A power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance

Abstract

The application discloses a power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance, relates to the technical field of power data fusion, and solves the technical problem that the prior art does not optimize the fault instantaneous data conflict characteristics, specifically, a modular multi-source heterogeneous data fusion research and judgment architecture is built in the system, a patrol data fusion unit, a power grid data fusion unit and an aging attenuation unit are respectively configured, joint patrol research and judgment of a power transmission line, accurate disposal of sudden faults of a distribution network and aging attenuation evaluation of primary equipment are realized, and full-scene intelligent operation and maintenance is realized; through multi-scene, multi-dimensional and hierarchical multi-source heterogeneous data fusion logic, traditional power grid operation and maintenance data islands are broken, passive disposal of power grid hidden dangers is changed to active prediction, fault research and judgment is changed from one-sided judgment to accurate positioning, and operation and maintenance work is changed from experience to data quantification, overall upgrading is realized, power grid operation and maintenance efficiency, operation and maintenance accuracy and power grid operation stability are effectively improved, and the modern power grid full-scene intelligent operation and maintenance demand is adapted.

Description

Technical Field

[0001] This invention relates to the field of power data fusion technology, specifically to a power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance. Background Technology

[0002] With the continuous advancement of the construction of new power systems, the power grid structure is becoming increasingly complex. Transmission and distribution equipment are being put into operation on a large scale and routinely. Power grid operation and maintenance scenarios are showing a diversified, refined, and dynamic development trend. As the basic carriers of power grid operation, core power equipment such as overhead transmission lines, transformers, and circuit breakers are exposed to complex outdoor environments for a long time. Affected by multiple factors such as natural environment, operating load, equipment aging, and external damage, they are prone to various hidden dangers such as line damage, hardware corrosion, insulation attenuation, abnormal operating parameters, and sudden short-circuit grounding faults, which directly affect the stability and security of power grid supply.

[0003] However, in existing technologies, the detection of overhead lines in the power grid lacks data fusion, which poses a risk of reduced accuracy. Furthermore, it has not been optimized for the characteristics of instantaneous data conflict during faults. When heterogeneous data evidence from multiple sources of faults conflicts, the fusion results deviate greatly, leading to fault location deviation and misjudgment of causes, which cannot meet the needs of emergency fault response.

[0004] To address the aforementioned technical shortcomings, a solution is proposed. Summary of the Invention

[0005] The purpose of this invention is to solve the problems mentioned above by proposing a power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] A power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance, the system includes:

[0008] The inspection data fusion unit collects multi-source data from overhead line inspections and multi-source data from power grid faults, and labels the multi-source data from overhead line inspections as background data, feature data, and fact data according to their types; and performs fusion analysis on the background data, feature data, and fact data to output line risk signals, line safety signals, line damage signals, inefficient operation and maintenance signals, or efficient operation and maintenance signals.

[0009] The power grid data fusion unit performs fusion analysis on alarm data from power grid protection devices, electrical quantity change data from measurement and control devices, and waveform data from line sensor faults in the scenario of instantaneous handling of distribution network faults, and outputs power grid sensor fault signals, instantaneous electrical quantity change signals, or power grid response lag signals.

[0010] The aging attenuation unit acquires structured insulation detection data, real-time temperature and vibration sensing time-series data, and load operation record data within the aging determination period when power failures are frequent and there is no data traceability. It outputs the aging determination period or the stable operation period and generates aggravated or smoothed time-series markers.

[0011] Furthermore, the background data consists of structured ledger data on tower numbers, years of operation, and rated parameters; the feature data consists of inspection images of line damage and hardware corrosion taken by drones; and the factual data consists of manually entered textual records of defect locations and phenomena.

[0012] Furthermore, when the inspection data fusion unit performs fusion analysis on multi-source data containing only background data, it extracts the rated parameter ledger data, identifies the frequency of rated parameters exceeding the set threshold, and calculates the rate of increase of the frequency of parameter exceeding the threshold based on the years of operation; if the rate of increase exceeds the peak rate of the historical fault cycle, a line risk signal is generated; otherwise, a line safety signal is generated.

[0013] Furthermore, when the inspection data fusion unit performs fusion analysis on multi-source data including background data and feature data, it analyzes the area change trend of the line damage area captured by the drone. If the damage area shows an increasing trend, or if the area of ​​the rusted location in the hardware corrosion inspection image data exceeds the warning area red line value, a line damage signal is generated. If the current background data indicates a line risk, an aging label is set for the line damage signal. If the current background data indicates a line safety, an external force label is set for the line damage signal.

