Multi-source data power equipment risk intelligent early warning method and system

Abstract

The invention discloses a multi-source data power equipment risk intelligent early warning method and system, and relates to the technical field of power equipment, and the method comprises a first step of power data analysis, a second step of power operation analysis, and a third step of power intelligent early warning. Whether the power equipment of the power grid has an abnormal state is judged, single data limitation is broken through, association analysis is performed on the abnormal power equipment and other normal operation equipment of the power grid, an operation influence network between the equipment is constructed, power grid influence indexes of the abnormal power equipment of the power grid are obtained, multi-level risk early warning is matched, and early warning is performed. Systematic carding of the association relationship between the abnormal power equipment and the normal equipment is realized, the limitation of single equipment analysis in the prior art is broken, the influence conduction path of the abnormal equipment can be comprehensively captured, the risk of power grid fault expansion is reduced, and the overall operation stability of a power grid is guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power equipment, in particular to a multi-source data power equipment risk intelligent early warning method and system. BACKGROUND

[0002] With the continuous expansion of the power grid scale, the number of power equipment continues to increase, and the operation correlation between equipment is more complex. When an abnormal power equipment appears in the power grid, its influence on the overall stability of the power grid often has conduction and diffusion. In the prior art, the processing of abnormal power equipment is mostly focused on fault diagnosis and maintenance of a single device, lacking systematic analysis of the operation correlation between abnormal equipment and normal equipment, and it is difficult to quantify the influence of abnormal equipment on the overall stability of the power grid, resulting in insufficient pertinence and accuracy of risk early warning, which cannot provide accurate decision support for power grid operation and maintenance, and may cause problems such as expansion of power grid failure and decrease of stability. Therefore, it is necessary to analyze a multi-source data power equipment risk intelligent early warning method and system.

[0003] The prior art such as the invention application patent with the publication number CN120873747A discloses a target detection and early warning method and system for power equipment safety active intervention. The method comprises: continuously monitoring the operation environment of the power equipment to obtain multi-source data; based on the multi-source data, identifying a safety hazard target through a target detection algorithm, when the safety hazard target is identified, extracting corresponding multi-source risk features from the multi-source data according to the type of the safety hazard target, and constructing a risk feature spectrum; based on the risk feature spectrum, quantifying the risk of the safety hazard target to obtain a quantized risk vector; determining the risk level corresponding to the quantized risk vector, and executing a warning strategy corresponding to the risk level.

[0004] The prior art of a multi-source data power equipment risk intelligent early warning method and system can meet the basic requirements, but there are also some potential defects and challenges, which are embodied in the following aspects: first, in the prior art, the analysis of the work operation compliance index of each power equipment of the power grid is not accurate enough, which affects the judgment of whether the power equipment of the power grid is in an abnormal state, affects the screening of each abnormal power equipment, reduces the quality of the fused data set, affects the judgment of the abnormal state, increases the problems caused by subsequent control measures due to incorrect identification of abnormal equipment, and affects the link control of the power grid risk.

[0005] II. In the prior art, the correlation analysis of each abnormal power equipment and other normal operation equipment of the power grid is not high enough, which affects the construction of the operation influence network between the equipment, the analysis of the power grid influence index of each abnormal power equipment of the power grid, the matching of multi-level risk early warning, improves the limitations of single equipment analysis in the prior art, affects the capture of the influence transmission path of abnormal equipment, increases the risk of power grid failure expansion, and reduces the overall operation stability of the power grid. SUMMARY

[0006] The purpose of the present application is to provide a multi-source data power equipment risk intelligent early warning method and system, which solves the problems in the background art.

[0007] To solve the above technical problems, the present application adopts the following technical scheme: the present application provides a multi-source data power equipment risk intelligent early warning method, which includes step one, power data analysis, step two, power operation analysis and step three, power intelligent early warning.

[0008] Step one, power data analysis: based on the sensor data, meteorological data and equipment state data extraction of the power equipment of the power grid, and through the dynamic weight distribution algorithm, the multi-source data is standardized, a fusion data set is generated, and the work operation compliance index of each power equipment of the power grid is analyzed.

[0009] Step two, power operation analysis: based on the work operation compliance index of each power equipment of the power grid, it is judged whether the power equipment of the power grid has abnormal state, and each abnormal power equipment is screened.

