Lightning strike positioning system and method with big data intelligent analysis

Abstract

The application provides a lightning stroke positioning system and method with big data intelligent analysis, and is applied to the field of data processing applications. The application collects multidimensional data, including monitoring, historical lightning strikes, real-time weather, etc., around lightning stroke positioning and power grid risk research and judgment. The monitoring data is processed by a single-chip microcomputer, and the power grid equipment parameters include line insulation level, etc. The original lightning stroke signal is preprocessed by a signal processing algorithm to obtain a data set, then an AI big data model combined with a neural network is used to analyze and fuse data to predict lightning strikes, and power outage research and judgment rules are established combined with electric field data and power grid parameters. The real-time power grid data is used to verify and judge the power-off state. Subsequent synchronous analysis results are presented, time synchronization is ensured by Beidou + dual-mode backup and delay compensation, maintenance data is generated by regular calibration, and finally the maintenance data is processed, related system data is stored, and results are generated.

Description

Technical Field

[0001] This invention relates to the field of data processing applications, and in particular to a lightning strike location system and method with big data intelligent analysis. Background Technology

[0002] Currently, in the field of lightning strike location and monitoring of high-voltage transmission lines, traditional technologies related to lightning strike monitoring, location and timing, power supply, and data analysis exist. However, they still have shortcomings in terms of adaptability to complex environments, monitoring accuracy, intelligent judgment capabilities, and system stability, as detailed below:

[0003] Limitations of lightning strike monitoring technology: Traditional electromagnetic induction lightning strike sensors have low sensitivity (high lower limit of current detection) and a narrow current acquisition range (mostly 500A-6000A), making it difficult to capture weak lightning strike signals (such as long-distance, low-energy lightning strikes) in the complex lightning environment of Angola. They are also susceptible to power grid harmonics and surrounding electromagnetic interference, resulting in high noise in the original signal. This leads to insufficient integrity and accuracy of the monitoring data, making it impossible to provide a reliable data foundation for subsequent fault location and analysis.

[0004] Insufficient positioning and timing accuracy: Traditional base station positioning algorithms have a positioning accuracy of only 200 meters, which is difficult to adapt to the 50-meter / pole spacing of the Angolan transmission lines and cannot accurately pinpoint the specific tower location where lightning strikes occurred; time synchronization relies on a single timing source without a backup mechanism, and signal loss is prone to occur in complex terrain (such as tropical rainforest obstruction), resulting in large clock synchronization deviations between different monitoring terminals and the monitoring center, making it difficult to effectively integrate and analyze multi-source data.

[0005] Weak data analysis and judgment capabilities: Existing technologies mostly rely on human experience or simple algorithms to analyze data, lacking the ability to deeply integrate multi-source data such as historical lightning strikes, real-time weather, power grid operation, and geographic information. This makes it impossible to uncover potential patterns in lightning strikes for prediction, and power outage judgments rely solely on single electric field data without combining power grid equipment parameters (such as line insulation level and tower grounding resistance) with real-time power grid operation data for verification. This can easily lead to misjudgments and is difficult to meet the needs of Angolan power grid operation and maintenance for accurate decision-making.

[0006] Defects in outdoor power supply and system maintenance: Some outdoor monitoring terminals rely on grid power and cannot operate normally in areas with incomplete grid coverage in Angola; traditional solar power supply solutions are not optimized for local sunlight characteristics (unstable sunlight during the rainy season and strong sunlight during the dry season), resulting in low photoelectric conversion efficiency, poor high-temperature stability of energy storage batteries, and a lack of targeted maintenance mechanisms; system maintenance data is not classified and analyzed according to equipment type and fault level, making it difficult to generate operation and maintenance suggestions adapted to the local environment, which affects the long-term stable operation of the system.

[0007] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore includes information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0008] The purpose of this application is to provide a lightning strike location system and method with big data intelligent analysis, which at least to some extent overcomes the problems existing in the prior art. Focusing on the monitoring of transmission lines in Angola, it constructs a full-process system based on "data-driven + intelligent judgment." It collects seven types of data, achieving meter-level positioning and microsecond-level synchronization using BeiDou; after preprocessing, it generates high-quality data, with an LSTM model achieving 85% accuracy in lightning strike prediction and over 90% accuracy in power outage assessment. BeiDou + dual-mode ensures time synchronization, solar energy is used for MPPT self-sufficiency, and maintenance data generates operation and maintenance suggestions, adapting to the local environment.