[0014] Furthermore, when the inspection data fusion unit performs fusion analysis on multi-source data including background data, feature data, and factual data, it extracts factual data from the pre-maintenance and post-maintenance periods, collects the frequency of the same defect at the defect location in the two periods, and the frequency of consecutive no text records at the defect location in the pre-maintenance period in the post-maintenance period; if the frequency of the same defect exceeds the same defect frequency threshold, or the frequency of consecutive no text records is lower than the no text record frequency threshold, an inefficient maintenance signal is generated; otherwise, an efficient maintenance signal is generated.

[0015] Furthermore, the power grid data fusion unit marks the fault time when the power grid enters the set fault range based on the power grid characteristic parameters, forming a preset fault time period, and collects power grid protection device action alarm data, electrical quantity change data of the measurement and control device, and line sensor fault waveform data; it establishes the initial response time based on the protection device action alarm data and marks it as the power grid response time; it obtains the change time and change waveform based on the electrical quantity change data; and it obtains the monitoring data waveform and the initial fluctuation time based on the line sensor fault waveform data and marks it as the monitoring time.

[0016] Furthermore, the power grid data fusion unit compares the consistency between the changing waveform and the monitoring data waveform, as well as the consistency between the moment of abrupt change and the monitoring moment:

[0017] If the changing waveform is consistent with the monitoring data waveform and the abrupt change time is consistent with the monitoring time, it is inferred that the power grid monitoring fluctuation is updated in real time; if the changing waveform is inconsistent with the monitoring data waveform and the abrupt change time is earlier than the monitoring time, a power grid sensing fault signal is generated when the changing waveform has changed by several units of time before the power grid response time; if the changing waveform is inconsistent with the monitoring data waveform and the abrupt change time is later than the monitoring time, an instantaneous change signal of electrical quantity is generated.

[0018] Furthermore, in the scenario of real-time updates of power grid fluctuations, if the power grid response time is later than the electrical quantity change time, a power grid response lag signal is generated; if the power grid response time is synchronized with the electrical quantity change time, electrical quantity fluctuation detection is performed on the historical adjacent time corresponding to the changed waveform to identify whether the electrical quantity fluctuation trend at adjacent time is consistent with the changed waveform. If they are consistent, the sensing identification cycle is expanded during operation and maintenance and the accuracy of sensor monitoring values ​​in a single monitoring cycle is increased.

[0019] Furthermore, the aging attenuation unit records periods of frequent power failures with no data traceability, marking these periods as aging determination periods. It acquires structured insulation test data, real-time temperature and vibration sensing time-series data, and load operation record data of the power equipment within the aging determination period. When the structured insulation test data or real-time temperature and vibration sensing time-series data exceeds the set value range, and the load operation record data continues to increase and exceeds the record peak of adjacent historical cycles, the aging determination period is marked as the aging determination period; otherwise, it is marked as the stable operation period.

[0020] Furthermore, the aging attenuation unit uploads the aging determination period to the system, collects the time of power failure and the time of operation and maintenance of power equipment. If the frequency of power failure exceeds the set frequency threshold, or the time interval of operation and maintenance of power equipment is lower than the set time interval threshold, the current aging determination period is marked as an aggravated time sequence; otherwise, it is marked as a smooth time sequence. The system performs localized targeted operation and maintenance of the power grid based on the aggravated or smooth time sequence.

[0021] Compared with the prior art, the beneficial effects of the present invention are:

[0022] 1. It achieves time-series linkage and fusion of three types of heterogeneous data: structured ledgers, image features, and text records. It no longer relies on single data to judge the line status, significantly improving the comprehensiveness and accuracy of overhead line anomaly identification. It effectively avoids the problems of missed and false judgments, and can identify operational risks in advance when equipment parameters deteriorate slowly and no obvious defects appear. It promotes the transformation of line operation and maintenance from passive emergency repair to proactive prediction, and reduces the probability of sudden line failures from the root. Based on the risk status of background data, it accurately matches aging labels and external force labels for damage and corrosion defects identified by UAV images, accurately distinguishes the causes of defects, and solves the problem of homogenization of traditional operation and maintenance solutions.

[0023] 2. It can accurately identify various fault types, effectively distinguish between line sensing hardware lag faults, power grid short-circuit faults, and protection device response lag problems, and solve the problems of traditional fault analysis and judgment location deviation and cause misjudgment, so as to achieve accurate location of fault root cause and provide accurate data support for emergency response.