[0010] Step three, power intelligent early warning: based on each abnormal power equipment, the correlation analysis of each abnormal power equipment and other normal operation equipment of the power grid is carried out, the operation influence network between the equipment is constructed, the influence index of abnormal equipment on the overall power grid stability is evaluated by using graph neural network model, the power grid influence index of each abnormal power equipment of the power grid is obtained, the multi-level risk early warning is matched, and early warning is carried out.

[0011] Further, the fusion data set specifically includes: operation parameter data, external meteorological data and operation state data.

[0012] Further, the working operation compliance index of each power equipment of the power grid is analyzed based on the obtained operation parameter data, external meteorological data and equipment state data, wherein the operation parameter data includes voltage difference, current difference and vibration frequency of each time period of each power equipment of the power grid, the external meteorological data includes humidity difference and external temperature difference of each time period of each power equipment of the power grid, and the equipment state data includes fault stop number, maintenance number and operation time length of each time period of each power equipment of the power grid, and then the working operation compliance index of each power equipment of the power grid is analyzed , external meteorological influence index and equipment state influence index , and the working operation compliance index of each power equipment of the power grid is calculated, and the specific calculation formula is: , wherein, represents the number of each power equipment, , represents the number of each power equipment.

[0013] Further, the operation influence index of each power equipment of the power grid is analyzed based on the obtained voltage difference, current difference and vibration frequency of each time period of each power equipment of the power grid, and the historical voltage difference mean value, historical current difference mean value and historical vibration frequency mean value are extracted from the database and subjected to difference processing, and then divided by the historical standard deviation of the operation parameter data to obtain the voltage standard deviation value, current standard deviation value and vibration standard frequency of each time period of each power equipment of the power grid, and the voltage standard deviation value, current standard deviation value and vibration standard frequency of each time period of each power equipment of the power grid are compared with the voltage standard allowable difference safety interval, current standard allowable difference safety interval and vibration standard allowable frequency safety interval of the power equipment of the power grid stored in the database, if the voltage standard deviation value of a certain time period of a certain power equipment of the power grid is in the voltage standard allowable difference safety interval, the current standard deviation value is in the current standard allowable difference safety interval, and the vibration standard frequency is in the vibration standard allowable frequency safety interval, then the operation influence index of the power equipment of the power grid is recorded as 1, otherwise as -1, and then the operation influence index of each power equipment of the power grid is obtained , wherein, The value of includes 1 and -1.

[0014] Further, the external meteorological influence index of each power equipment of the power grid is obtained by: based on the obtained humidity difference value and the outer surface temperature difference value of each time period of each power equipment of the power grid, the humidity difference value and the outer surface temperature difference value of each time period of each power equipment of the power grid are respectively subjected to mean value processing to obtain the humidity difference mean value and the outer surface temperature difference mean value of each power equipment of the power grid, and the humidity difference mean value and the outer surface temperature difference mean value of each power equipment of the power grid are compared with the humidity difference allowed mean value safety interval and the outer surface temperature difference allowed mean value safety interval of the power equipment of the power grid stored in the database, if the humidity difference mean value of a certain power equipment of the power grid is in the humidity difference allowed mean value safety interval and the outer surface temperature difference mean value is in the outer surface temperature difference allowed mean value safety interval, the external meteorological influence index of the power equipment of the power grid is recorded as 1, otherwise as -1, and then the external meteorological influence index of each power equipment of the power grid is obtained. wherein, including 1 and -1.

[0015] Further, the equipment state influence index of each power equipment of the power grid is obtained by: based on the obtained fault stop number, the maintenance number and the running time length of each time period of each power equipment of the power grid, and compared with the fault stop allowed number safety interval, the maintenance allowed number safety interval and the running time length safety interval of the power equipment of the power grid stored in the database, if the fault stop number of a certain time period of a certain power equipment of the power grid is in the fault stop allowed number safety interval, the maintenance number is in the maintenance allowed number safety interval and the running time length is in the running time length safety interval, the equipment state influence index of the power equipment of the power grid is recorded as 1, otherwise as -1, and then the equipment state influence index of each power equipment of the power grid is obtained. wherein, the value of including 1 and -1.