[0009] Other features and advantages of this application will become apparent from the following detailed description, or may be learned in part by practice of the invention.

[0010] According to one aspect of this application, a lightning strike location method with big data intelligent analysis is provided, comprising: acquiring monitoring data and historical lightning strike data, real-time meteorological data, power grid operation data, geographic information data, power grid equipment parameter data, and historical power outage correlation data; the monitoring data includes lightning current data, electric field intensity change data, monitoring terminal location data, and microsecond-level clock synchronization time data processed by a microcontroller; the power grid equipment parameter data includes line insulation level and tower grounding resistance; preprocessing the original lightning strike signal based on a signal processing algorithm to generate a preprocessed dataset; processing the preprocessed dataset, using an AI big data model combined with a neural network architecture to perform deep analysis of the fused data, mining potential patterns of lightning strike occurrence to achieve lightning event prediction, and running a power outage judgment algorithm. The system combines electric field intensity change data with power grid equipment parameters to construct judgment rules, verifies model outputs using real-time power grid operation data, determines line outage status, and generates analysis results. The analysis results are then presented and synchronized, generating data presentation and synchronization results. System time and model are managed, achieving time reference synchronization through 1PPS pulse signals + NMEA data parsing, and a transmission delay compensation algorithm is designed. GPS / GLONASS dual-mode backup is configured, automatically switching to the backup source when the BeiDou signal is lost, ensuring synchronization continuity. Calibration and maintenance commands are periodically triggered, recording synchronization accuracy deviations and generating system maintenance data. This system maintenance data is then processed, receiving and storing maintenance data from the solar power supply system and time synchronization system, generating system maintenance data processing results.

[0011] Another aspect of this application discloses a lightning strike location device with big data intelligent analysis, characterized by comprising: an acquisition module for acquiring monitoring data and historical lightning strike data, real-time meteorological data, power grid operation data, geographic information data, power grid equipment parameter data, and historical power outage correlation data; the monitoring data includes lightning current data, electric field intensity change data, monitoring terminal location data, and microsecond-level clock synchronization time data processed by a microcontroller; the power grid equipment parameter data includes line insulation level and tower grounding resistance; and a processing module for preprocessing the original lightning strike signal based on a signal processing algorithm to generate a preprocessed dataset; and further processing the preprocessed dataset, using an AI big data model combined with a neural network architecture to perform deep analysis of the fused data, and mining potential patterns in lightning strike occurrence to achieve lightning event prediction. The system utilizes power outage assessment algorithms and models to construct assessment rules based on electric field intensity change data and power grid equipment parameters. Real-time power grid operation data is used to verify the model output, determine line outage status, and generate analysis results. These results are then presented and synchronized, generating data presentation and synchronization results. System time and model management are implemented, achieving time reference synchronization through 1PPS pulse signals and NMEA data parsing, and a transmission delay compensation algorithm is designed. GPS / GLONASS dual-mode backup is configured, automatically switching to the backup source when the BeiDou signal is lost, ensuring synchronization continuity. Calibration and maintenance commands are periodically triggered, recording synchronization accuracy deviations and generating system maintenance data. This system maintenance data is then processed, receiving and storing maintenance data from the solar power supply system and time synchronization system, generating system maintenance data processing results.

[0012] According to another aspect of this application, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a second processor, implements the above-described lightning strike location method with big data intelligent analysis.

[0013] This application presents a lightning strike location system and method with big data intelligent analysis, focusing on the monitoring of transmission lines in Angola. It constructs a complete technical system based on "data-driven + intelligent judgment." In the data acquisition stage, seven types of data are acquired, including monitoring data such as lightning strike current with a range of 20,000A. BeiDou navigation is used to achieve meter-level positioning and microsecond-level clock synchronization at the terminal, adapting to local terrain. Signal preprocessing involves multiple steps to remove invalid signals, generating a high-quality dataset. Analysis and judgment use an LSTM model to predict lightning strikes with an accuracy of 85%. Power outage rules are established by combining electric field formulas and grid parameters, with real-time data verification showing an accuracy exceeding 90%, outputting structured results. In terms of system support, BeiDou + dual-mode ensures time synchronization, with monthly off-peak calibration; solar energy combined with MPPT technology provides self-sufficiency; maintenance data processing generates operation and maintenance suggestions, adapting to the Angolan environment and supporting grid stability.