[0024] 3. It can accurately capture weak aging and attenuation signals that are covered by regular data fluctuations during long-term equipment operation, effectively solving the problem of hidden equipment aging risks that cannot be identified in advance, and realizing a comprehensive and objective assessment of equipment aging status; through multi-dimensional data linkage judgment, it can accurately distinguish between stable equipment operation status and aging operation status, and realize accurate identification of equipment aging problems; based on fault frequency and maintenance interval data, it can further distinguish between aging acceleration time sequence and smooth time sequence, and realize graded judgment of equipment aging status. Attached Figure Description

[0025] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0026] Figure 1 This is a system principle block diagram of the present invention;

[0027] Figure 2 This is a flowchart of the method of the present invention. Detailed Implementation

[0028] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] 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 the invention. 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.

[0030] Please see Figures 1-2 As shown, a power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance is used for power grid operation and maintenance, that is, to perform multi-scenario analysis and detection of the power grid to determine the direction of operation and maintenance, thereby improving the efficiency of power grid operation; and to improve the accuracy and relevance of power grid data analysis through multi-source heterogeneous data fusion.

[0031] Overhead lines are a major component of the power grid, and the inspection of overhead lines is currently carried out independently in the existing operation and maintenance inspection methods, which lacks the accuracy brought by data fusion. Therefore, the inspection data fusion unit in this system performs inspection data fusion analysis on overhead lines to infer whether there are any abnormalities in the overhead lines, and improves the efficiency of power grid operation and maintenance through fusion and alignment.

[0032] This unit is adapted to the scenario of daily joint inspection of power transmission lines by manpower and drones;

[0033] Multi-source data was collected for overhead line inspections, including tower numbers, years of operation, structured ledger data of rated parameters, damaged areas of the line captured by drones, inspection images of rusted fittings, manually entered defect locations, and textual records of phenomena. Among these, tower numbers, years of operation, and structured ledger data of rated parameters were uniformly marked as background data; drone-captured images of damaged lines and rusted fittings were marked as feature data; manually entered textual records of defect locations and phenomena were marked as factual data; and the textual records of phenomena were represented as records of defects, wear, etc., at each location.

[0034] The system sets the time sequence of multi-source data collection based on the collection time of the multi-source data, and identifies the multi-source data collected by the overhead line at the current time. When only background data exists, it extracts the rated parameter structured ledger data from the background data, and identifies the frequency of rated parameters exceeding the set threshold in the ledger data. Based on the increase in the number of years of operation, it obtains the rate of increase of the frequency of parameter exceeding the threshold. When the rate of increase exceeds the peak rate of the historical fault cycle, a line risk signal is generated and uploaded to the system. After receiving the data, the system continuously collects data on the current overhead line, such as increasing the range or frequency of drone or manual inspections. If the rate of increase exceeds the peak rate of the historical fault cycle, a line safety signal is generated and uploaded to the system.

[0035] When generating feature data, that is, multi-source data includes background data and feature data;

[0036] The system analyzes the feature data. If the area of ​​line damage captured by the drone shows an increasing trend, or if the area of ​​rust in the hardware corrosion inspection image data exceeds the warning area red line value of the corresponding area, a line damage signal is generated. When the background data indicates line risk, the current line damage signal is tagged with an aging label and uploaded to the system. When the background data indicates line safety, the current line damage signal is tagged with an external force label and uploaded to the system. After receiving the data, the system formulates an overhead line maintenance plan based on the label type, performs maintenance on the overhead line, and sets up manual inspection.

[0037] When generating factual data, multi-source data includes background data, feature data, and factual data.

[0038] During maintenance, factual data from the period before and after maintenance are extracted. The frequency of the same defect at the defect location is collected in both periods, as well as the frequency of consecutive undocumented defects at the defect location in the period before maintenance. If the frequency of the same defect at the defect location in both periods exceeds the set threshold for the frequency of the same defect, or if the frequency of consecutive undocumented defects at the defect location in the period before maintenance is lower than the threshold for the frequency of undocumented defects, an inefficient maintenance signal is generated and uploaded to the system. After receiving the signal, the system re-formulates the maintenance plan, re-inspects the overhead lines, and terminates the current line task.

