of the power consumption area to be tested, as well as several initial power inspection verification data for data verification.

[0107] The aforementioned target weight determination module is used to repeatedly execute the early warning judgment operation until the current error between the current power inspection data and the corresponding current power inspection verification data in each mode is less than the corresponding preset error threshold, and the current weight is used as the target weight; wherein, the aforementioned early warning judgment operation is as follows:

[0108] Obtain the current weight, the current power inspection data under each mode, and the current power inspection verification data; wherein, the initial weight is the preset initial weight, the initial power inspection data is the aforementioned initial power inspection data, and the initial power inspection verification data is the aforementioned initial power inspection verification data.

[0109] The current warning judgment value is calculated based on the current weight and the current power inspection data. If the current warning judgment value is not less than the above-mentioned preset warning judgment threshold, the current error between the current power inspection data and the current power inspection verification data under each mode is calculated.

[0110] If any current error is greater than the corresponding preset error threshold, the current weight corresponding to the current error that is greater than the preset error threshold is updated, and the power inspection data and power inspection verification data in the next preset time period are obtained.

[0111] The aforementioned monitoring and early warning module is used to calculate the target early warning judgment value based on the current power inspection data and target weight under each mode, and to issue an early warning for potential electrical hazards in the power consumption area to be inspected if the target early warning judgment value is not less than the preset early warning judgment threshold; otherwise, it will not issue an early warning for potential electrical hazards in the power consumption area to be inspected.

[0112] In a preferred embodiment, the target weight determination module includes:

[0113] Feature extraction unit and judgment value calculation unit;

[0114] The aforementioned feature extraction unit is used to extract features from the current power inspection data under each modality using a preset feature extraction model, thereby obtaining the dimension values ​​of the power inspection data under each modality.

[0115] The aforementioned judgment value calculation unit is used to calculate the current warning judgment value based on the values ​​of each dimension and the corresponding current weights.

[0116] In another preferred embodiment, the feature extraction unit includes: a model training subunit;

[0117] The aforementioned model training subunit is used to acquire several historical power grid inspection data with real labels; wherein, the real labels are used to represent the real dimension values ​​of each historical power grid inspection data; the aforementioned historical power grid inspection data is input into the feature extraction model to be trained for iterative training until the loss function converges, resulting in trained preset historical power grid inspection data; wherein, in each iteration of training, the current predicted dimension value is obtained based on the current historical power grid inspection data; the current loss function is calculated based on the current predicted dimension value and the corresponding real label; and it is determined whether the current loss function has converged; if it has converged, the current feature extraction model is used as the aforementioned preset feature extraction model; otherwise, the model parameters in the current feature extraction model are adjusted, and training continues.

[0118] It should be noted that the device embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationship between modules indicates that they have a communication connection, which can be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without creative effort. The above schematic diagram is merely an example of a monitoring and early warning device for potential electrical hazards in a power system and does not constitute a limitation on such a device. It may include more or fewer components than shown, or combine certain components, or use different components.

[0119] Based on the above method embodiments, the present invention provides corresponding terminal device embodiments.

[0120] Another embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the power system electrical hazard monitoring and early warning method described in any embodiment of the present invention.

[0121] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the device.

[0122] The aforementioned terminal devices may be computing devices such as desktop computers, laptops, handheld computers, and cloud servers. These devices may include, but are not limited to, processors and memory.

[0123] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. This processor is the control center of the device, connecting various parts of the device via various interfaces and lines.

[0124] The aforementioned memory can be used to store the aforementioned computer programs and / or modules. The aforementioned processor implements various functions of the aforementioned device by running or executing the computer programs and / or modules stored in the aforementioned memory, and by calling data stored in the memory. The aforementioned memory may mainly include a program storage area and a data storage area, wherein the program storage area may store the operating system, at least one application program required for a function, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0125] Based on the above method embodiments, the present invention provides corresponding storage medium embodiments.

[0126] Another embodiment of the present invention provides a storage medium including a stored computer program, wherein, when the computer program is running, it controls the device where the storage medium is located to execute the power system electrical hazard monitoring and early warning method described in any embodiment of the present invention.

[0127] In this embodiment, the storage medium is a computer-readable storage medium, and the computer program includes computer program code, which may be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium may include any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0128] The above are preferred embodiments of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for monitoring and early warning of electrical hazards in a power system, characterized in that, The method comprises the following steps: acquiring initial power inspection data of a to-be-detected power consumption area in a same preset time period under several modes of a power system and several initial power inspection verification data for data verification; repeatedly performing a pre-warning judgment operation until current errors between current power inspection data under each mode and corresponding current power inspection verification data are all less than corresponding preset error thresholds, and a current weight is taken as a target weight; calculating a target pre-warning judgment value according to the current power inspection data under each mode and the target weight, and performing an electricity-related hidden danger pre-warning for the to-be-detected power consumption area in a case where the target pre-warning judgment value is not less than a preset pre-warning judgment threshold; otherwise, not performing the electricity-related hidden danger pre-warning for the to-be-detected power consumption area; wherein the pre-warning judgment operation comprises the following steps: acquiring a current weight, current power inspection data under each mode and current power inspection verification data; wherein an initial weight is a preset initial weight, initial power inspection data is the initial power inspection data, and initial power inspection verification data is the initial power inspection verification data; calculating a current pre-warning judgment value according to the current weight and the current power inspection data, and calculating current errors between the current power inspection data under each mode and the current power inspection verification data in a case where the current pre-warning judgment value is not less than the preset pre-warning judgment threshold; in a case where any one of the current errors is greater than a corresponding preset error threshold, updating a current weight corresponding to the current error greater than the preset error threshold, and acquiring power inspection data and power inspection verification data in a next preset time period.