[0016] Further, whether the power equipment of the power grid has an abnormal state is judged, and each abnormal power equipment is screened, and the specific analysis method is: based on the obtained working operation compliance index of each power equipment of the power grid, the working operation compliance index of each power equipment of the power grid is compared with the working operation compliance index safety interval of the power equipment of the power grid stored in the database, when the working operation compliance index of a certain power equipment of the power grid is in the working operation compliance index safety interval, it is indicated that the power equipment of the power grid has an abnormal state, and the power equipment is recorded as an abnormal power equipment, and then each abnormal power equipment of the power grid is obtained.

[0017] Further, the constructed inter-device operation influence network is used to evaluate an influence index of the abnormal device on the overall power grid stability by a graph neural network model to obtain the power grid influence index of each abnormal power device of the power grid, and the specific analysis method is: based on the obtained each abnormal power device of the power grid, and each abnormal power device is excluded, each normal operation power device of the power grid is obtained, the operation correlation data set between each abnormal power device and each normal operation power device is obtained, and the inter-device operation influence network is established, the inter-device operation influence network is input into the graph neural network model, and the power grid influence index of each abnormal power device of the power grid is obtained.

[0018] Further, the multi-level risk early warning is matched and early warning is performed, and the specific analysis method is: the preset multi-level risk early warning standard is extracted from the database, the power grid influence index of each abnormal power device of the power grid is matched with the multi-level risk early warning standard, the early warning grade corresponding to each abnormal power device is obtained, and according to the early warning grade obtained by matching, the corresponding early warning mechanism is triggered, and is sent to the power grid operation and management platform, and early warning execution is completed.

[0019] The second aspect of the present application provides a system for executing the power device risk intelligent early warning method of the multi-source data, and the system is characterized by comprising: a power data analysis module: based on the sensor data, meteorological data and device state data of the power device of the power grid, and through a dynamic weight distribution algorithm, the multi-source data is standardized to generate a fusion data set, and then the working operation compliance index of each power device of the power grid is analyzed.

[0020] The power operation analysis module: based on the obtained working operation compliance index of each power device of the power grid, whether the power device of the power grid exists an abnormal state is judged, and each abnormal power device is screened.

[0021] The power intelligent early warning module: based on the obtained each abnormal power device, each abnormal power device and other normal operation devices of the power grid are associated and analyzed, the inter-device operation influence network is constructed, the influence index of the abnormal device on the overall power grid stability is evaluated by a graph neural network model to obtain the power grid influence index of each abnormal power device of the power grid, the multi-level risk early warning is matched, and early warning is performed.

[0022] The beneficial effects of the present application are that in step one, power data analysis and step two, power operation analysis: based on the extraction of sensor data, meteorological data and equipment state data of the power equipment of the power grid, and through a dynamic weight distribution algorithm, the multi-source data is standardized to generate a fusion data set, and then analyze the work operation compliance index of each power equipment of the power grid, judge whether the power equipment of the power grid is in an abnormal state, and screen each abnormal power equipment, which breaks through the limitation of single data, improves the quality and pertinence of the fusion data set, improves the judgment of the abnormal state, avoids the subsequent control measures caused by the identification error of the abnormal equipment, and ensures the link control of the power grid risk.

[0023] Step three, power intelligent early warning: based on the obtained each abnormal power equipment, the correlation analysis of each abnormal power equipment and other normal operation equipment of the power grid is carried out, the operation influence network between the equipment is constructed, the influence index of the abnormal equipment on the overall power grid stability is evaluated by using the graph neural network model, the power grid influence index of each abnormal power equipment of the power grid is obtained, the multi-level risk early warning is matched, and early warning is carried out, which realizes the systematic analysis of the correlation between the abnormal power equipment and the normal equipment, breaks through the limitation of single equipment analysis in the prior art, can fully capture the influence transmission path of the abnormal equipment, reduces the risk of power grid failure expansion, and ensures the overall operation stability of the power grid. BRIEF DESCRIPTION OF DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0025] Figure 1 The present application is a method for implementing step flowchart.

[0026] Figure 2 The present application is a system structure connection diagram. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present application will be described in detail below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0028] REFERENCE Figure 1As shown, the present application provides a power equipment risk intelligent early warning method of multi-source data, which includes: step one, power data analysis, step two, power operation analysis and step three, power intelligent early warning.