[0014] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0015] Figure 1 The flowchart illustrates a lightning strike location method with big data intelligent analysis provided in an embodiment of this application;

[0016] Figure 2 This illustration shows a schematic diagram of a lightning strike location device with big data intelligent analysis provided in one embodiment of this application. Detailed Implementation

[0017] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0018] The following is combined with Figure 1 This application describes a lightning strike location method with big data intelligent analysis according to exemplary embodiments. In one embodiment, this application also proposes a lightning strike location system and method with big data intelligent analysis. Figure 1 As shown:

[0019] S101 acquires monitoring data and historical lightning strike data, real-time meteorological data, power grid operation data, geographic information data, power grid equipment parameter data, and historical power outage related data.

[0020] In one implementation, focusing on the core needs of lightning strike location and power grid risk assessment for Angolan transmission lines, the system systematically acquires seven types of data according to the logic of "real-time monitoring - historical reference - environmental support - equipment benchmark": monitoring data, historical lightning strike data, real-time meteorological data, power grid operation data, geographic information data, power grid equipment parameter data, and historical power outage correlation data. All data must be associated with a unique equipment identifier and timestamp to ensure the effectiveness of subsequent data fusion and analysis, providing complete data support for lightning strike prediction and power outage assessment.

[0021] Monitoring data is collected by lightning strike monitoring terminals deployed on transmission line towers in Angola. After processing by a high-speed, low-power microcontroller, the data is uploaded. The core data comprises four types, all adapted to Angola's complex terrain (such as the northern rainforest and the southern semi-arid grassland) and lightning environment: Lightning current data uses a new generation of electromagnetic induction lightning sensors to collect raw current signals. The sensor's sensitivity is 200% higher than traditional sensors, and the current range is expanded to 20,000A, covering both weak and high-intensity lightning strikes. The signal is amplified by a high-speed, high-precision, low-distortion instrumentation amplifier, and then the microcontroller performs data format conversion and preliminary processing. The monitoring terminal on tower number 89 in the northern rainforest region of Angola collected a weak long-distance lightning strike signal with a raw current intensity of 3000A. After amplification by the instrumentation amplifier and processing by the microcontroller, structured data was generated showing a peak current of 3000A, duration of 50μs, and rise slope of 60A / μs. The terminal number "JT-089" and the acquisition time were simultaneously marked.

[0022] The electric field strength change data relies on a capacitance-based electric field sensor to sense changes in the electric field around a high-voltage transmission line. A precision circuit converts the capacitance changes between electrodes into quantified electric field strength. The electric field is stable during normal power transmission and disappears instantaneously after a power outage. The sensor captures this change in real time and transmits it to a microcontroller for processing. For a 10kV high-voltage transmission line in Angola, the electric field sensor is installed 10m away from the conductor. According to the formula... (in, λ is the electric field strength, measured in volts per meter (V / m), used to quantify the strength of the electric field around high-voltage transmission lines; λ is the linear charge density, measured in coulombs per meter (C / m), representing the amount of charge carried per unit length of the transmission line. It is the vacuum permittivity; The distance parameter, in meters (m), represents the vertical distance from the calculation point to the transmission line and the linear charge density. vacuum permittivity The calculation shows that the electric field strength is 1804V / m when the power is being transmitted normally, after processing by the microcontroller; when the line is out of power due to a fault, the output data shows that the electric field strength drops to 0V / m within 1 second.

[0023] Each lightning strike monitoring terminal is equipped with a high-performance BeiDou positioning module, achieving meter-level positioning accuracy in Angola's complex terrain (a 200-meter improvement over traditional base station positioning). This allows for precise location of the tower where the terminal is situated, adapting to the local tower spacing of 50 meters per tower. In a monitoring terminal located in the semi-arid grasslands of southern Angola, the coordinates "11.23°S, 14.56°E" were obtained via the BeiDou positioning module. After processing by a microcontroller, these coordinates were matched against a pre-set tower coordinate database to determine the corresponding tower number 215. The positioning deviation was ≤2 meters, meeting the requirement for accurate tower correlation of lightning strike locations.