[0039] If the frequency of the same defect at the same location does not exceed the set threshold for the frequency of the same defect in two time periods, and the frequency of no text record at the defect location in the previous time period does not fall below the threshold for the frequency of no text record, then an efficient operation and maintenance signal is generated and uploaded to the system. It should be explained that the thresholds involved in the scheme are not based on data collection during the historical detection phase of overhead lines, and are used to evaluate the corresponding type of data based on manually refined and set parameters.

[0040] After confirming that the overhead line inspection is qualified, the built-in power grid data fusion unit of this system performs fault data fusion on the power grid. The specific application scenario is the instantaneous handling of sudden short circuits and grounding faults in the distribution network. At the moment of the fault, protection devices, measurement and control devices, and sensing devices simultaneously upload multi-source fault data, which has the problems of data timing conflicts and numerical contradictions. This is used for accurate fault location and cause analysis. In the existing technology, there is no optimization for the characteristics of instantaneous fault data conflicts. When the evidence of heterogeneous multi-source fault data conflicts, the fusion result is extremely biased, resulting in fault location deviation and misjudgment of cause, which cannot meet the needs of emergency fault handling.

[0041] Fault identification is performed using power grid characteristic parameters, such as voltage and current. When the power grid characteristic parameters fall within the set fault range, the current time is marked as the fault time.

[0042] Extract all fault times to form a preset fault time period, and simultaneously collect power grid protection device action alarm data, electrical quantity change data of measurement and control device, and line sensor fault waveform data; power grid protection device action alarm data is represented as the action data of the protection device, such as circuit breaker tripping action data, etc.

[0043] The initial response time of the protection device is determined based on the alarm data of the power grid protection device, and the current time is marked as the power grid response time; the change time of the electrical quantity is obtained based on the change data of the electrical quantity of the measurement and control device, and the change waveform of the electrical quantity is obtained according to the change time; the waveform of the corresponding electrical quantity identification and monitoring data is obtained based on the line sensor fault waveform data, and the initial fluctuation time is extracted and set as the monitoring time.

[0044] When the waveform of the electrical quantity change is consistent with the waveform of the monitoring data, and the abrupt change time of the electrical quantity is consistent with the monitoring time, it is inferred that the power grid monitoring fluctuation is updated in real time, that is, the line sensor monitoring hardware is fault-free, and the risk of line sensor hardware failure is eliminated. When the waveform of the electrical quantity change is inconsistent with the waveform of the monitoring data, and the abrupt change time of the corresponding electrical quantity is earlier than the monitoring time, it is inferred that the sensor line identification is lagging. If the current power grid has changed the waveform several units of time before the time of response, a power grid sensor fault signal is generated and uploaded to the system. After receiving the signal, the system accurately identifies the current power grid fluctuation electrical quantity and controls the corresponding hardware to slow down the fluctuation period of the waveform, reduce the damage to the power grid hardware, and simultaneously control the electrical quantity until it returns to the set range. It should be explained that the number of several units of time is a manually set number based on the abnormal fluctuation that the current power grid specifications can withstand.

[0045] When the waveform of the change in electrical quantity is inconsistent with the waveform of the monitoring data, and the sudden change time of the corresponding electrical quantity is later than the monitoring time, an instantaneous change signal of electrical quantity is generated and uploaded to the system. After receiving the signal, the system identifies short circuits in all branches of the current power grid and marks the branch location where the electrical quantity value increases instantaneously as the short circuit fault location of the power grid for targeted maintenance.

[0046] When the power grid is in a real-time update scenario of power grid fluctuation monitoring, if the power grid response time is later than the electrical quantity change time, a power grid response lag signal will be generated and uploaded to the system. After receiving the signal, the system will adjust the trigger parameters of the protection device during the current power grid operation and maintenance.

[0047] If the power grid response time is synchronized with the electrical quantity change time, then the electrical quantity fluctuation detection is performed on the historical adjacent time corresponding to the electrical quantity change waveform. If the electrical quantity fluctuation trend and change waveform of adjacent time are consistent, then the sensor identification period is extended during operation and maintenance to increase the accuracy of the sensor monitoring value of a single monitoring period.

[0048] If none of the above situations occur, the system determines it to be a power failure, there is no data traceability, only maintenance is performed and recorded in the work log;

[0049] When the system identifies frequent power outages without data traceability, the built-in aging and degradation identification unit performs aging and degradation analysis on the hardware covered by the power grid. The application scenario is the aging and maintenance of primary equipment such as transformers and circuit breakers that have been in operation for a long time. The temperature rise, insulation resistance, vibration amplitude, and operating load of the equipment are multi-source heterogeneous time-series data with slow and gradual degradation characteristics. Data fusion is required to assess the degree of equipment aging. Therefore, the existing data fusion method only designs fusion logic for abrupt abnormal data and cannot identify the small gradual aging characteristics of equipment in long-term operation. The weak degradation signals of multi-source gradual heterogeneous data are covered by the fluctuations of conventional data, which makes it impossible to predict the hidden dangers of equipment aging in advance and can only identify sudden faults.