[0029] It should be noted that step one, power data analysis and step two, power operation analysis are connected, and step two, power operation analysis and step three, power intelligent early warning are connected.

[0030] Step one, power data analysis: based on the sensor data, meteorological data and equipment state data extraction of the power equipment of the power grid, and through the dynamic weight distribution algorithm, the multi-source data is standardized to generate a fusion data set, and then the working operation compliance index of each power equipment of the power grid is analyzed.

[0031] In the above embodiment, the fusion data set specifically includes: operating parameter data, external meteorological data and operating state data.

[0032] In the above embodiment, the working operation compliance index of each power equipment of the power grid is specifically analyzed by: based on the obtained operating parameter data, external meteorological data and equipment state data, wherein the operating parameter data includes: voltage difference, current difference and vibration frequency of each time period of each power equipment of the power grid, the external meteorological data includes: humidity difference and external temperature difference of each time period of each power equipment of the power grid, and the equipment state data includes: fault stop number, maintenance number and running time of each time period of each power equipment of the power grid, and then the running influence index , external meteorological influence index and equipment state influence index of each power equipment of the power grid are analyzed, and then the working operation compliance index of each power equipment of the power grid is analyzed, and the specific calculation formula is: , wherein, represents the number of each power equipment, , represents the number of each power equipment.

[0033] In the above embodiment, the operation influence index of each power device of the power grid is obtained by the following method: based on the obtained voltage difference value, current difference value and vibration frequency of each time period of each power device of the power grid, the historical voltage difference value mean, historical current difference value mean and historical vibration frequency mean are extracted from the database, and the difference value is processed, and then divided by the historical standard deviation of the operation parameter data, to obtain the voltage standard deviation value, current standard deviation value and vibration standard frequency of each time period of each power device of the power grid; the voltage standard deviation value, current standard deviation value and vibration standard frequency of each time period of each power device of the power grid are compared with the voltage standard allowed difference value safety interval, current standard allowed difference value safety interval and vibration standard allowed frequency safety interval of the power device of the power grid stored in the database; if the voltage standard deviation value of a certain time period of a certain power device of the power grid is in the voltage standard allowed difference value safety interval, the current standard deviation value is in the current standard allowed difference value safety interval, and the vibration standard frequency is in the vibration standard allowed frequency safety interval, the operation influence index of the power device of the power grid is recorded as 1, otherwise as -1, and then the operation influence index of each power device of the power grid is obtained , wherein The value of the operation influence index includes 1 and -1.

[0034] In the above embodiment, the external meteorological influence index of each power device of the power grid is obtained by the following method: based on the obtained humidity difference value and surface temperature difference value of each time period of each power device of the power grid, the humidity difference value and surface temperature difference value of each time period of each power device of the power grid are respectively processed by mean value, to obtain the humidity difference mean value and surface temperature difference mean value of each power device of the power grid; the humidity difference mean value and surface temperature difference mean value of each power device of the power grid are compared with the humidity difference allowed mean value safety interval and surface temperature difference allowed mean value safety interval of the power device of the power grid stored in the database; if the humidity difference mean value of a certain power device of the power grid is in the humidity difference allowed mean value safety interval and the surface temperature difference mean value is in the surface temperature difference allowed mean value safety interval, the external meteorological influence index of the power device of the power grid is recorded as 1, otherwise as -1, and then the external meteorological influence index of each power device of the power grid is obtained , wherein The value of the external meteorological influence index includes 1 and -1.

[0035] In the above embodiment, the device state influence index of each power device of the power grid is obtained by the following method: based on the obtained fault stop number, maintenance number and operation time length of each time period of each power device of the power grid, and compared with the fault stop number safety interval, maintenance number safety interval and operation time length safety interval of the power device of the power grid stored in the database, if the fault stop number of a certain time period of a certain power device of the power grid is in the fault stop number safety interval, the maintenance number is in the maintenance number safety interval, and the operation time length is in the operation time length safety interval, the device state influence index of the power device of the power grid is recorded as 1, otherwise as -1, and then the device state influence index of each power device of the power grid is obtained , wherein The value of the device state influence index includes 1 and -1.

[0036] Step two, power operation analysis: based on the obtained work operation compliance index of each power device of the power grid, whether the power device of the power grid exists abnormal state is judged, and each abnormal power device is screened.