[0024] Microsecond-level clock synchronization data, utilizing the BeiDou satellite time service, combined with 1PPS pulse signals and NMEA data parsing, enables microsecond-level clock synchronization between each monitoring terminal and the monitoring center. The microcontroller records the acquisition time of each type of monitoring data, ensuring time alignment of multi-source data. For example, under BeiDou time synchronization, a monitoring terminal recorded the lightning current data acquisition time as "2024-07-05 10:20:33.456789μs". The electric field intensity change data acquisition time deviates from this time by ≤300 microseconds, providing a unified time reference for subsequent multi-data correlation analysis of "lightning current-electric field change-location".

[0025] Records of lightning strikes on Angolan transmission lines over the past three years or more were collected, including the time of the strike, the corresponding tower number, the current intensity, and the weather conditions at the time. A historical lightning strike database was constructed to train AI models on lightning strike patterns. Data from 2021 to 2023 was extracted from the Angolan power company's maintenance archives, including structured records such as "2022-09-03 15:40, Tower No. 178, lightning current 9500A, thunderstorm weather, wind speed 6m / s" and "2023-01-18 08:25, Tower No. 302, lightning current 4200A, light rain weather, wind speed 3m / s," totaling 1080 valid historical lightning strike records. Real-time data on temperature, humidity, air pressure, wind speed, precipitation, and lightning warnings are collected through interfaces with local Angolan meteorological departments or small weather stations deployed along the route. This provides environmental factor support for lightning strike prediction and adapts to the local climate differences between rainy tropical rainforests and arid grasslands. Real-time data is acquired for the northern Angola monitoring area: temperature 29°C, relative humidity 80%, air pressure 1008 hPa, wind speed 4 m / s, with an orange lightning warning issued by the meteorological department; data for the southern monitoring area: temperature 32°C, relative humidity 45%, air pressure 1015 hPa, wind speed 2 m / s, with no lightning warning. Both types of data are updated and transmitted to the system every 5 minutes.

[0026] Real-time grid data, including transmission line current, voltage, and power factor, is accessed from the Angolan power grid dispatch system to verify outage assessment results and determine whether lightning strikes affected normal line operation. Real-time operational data for the line between towers 210-220 is acquired: line current 620A, voltage 110kV, power factor 0.93, and load stable with no fluctuations. When a lightning strike occurs near this line, the line current is briefly monitored to 580A before returning to normal after 10 seconds. This data is used to verify whether changes in electric field strength caused abnormal line operation. Geospatial data of Angolan transmission lines, including line direction, tower distribution, terrain type (rainforest, grassland, mountainous), and tower spacing, is integrated to provide a reference for the spatial correlation between monitoring terminal deployment and lightning strike location. Geographic information of a main line in Angola was obtained based on a GIS system: the line runs from the capital Luanda to the southern city of Benguela, passing through 40 kilometers of tropical rainforest and 60 kilometers of semi-arid grassland, with a total of 1,800 poles and towers spaced 50 meters apart; among them, poles 500-600 are located in mountainous areas with no tall trees or buildings within 30 meters. This data was used for the planning of the installation points of monitoring terminals in the early stage (this section is deployed according to "one unit per pole").

[0027] Key equipment parameters for Angolan transmission lines were collected, including line insulation class and tower grounding resistance. This data provides a benchmark for equipment lightning protection capabilities to assess power outages and is adapted to the varying withstand characteristics of equipment in different regions. Equipment parameters for towers 100-300 were obtained: the line insulation class is 110kV, using composite insulators; the grounding resistance value for towers 100-200 (tropical rainforest area) is ≤8Ω, and the grounding resistance value for towers 201-300 (grassland area) is ≤10Ω. This data is used to determine whether "electric field intensity changes exceed the equipment withstand threshold." For example, for lines with grounding resistance ≤8Ω, the withstand electric field fluctuation threshold is set at 2000V / m; exceeding this threshold indicates a risk of power outage.

[0028] This study compiled records of past power outages on Angolan transmission lines, focusing on outage time, affected tower sections, causes (e.g., lightning strikes, equipment failures), and available lightning strike monitoring data to construct a "lightning strike-power outage" correlation database for use in training AI models to determine outage judgment rules. Outage data from 2022 was extracted: "2022-08-20 16:30, power outage on towers 250-255, caused by lightning strike leading to insulator breakdown; the corresponding terminal recorded a lightning current of 14000A, and the electric field strength dropped sharply from 1800V / m to 0V / m"; "2023-03-12 11:15, power outage on towers 120-122, caused by equipment aging, no lightning strike monitoring data." A total of 280 valid correlation records were integrated, of which 190 were directly related to lightning strikes.