[0050] Record periods of frequent power outages with no data traceability and mark them as aging judgment periods. Obtain time-series heterogeneous data of power equipment within the aging judgment period. Power equipment is represented as circuit breakers, transformers and other equipment in the power grid. The time-series heterogeneous data is divided into insulation detection structured data, real-time temperature and vibration sensing time-series data and load operation record data.

[0051] If any of the structured data from insulation testing or the time-series data from real-time temperature and vibration sensing exceeds the set value range, it is inferred that there is a risk in the power grid quantitative data; and if the load operation record data continues to increase and exceeds the record peak of the adjacent historical period, it is inferred that the power equipment is in the aging stage, and the aging judgment period is marked as the aging determination period; historical data is represented as the continuous operating time of the power equipment, the frequency of continuous operation, etc.

[0052] Conversely, if any of the structured data from insulation testing or the real-time temperature and vibration sensing time-series data does not exceed the set value range, it is inferred that there is no risk in the power grid quantification data. The load operation record data is continuously recorded, and the aging judgment period is marked as the stable operation period. It should be explained that the set value range refers to the set threshold range of insulation performance parameters and the set threshold range of temperature and vibration parameters.

[0053] The aging time period is uploaded to the system, which simultaneously identifies the aging time period and collects the time of power failure and the time of operation and maintenance of power equipment. If the frequency of power failure exceeds the set frequency threshold, or the operation and maintenance interval of power equipment is lower than the set interval threshold, the time series of the current aging time period is marked as an aggravated time series; conversely, if the frequency of power failure does not exceed the set frequency threshold and the operation and maintenance interval of power equipment is not lower than the set interval threshold, the time series of the current aging time period is marked as a smooth time series.

[0054] The system performs targeted maintenance on specific parts of the power grid based on the timing type; it should be explained that the maintenance interval is the time interval between the maintenance time and the time of the adjacent fault.

[0055] In summary, for the scenario of joint manual and drone inspection of transmission lines, this approach achieves layered fusion and comprehensive analysis of multi-source heterogeneous inspection data, completely resolving the technical problems of isolated data analysis, one-sided risk identification, unclear defect causes, and inability to verify maintenance effectiveness in traditional inspection methods. For the scenario of instantaneous handling of sudden short circuits and grounding faults in distribution networks, this approach specifically addresses the technical challenges of conflicting multi-source data, large fusion deviations, inaccurate fault location, misjudgment of causes, and inability to identify abnormal equipment responses, significantly improving the emergency response capabilities for sudden distribution network faults. For the aging maintenance scenario of long-term operational primary equipment such as transformers and circuit breakers, this approach overcomes the technical limitations of traditional technologies that can only identify sudden faults and cannot capture subtle, gradual aging characteristics, enabling early prediction and accurate assessment of potential aging hazards in power equipment.

[0056] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance, characterized in that, The system includes: The inspection data fusion unit collects multi-source data from overhead line inspections and multi-source data from power grid faults, and labels the multi-source data from overhead line inspections as background data, feature data, and fact data according to their types; and performs fusion analysis on the background data, feature data, and fact data to output line risk signals, line safety signals, line damage signals, inefficient operation and maintenance signals, or efficient operation and maintenance signals. The power grid data fusion unit performs fusion analysis on alarm data from power grid protection devices, electrical quantity change data from measurement and control devices, and waveform data from line sensor faults in the scenario of instantaneous handling of distribution network faults, and outputs power grid sensor fault signals, instantaneous electrical quantity change signals, or power grid response lag signals. The aging attenuation unit acquires structured insulation detection data, real-time temperature and vibration sensing time-series data, and load operation record data within the aging determination period when power failures are frequent and there is no data traceability. It outputs the aging determination period or the stable operation period and generates aggravated or smoothed time-series markers.

2. The power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance according to claim 1, characterized in that, Background data consists of structured ledger data on tower numbers, years of operation, and rated parameters; feature data consists of inspection images of line damage and hardware corrosion taken by drones; factual data consists of manually entered textual records of defect locations and phenomena.