[0037] In the above embodiment, whether the power device of the power grid exists abnormal state is judged, and each abnormal power device is screened by the following method: based on the obtained work operation compliance index of each power device of the power grid, the work operation compliance index of each power device of the power grid is compared with the work operation compliance index safety interval of the power device of the power grid stored in the database, when the work operation compliance index of a certain power device of the power grid is in the work operation compliance index safety interval, it is indicated that the power device of the power grid exists abnormal state, and the power device is recorded as an abnormal power device, and then each abnormal power device of the power grid is obtained.

[0038] Step three, power intelligent early warning: based on the obtained each abnormal power device, each abnormal power device and other normal operation devices of the power grid are associated and analyzed, a device-to-device operation influence network is constructed, a graph neural network model is used to evaluate the influence index of abnormal devices on the stability of the whole power grid, the power grid influence index of each abnormal power device of the power grid is obtained, multi-level risk early warning is matched, and early warning is carried out.

[0039] In the above embodiment, the device-to-device operation influence network is constructed, the graph neural network model is used to evaluate the influence index of abnormal devices on the stability of the whole power grid, and the power grid influence index of each abnormal power device of the power grid is obtained by the following method: based on the obtained each abnormal power device of the power grid, each abnormal power device is excluded, each normal operation power device of the power grid is obtained, the operation correlation data set between each abnormal power device and each normal operation power device is obtained, and the device-to-device operation influence network is constructed, the device-to-device operation influence network is input into the graph neural network model, and the power grid influence index of each abnormal power device of the power grid is obtained.

[0040] It should be noted that the correlation data set includes but is not limited to electrical connection relationship, load transmission path relationship, topological structure dependency relationship, operating parameter coupling relationship, and specifically abnormal power equipment and normal operating power equipment, load transmission between normal operating power equipment, parameter coupling value, and connection tightness value.

[0041] It should be noted that the operating influence network between devices is established by taking a single power device as a network node and the operating influence relationship between devices as a network edge, and based on the correlation operating correlation data set, a corresponding weight value is assigned to each network edge.

[0042] It should be noted that the weight value ranges from 0 to 1, and the greater the weight value, the stronger the operating influence between the corresponding two devices.

[0043] It should be noted that the power grid influence index of each abnormal power device of the power grid refers to the influence on the entire power device.

[0044] It should be noted that the graph neural network model involved in the present application has a core innovation in the feature processing logic designed for the power grid device operating influence network. The remaining parameter settings and operation details are all disclosed in the prior art and can be flexibly set by the person skilled in the art according to the actual scene, as follows: Except for the core feature processing and output calculation logic, the standardization settings of the model do not need to be specially limited, including model selection, basic network structure parameters, training process parameters, and graph data preprocessing routine operations, basic network structure parameters such as the number of hidden layers, feature dimensions of each layer, and activation function selection, training process parameters such as batch size, training iteration rounds, regularization coefficient, optimizer type and corresponding learning rate adjustment strategy, graph data preprocessing routine operations such as node feature normalization and standardization method, edge feature encoding format, and graph data batch sampling strategy. The above standardization settings are all general configurations of the graph neural network model, and the person skilled in the art can adjust them according to the power grid device data scale, computing resource conditions and evaluation accuracy requirements, which will not affect the implementation and reproduction of the technical solution of the present application.

[0045] In the above embodiment, the matching multi-level risk early warning is performed, and the early warning is performed, and the specific analysis method is: extracting the preset multi-level risk early warning standard from the database, matching the power grid influence index of each abnormal power device of the power grid with the multi-level risk early warning standard, obtaining the corresponding early warning level of each abnormal power device, and according to the early warning level obtained by matching, triggering the corresponding early warning mechanism and sending it to the power grid operation and maintenance management platform, and completing the early warning execution.

[0046] It should be noted that the early warning level includes: first-level early warning and second-level early warning.

[0047] It should be noted that the corresponding early warning mechanism includes: the corresponding early warning mechanism of the first level early warning is to arrange operation and maintenance personnel for on-site maintenance immediately, reduce the load of the associated line, and issue an early warning signal; the corresponding early warning mechanism of the second level early warning is to continuously observe the abnormal power grid equipment.