[0029] S102, based on signal processing algorithms, preprocesses the original lightning strike signal to generate a preprocessed dataset.

[0030] In one implementation, the raw lightning strike signal collected by the sensor is input into a high-speed, high-precision, low-distortion instrumentation amplifier. The amplifier parameters are optimized for the signal characteristics of the new generation of electromagnetic induction lightning strike sensors (with 200% higher sensitivity than traditional sensors and an extended current acquisition range of 20,000 A), amplifying the weak signal to improve the signal-to-noise ratio and ensuring that subsequent processing can capture the characteristics of weak lightning strikes. The raw lightning strike signal contains two core types of data: first, the lightning strike current signal within the 20,000 A range (such as the 3,000 A current of a weak lightning strike at a long distance and the 15,000 A current of a strong lightning strike at a short distance); second, the electric field strength signal converted from the capacitance principle (such as the electric field strength of 1804 V / m during normal power transmission on a 10 kV line). A sensor in the tropical rainforest region of northern Angola collected a raw lightning strike current signal with an intensity of only 2500A (signal attenuation due to rainforest obstruction). After being amplified by an instrumentation amplifier, the signal amplitude increased tenfold, and the signal-to-noise ratio increased from 20dB to 45dB. The originally weak current waveform characteristics were clearly distinguishable without introducing any distortion.

[0031] The amplified lightning strike signal is input into a signal denoising module, where wavelet transform denoising and adaptive filtering algorithms are run simultaneously. Wavelet transform is used to separate the transient characteristics of the lightning strike from broadband noise (such as 50Hz harmonics from the power grid), while adaptive filtering is used to dynamically suppress interference that varies with the environment (such as electromagnetic interference from gusts in grassland areas). Finally, a clear lightning strike characteristic signal is extracted, adapting to the environmental differences in different regions of Angola. In the monitoring terminal in the semi-arid grassland region of southern Angola, the amplified lightning strike signal is mixed with power grid harmonic noise (the amplitude of which is about 30% of the useful signal) and electromagnetic interference generated by wind and sand friction. After wavelet transform denoising (using 5 layers of db4 wavelet basis decomposition) and adaptive filtering (using the power grid harmonic frequency as the reference signal), the noise ratio is reduced to below 5%, and the peak value, rise time, and duration of the lightning strike current are completely preserved. The fluctuation amplitude of the electric field strength signal is reduced from ±50V / m to ±10V / m.

[0032] The noise-reduced lightning strike signal is input into the outlier processing module, which uses the 3σ criterion (based on the signal mean and standard deviation) to identify abnormal data. Outliers mainly originate from sensor drift (e.g., high temperatures causing current detection value shifts) and data transmission packet loss (e.g., signal disconnection due to BeiDou communication interruptions). After identification, outliers are corrected using interpolation repair methods (e.g., linear interpolation, nearest neighbor interpolation) to ensure signal data integrity. For example, due to high temperatures during the dry season (daily average above 35℃), a monitoring terminal experienced sensor drift, causing a sudden change in the collected lightning strike current data to 0A (normally around 8000A). The system determined this value to be abnormal (exceeding the mean ± 3 times the standard deviation) using the 3σ criterion. Linear interpolation using 10 consecutive normal data points corrected the outlier value to 7980A, with a deviation of less than 1% from the actual lightning strike current. For the two missing consecutive electric field strength data points caused by transmission packet loss, nearest neighbor interpolation was used to complete them to 1802V / m and 1805V / m, consistent with the trend of surrounding data.

[0033] The outlier-corrected lightning strike signal is input into the time synchronization module. Combined with the microsecond-level clock synchronization time data provided by BeiDou time service (such as "2024-07-10 11:30:45.123456μs"), a time axis alignment algorithm is adopted to unify the time granularity (to the microsecond level) of the microsecond-level lightning strike current signal, the millisecond-level electric field intensity signal, the second-level real-time meteorological data (such as temperature and humidity), and the minute-level power grid operation data (such as line current). Multi-source signal spatiotemporal alignment is achieved through interpolation or resampling to ensure that the data can be correlated in the time dimension. When a lightning strike occurred, the lightning current data was collected at "2024-07-10 11:30:45.123456μs", the electric field strength data was collected at "2024-07-10 11:30:45.123ms" (i.e., 123000μs), and the real-time meteorological data was collected at "2024-07-10 11:30:45s" (i.e., 45000000μs). Using a time-axis alignment algorithm, the electric field strength data was interpolated to the microsecond level (corresponding to the electric field strength value at 123456μs), and the meteorological data was resampled to a microsecond-level timestamp. Finally, all three types of data were associated with the time point "2024-07-10 11:30:45.123456μs", achieving spatiotemporal alignment.