3. The power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance according to claim 1, characterized in that, When the inspection data fusion unit performs fusion analysis on multi-source data containing only background data, it extracts the rated parameter ledger data, identifies the frequency of rated parameters exceeding the set threshold, and calculates the rate of increase of the frequency of parameters exceeding the threshold based on the years of operation. If the rate of increase exceeds the peak rate of the historical fault cycle, a line risk signal is generated; otherwise, a line safety signal is generated.

4. The power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance according to claim 1, characterized in that, When the inspection data fusion unit performs fusion analysis on multi-source data including background data and feature data, it analyzes the area change trend of the line damage area captured by the drone. If the damage area shows an increasing trend, or if the area of ​​the rusted location in the metal fitting rust inspection image data exceeds the warning area red line value, a line damage signal is generated. If the current background data indicates a line risk, then an aging label is assigned to the line damage signal. If the current background data indicates that the line is safe, then an external force label is assigned to the line damage signal.

5. A power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance according to claim 1, characterized in that, When the inspection data fusion unit performs fusion analysis on multi-source data including background data, feature data, and factual data, it extracts factual data from the pre-maintenance and post-maintenance periods, collects the frequency of the same defect at the defect location in the two periods, and the frequency of consecutive no text records at the defect location in the pre-maintenance period in the post-maintenance period. If the frequency of the same defect exceeds the same defect frequency threshold, or the frequency of consecutive no text records is lower than the no text record frequency threshold, an inefficient maintenance signal is generated; otherwise, an efficient maintenance signal is generated.

6. The power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance according to claim 1, characterized in that, The power grid data fusion unit marks the fault time when the power grid enters the set fault range based on the power grid characteristic parameters, forms a preset fault time period, and collects power grid protection device action alarm data, electrical quantity change data of measurement and control device, and line sensor fault waveform data. The initial response time is determined based on the alarm data of the protection device and marked as the power grid response time; the sudden change time and change waveform are obtained based on the electrical quantity change data; the monitoring data waveform and the initial fluctuation time are obtained based on the line sensor fault waveform data and marked as the monitoring time.

7. A power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance according to claim 6, characterized in that, The power grid data fusion unit compares the consistency between the changing waveform and the monitoring data waveform, as well as the consistency between the moment of abrupt change and the monitoring moment: If the changing waveform is consistent with the monitoring data waveform and the abrupt change time is consistent with the monitoring time, it is inferred that the power grid monitoring fluctuation is updated in real time; if the changing waveform is inconsistent with the monitoring data waveform and the abrupt change time is earlier than the monitoring time, a power grid sensing fault signal is generated when the changing waveform has changed by several units of time before the power grid response time; if the changing waveform is inconsistent with the monitoring data waveform and the abrupt change time is later than the monitoring time, an instantaneous change signal of electrical quantity is generated.

8. A power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance according to claim 7, characterized in that, In the scenario of real-time updates of power grid fluctuations, if the power grid response time is later than the electrical quantity change time, a power grid response lag signal will be generated; if the power grid response time is synchronized with the electrical quantity change time, electrical quantity fluctuation detection will be performed on the historical adjacent time corresponding to the change waveform to identify whether the electrical quantity fluctuation trend and the change waveform are consistent at adjacent time. If they are consistent, the sensing identification cycle will be expanded during operation and maintenance and the accuracy of sensor monitoring values ​​in a single monitoring cycle will be increased.

9. A power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance according to claim 1, characterized in that, The aging attenuation unit records periods of frequent power failures with no data traceability, marks them as aging judgment periods, and acquires structured insulation test data, real-time temperature and vibration sensing time-series data, and load operation record data of power equipment within the aging judgment period. When the structured data of insulation testing or the time-series data of real-time temperature and vibration sensing exceed the set value range, and the load operation record data continues to increase and exceeds the record peak of the adjacent historical cycle, the aging judgment period is marked as the aging determination period. Otherwise, it is marked as a stable operating period.

10. A power communication multi-source heterogeneous data fusion system based on power grid operation and maintenance according to claim 9, characterized in that, The aging and attenuation unit uploads the aging time period to the system, collects the time of power failure and the operation and maintenance time of power equipment. If the frequency of power failure exceeds the set frequency threshold, or the operation and maintenance interval of power equipment is lower than the set interval time threshold, the current aging time period is marked as an aggravated time series; otherwise, it is marked as a smooth time series. The system performs localized targeted operation and maintenance of the power grid based on the aggravated or smooth time series.

Images