[0048] Referring to Figure 2 As shown in the drawings, the present application provides a kind of power equipment risk intelligent early warning method of multi-source data system, it is characterized in that, include: power data analysis module: based on the power equipment of power grid is sensor data, meteorological data and equipment state data extraction, and through dynamic weight distribution algorithm is standardized to multi-source data, generate fusion data set, further analyze the work operation compliance index of each power equipment of power grid.

[0049] Power operation analysis module: based on the work operation compliance index of each power equipment of power grid obtained, judge whether the power equipment of power grid exists abnormal state, and screen each abnormal power equipment.

[0050] Power intelligent early warning module: based on each abnormal power equipment obtained, each abnormal power equipment and other normal operation equipment of power grid are associated analysis, construct the running influence network between equipment, the influence index of abnormal equipment on overall power grid stability is evaluated using graph neural network model, obtain the power grid influence index of each abnormal power equipment of power grid, match multi-level risk early warning, and early warning.

[0051] The above content is only an example and description of the concept of the present application. Those skilled in the art can make various modifications, supplements or substitutions of the described specific embodiments using similar methods, as long as they do not deviate from the concept of the present application or exceed the scope defined by the present application.

Claims

1. A method for intelligent early warning of power equipment risks based on multi-source data, characterized in that, include: Step 1: Power Data Analysis: Based on the extraction of sensor data, meteorological data, and equipment status data of power grid equipment, and the standardization of multi-source data through a dynamic weight allocation algorithm, a fused dataset is generated, and then the operational compliance index of each power grid equipment is analyzed. Step 2: Power Operation Analysis: Based on the obtained operating compliance index of each power equipment in the power grid, determine whether there are any abnormal states of the power equipment in the power grid, and screen out each abnormal power equipment; Step 3: Intelligent Power Early Warning: Based on the identified abnormal power equipment, the abnormal power equipment is correlated with other normally operating equipment in the power grid to construct an operational impact network between equipment. A graph neural network model is used to evaluate the impact index of abnormal equipment on the overall power grid stability, obtain the power grid impact index of each abnormal power equipment, match multi-level risk warnings, and issue warnings.

2. The intelligent early warning method for power equipment risks based on multi-source data according to claim 1, characterized in that, The fused dataset specifically includes: operating parameter data, external meteorological data, and operating status data.

3. The intelligent early warning method for power equipment risks based on multi-source data according to claim 2, characterized in that, The operation of each power device in the power grid conforms to the index, and the specific analysis method is as follows: Based on the obtained operating parameter data, external meteorological data, and equipment status data, including the operating parameter data (voltage difference, current difference, and vibration frequency of each power equipment in the power grid at different time periods), the external meteorological data (humidity difference and surface temperature difference of each power equipment in the power grid at different time periods), and the equipment status data (number of outages due to faults, number of repairs, and operating time of each power equipment in the power grid at different time periods), the operational impact index of each power equipment in the power grid is analyzed. External meteorological impact index and equipment condition impact index Furthermore, the operational compliance index of various power equipment in the power grid is analyzed, and its specific calculation formula is as follows: ,in, This is represented by the serial number of each electrical device. , This represents the quantity of each electrical device.

4. The intelligent early warning method for power equipment risks based on multi-source data according to claim 3, characterized in that, The specific analysis method for the operational impact index of each power equipment in the power grid is as follows: Based on the voltage difference, current difference, and vibration frequency of each power device in the power grid over various time periods, and by extracting the historical average voltage difference, historical average current difference, and historical average vibration frequency from the database, and performing difference processing, the values ​​are divided by the historical standard deviations of the operating parameter data to obtain the standard deviations of voltage, current, and vibration frequency for each power device in the power grid over various time periods. These standard deviations are then compared with the safe ranges for permissible voltage, current, and vibration frequencies stored in the database. If the standard deviation of voltage, current, and vibration frequency of a power device in the power grid falls within the safe range for each of these ranges, the operational impact index of that power device is recorded as 1; otherwise, it is recorded as -1. This process yields the operational impact index of each power device in the power grid. ,in, The values ​​include 1 and -1.