[0034] The spatiotemporally aligned lightning strike signal is input into a high-speed, low-power microcontroller. The microcontroller performs two core tasks: first, format conversion, converting the analog signal format of the lightning strike current waveform and electric field intensity data into a digital structured format (such as JSON); second, preliminary integration, associating and integrating the purified lightning strike current waveform data (including peak value, duration, and rise slope), quantized electric field intensity data (including real-time value and rate of change), and monitoring terminal location data (meter-level coordinates from BeiDou positioning) to generate a structured signal containing multi-dimensional information. After processing the aligned signal, the high-speed, low-power microcontroller generates a structured signal: "Terminal Number: JT-123, Time: 2024-07-10 11:30:45.123456μs, Lightning Current: 8000A (peak value), 60μs (duration), 133A / μs (rise slope), Electric Field Strength: 1804V / m (real-time value), -50V / m (rate of change), Terminal Location: 12.45°S, 13.78°E". The data format meets the requirements of subsequent server storage and analysis.

[0035] The structured signals processed by the microcontroller are input into the data verification module. Based on preset signal quality constraints (such as lightning current amplitude needing to be within the range of 0-20000A, electric field strength data deviation needing to be ≤5%, and position data deviation needing to be ≤5 meters), invalid signals (such as current signals exceeding the measurement range and terminal data with excessive position deviation) are eliminated. Finally, all valid signals are integrated to generate a preprocessed dataset. The data verification module verifies 100 structured signals and finds two invalid signals: one with a lightning current value of 25000A (exceeding the 20000A range, determined to be a sensor malfunction), and one with a terminal position deviation of 8 meters (exceeding the 5-meter threshold, determined to be an anomaly in BeiDou positioning). After these are eliminated, the remaining 98 valid signals are integrated into a preprocessed dataset. Each record in the dataset contains complete lightning strike characteristics, spatiotemporal information, and quality indicators, and can be directly used for AI model training.

[0036] S103 processes the preprocessed dataset, uses an AI big data model combined with a neural network architecture to perform in-depth analysis of the fused data, mines potential patterns of lightning strikes to predict lightning events, runs a power outage judgment algorithm and model to construct judgment rules based on electric field intensity change data and power grid equipment parameters, verifies the model output using real-time power grid operation data, determines the power outage status of the line, and generates analysis results.

[0037] In one implementation, the preprocessed dataset (including purified lightning current, quantified electric field strength, etc.) is imported into the data fusion module along with historical lightning data, real-time meteorological data, power grid operation data, geographic information data, and historical power outage correlation data. After data cleaning, feature engineering, and data standardization preprocessing, the AI ​​big data model combined with a neural network architecture (preferably using LSTM adapted to lightning time series data) is used to mine potential patterns of lightning occurrence and generate lightning event prediction results.

[0038] Invalid records without specific tower locations in historical lightning strike data were removed (e.g., a record from 2022 that only indicated "southern region" without a associated tower number). Missing values ​​such as "empty humidity value" in real-time meteorological data due to sensor malfunctions were processed (by interpolation of nearby time periods). Key features were extracted from the fused data, such as "lightning warning level and hourly precipitation" in meteorological data, "real-time line load" in power grid data, and "terrain type (rainforest / grassland)" in geographic information data. Features of different magnitudes, such as "lightning current (0-20000A)" and "electric field strength (0-2000V / m)," were normalized to the [0,1] interval. Based on 1,080 historical lightning strike data points from 2021 to 2023 and real-time meteorological and power grid data from 2024, an LSTM model was trained to predict that 1-2 lightning strikes would occur in the tropical rainforest region of northern Angola (tower section 80-100) within the next 24 hours. The predicted current intensity is concentrated in the range of 5,000-8,000 A, and the most likely time of occurrence is during afternoon thunderstorms (14:00-16:00). The prediction accuracy reached 85% after cross-validation.