2. The method of claim 1, wherein the method comprises: The calculation of the current pre-warning judgment value according to the current weight and the current power inspection data comprises the following steps: extracting features of the current power inspection data under each mode by using a preset feature extraction model to obtain dimension values of the power inspection data under each mode; calculating the current pre-warning judgment value according to the dimension values and corresponding current weights.

3. The method of claim 2, wherein the method further comprises: The training of the preset feature extraction model comprises the following steps: acquiring several historical power inspection data with real labels; wherein the real labels are used to represent real dimension values of the historical power inspection data; inputting the historical power inspection data into a feature extraction model to be trained for iterative training until a loss function converges, and obtaining trained preset historical power inspection data; wherein in each iterative training, a current predicted dimension value is obtained according to current historical power inspection data, a current loss function is calculated according to the current predicted dimension value and a corresponding real label, and it is determined whether the current loss function converges; if the current loss function converges, the current feature extraction model is taken as the preset feature extraction model; otherwise, after adjusting model parameters in the current feature extraction model, the training is continued.

4. The method of claim 3, wherein the power system involves electrical hazards. Before obtaining the dimension values of the power inspection data under each mode, the current power inspection data under each mode is further subjected to data cleaning.

5. The method of claim 4, wherein the power system involves electrical hazards. The method further comprises the following steps: The current early warning judgment value is calculated according to the current weight and the current power inspection data, and in the case that the current early warning judgment value is less than the preset early warning judgment threshold, no power-related hidden danger early warning is performed on the to-be-detected power consumption area.

6. A monitoring and early warning device for electrical hazards in a power system, characterized in that, Comprise: Data acquisition module, target weight determination module and monitoring and early warning module; The data acquisition module is configured to acquire initial power inspection data of the to-be-detected power consumption area in a same preset time period under a plurality of modes of a power system and a plurality of initial power inspection verification data for data verification; The target weight determination module is configured to repeatedly perform an early warning judgment operation until a current error between the current power inspection data under each mode and the corresponding current power inspection verification data is less than a corresponding preset error threshold, and the current weight is taken as a target weight; wherein the early warning judgment operation comprises: Acquiring the current weight, the current power inspection data under each mode and the current power inspection verification data; wherein the weight at the initial time is a preset initial weight, the power inspection data at the initial time is the initial power inspection data, and the power inspection verification data at the initial time is the initial power inspection verification data; According to the current weight and the current power inspection data, a current early warning judgment value is calculated, and in the case that the current early warning judgment value is not less than the preset early warning judgment threshold, a current error between the current power inspection data and the current power inspection verification data under each mode is calculated; In the case that any one current error is greater than the corresponding preset error threshold, the current weight corresponding to the current error greater than the preset error threshold is updated, and the power inspection data and the power inspection verification data in the next preset time period are acquired; The monitoring and early warning module is configured to calculate a target early warning judgment value according to the current power inspection data under each mode and the target weight, and in the case that the target early warning judgment value is not less than the preset early warning judgment threshold, perform power-related hidden danger early warning on the to-be-detected power consumption area; otherwise, no power-related hidden danger early warning is performed on the to-be-detected power consumption area.

7. The device for monitoring and early warning of electrical hazards of power systems according to claim 6, characterized in that, The target weight determination module comprises: A feature extraction unit and a judgment value calculation unit; The feature extraction unit is configured to perform feature extraction on the current power inspection data under each mode through a preset feature extraction model to obtain a dimension value of the power inspection data of each mode; The judgment value calculation unit is configured to calculate a current early warning judgment value according to each dimension value and the corresponding current weight.

8. The device for monitoring and early warning of electrical hazards of power systems according to claim 7, characterized in that, The feature extraction unit comprises a model training subunit; The model training subunit is configured to obtain historical power inspection data with real labels, wherein the real labels are used to represent real dimension values of the historical power inspection data; input the historical power inspection data into a feature extraction model to be trained for iterative training until a loss function converges, and obtain preset historical power inspection data that has been trained; wherein, at each iteration, a current predicted dimension value is obtained according to the current historical power inspection data; a current loss function is calculated according to the current predicted dimension value and the corresponding real label; and it is determined whether the current loss function converges; if the current loss function converges, the current feature extraction model is used as the preset feature extraction model; otherwise, the model parameters in the current feature extraction model are adjusted, and the training is continued.

9. A terminal device, comprising: A computer program product includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and the processor implements the power system electrical hidden danger monitoring and early warning method according to any one of claims 1 to 5 when executing the computer program.

10. A storage medium, characterized by The storage medium includes a stored computer program, wherein the computer program controls the device where the storage medium is located to execute the power system electrical hidden danger monitoring and early warning method according to any one of claims 1 to 5 when the computer program is running.

Images