5. The intelligent early warning method for power equipment risks based on multi-source data according to claim 3, characterized in that, The specific analysis method for the external meteorological impact index of each power equipment in the power grid is as follows: Based on the humidity difference and surface temperature difference values ​​of each power device in the power grid over different time periods, the humidity difference and surface temperature difference values ​​of each power device in the power grid over different time periods are averaged to obtain the average humidity difference and average surface temperature difference values ​​of each power device in the power grid. These average humidity difference and average surface temperature difference values ​​of each power device in the power grid are compared with the safe ranges for the average humidity difference and average surface temperature difference stored in the database. If the average humidity difference of a certain power device in the power grid falls within the safe range for both the average humidity difference and the average surface temperature difference, then the external meteorological impact index of that power device in the power grid is recorded as 1; otherwise, it is recorded as -1. This process yields the external meteorological impact index of each power device in the power grid. ,in, Includes 1 and -1.

6. The intelligent early warning method for power equipment risks based on multi-source data according to claim 3, characterized in that, The specific analysis method for the equipment status impact index of each power equipment in the power grid is as follows: Based on the obtained fault outage counts, maintenance counts, and operating durations of each power equipment in the power grid for each time period, and compared with the safe intervals for the allowable outage counts, maintenance counts, and operating durations of the power equipment in the power grid stored in the database, if the number of fault outages, maintenance counts, and operating durations of a certain power equipment in the power grid for a certain time period are within the safe intervals for the allowable outage counts, maintenance counts, and operating durations, then the equipment state impact index of that power equipment in the power grid is recorded as 1; otherwise, it is recorded as -1. This process yields the equipment state impact indexes of each power equipment in the power grid. ,in, The values ​​include 1 and -1.

7. The intelligent early warning method for power equipment risks based on multi-source data according to claim 3, characterized in that, The specific analysis method for determining whether there are abnormal conditions in the power grid's electrical equipment and screening out each abnormal electrical equipment is as follows: Based on the obtained operating compliance index of each power equipment in the power grid, the operating compliance index of each power equipment in the power grid is compared with the safe range of the operating compliance index of the power equipment in the power grid stored in the database. When the operating compliance index of a certain power equipment in the power grid is within the safe range of the operating compliance index, it indicates that the power equipment in the power grid is in an abnormal state, and the power equipment is recorded as abnormal power equipment, thereby obtaining each abnormal power equipment in the power grid.

8. The intelligent early warning method for power equipment risks based on multi-source data according to claim 7, characterized in that, The aforementioned construction of the inter-device operational impact network utilizes a graph neural network model to evaluate the impact index of abnormal devices on the overall power grid stability, thereby obtaining the power grid impact index of each abnormal power device. The specific analysis method is as follows: Based on the abnormal power equipment of the power grid, the abnormal power equipment is removed to obtain the normal operating power equipment of the power grid. The operation correlation dataset between the abnormal power equipment and the normal operating power equipment is obtained, and the operation influence network between the equipment is established. The operation influence network between the equipment is input into the graph neural network model to obtain the power grid influence index of each abnormal power equipment of the power grid.

9. The intelligent early warning method for power equipment risks based on multi-source data according to claim 8, characterized in that, The specific analysis method for matching and issuing multi-level risk warnings is as follows: The system extracts preset multi-level risk warning standards from the database, matches the grid impact index of each abnormal power equipment with the multi-level risk warning standards to obtain the warning level corresponding to each abnormal power equipment, and triggers the corresponding warning mechanism based on the matched warning level, and sends it to the grid operation and maintenance management platform to complete the warning execution.

10. A system for executing the intelligent early warning method for power equipment risks based on multi-source data as described in any one of claims 1-9, characterized in that, include: Power Data Analysis Module: Based on the extraction of sensor data, meteorological data and equipment status data of power grid equipment, and the standardization of multi-source data through dynamic weight allocation algorithm, a fused dataset is generated, and then the operating compliance index of each power grid equipment is analyzed. Power Operation Analysis Module: Based on the obtained operating compliance index of each power equipment in the power grid, it determines whether there are abnormal states of the power equipment in the power grid and filters out each abnormal power equipment; Power Intelligent Early Warning Module: Based on the identified abnormal power equipment, the module performs correlation analysis between the abnormal power equipment and other normally operating equipment in the power grid, constructs an operational impact network between equipment, uses a graph neural network model to evaluate the impact index of abnormal equipment on the overall power grid stability, obtains the power grid impact index of each abnormal power equipment, matches multi-level risk warnings, and issues warnings.

Images