[0039] Based on the power outage assessment algorithm, the electric field intensity change data is processed in conjunction with power grid equipment parameters to generate a preliminary judgment result on the power outage status of the line. The electric field intensity is based on the formula... Calculations were performed to determine the preliminary assessment result by checking whether the change in electric field strength reached the line's outage threshold. For the No. 215 tower line in the southern grassland region of Angola, the power grid equipment parameters are "110kV insulation class, grounding resistance 10Ω". The outage threshold for this line was set as "a sudden drop in electric field strength ≥1500V / m for a duration ≥1 second". At a certain moment, the electric field sensor detected a sudden drop in electric field strength from 1800V / m to 200V / m, a change of 1600V / m, which meets the outage threshold condition. Therefore, it was preliminarily determined that the line was in a state of outage.

[0040] The preliminary judgment of the power outage status of the line was verified and optimized by combining it with real-time power grid operation data. False judgments were eliminated, and a target line power outage status judgment result was generated. The preliminary power outage status judgment result was combined with real-time power grid operation data (line current, voltage, power factor) to verify the correlation between changes in electric field strength and the actual operating status of the line. False judgments were eliminated, and a target judgment result was generated to ensure consistency with the actual operating conditions of the Angolan power grid. For the preliminary power outage judgment of the line at tower No. 215, real-time power grid data revealed that the line current dropped from 600A to 0A, the voltage dropped from 110kV to 0kV, and the power factor approached 0, perfectly matching the sudden drop in electric field strength, thus validating the preliminary judgment. In another case, the electric field strength of the line at tower No. 300 dropped sharply by 1200V / m due to a brief electromagnetic interference (not reaching the threshold). Real-time power grid data showed no significant fluctuations in line current and voltage, indicating a false judgment was made and eliminated, thus avoiding false power outage alarms.

[0041] The lightning strike prediction results and the line outage status assessment results are integrated according to the "time-location-risk" dimension to generate a structured analysis result that includes the predicted lightning strike time, location, intensity, and line outage risk level (low / medium / high). The risk level is determined by combining "lightning strike intensity + equipment withstand capability". The integrated output result is "2024-07-12 14:00-16:00, northern tower section 85-90 (tropical rainforest), predicted 1 lightning strike (current 7000A), line insulation level 110kV, grounding resistance 8Ω, outage risk level 'medium'; southern tower section 215, power outage confirmed, associated lightning current 7500A, priority should be arranged for emergency repair". The results can be directly used for visualization display in the monitoring center and synchronized with the mobile APP.

[0042] S104 presents and synchronizes the analysis results, generating data presentation and synchronization results.

[0043] In one implementation, focusing on the core objective of the Angolan transmission line lightning strike location system—"Operation and Maintenance Decision Support + On-site Emergency Repair Linkage"—the lightning strike prediction results and line outage status judgment results are processed according to the progressive logic of "monitoring center visualization → mobile APP synchronization → data interaction verification." This ensures that the analysis results can be accurately transmitted to operation and maintenance personnel and emergency repair teams, while also guaranteeing the real-time and consistency of data synchronization, thus meeting the needs of local power grid operation and maintenance scenarios (such as cross-terrain emergency repairs in tropical rainforests and semi-arid grassland areas).

[0044] On the high-performance server cluster of the intelligent monitoring center, professional power operation and maintenance monitoring software visualizes the analysis results, presenting them intuitively in a combination of "GIS map + data dashboard," providing operation and maintenance personnel with clear decision-making support. The visualized content must cover core information on lightning strike prediction and power outage status. The GIS map displays the predicted lightning strike location (e.g., the section between towers 85-90 in the north) and the already de-energized line section (e.g., the section between towers 215 in the south) on the Angola transmission line GIS map, using different colors to distinguish between "predicted lightning strike (yellow warning marker)," "power outage (red fault marker)," and "normal operation (green normal marker)." Clicking on a marker displays detailed information (e.g., predicted lightning strike current 7000A, power outage-related lightning strike current 7500A).

[0045] Key data trends are displayed in chart form, such as "Bar chart of lightning strike frequency in various regions in the past 7 days," "Pie chart of real-time power outage number statistics," and "Line chart of lightning current intensity distribution," while also showing core indicators (such as current prediction accuracy of 85% and power outage assessment accuracy of 92%). Through the visual interface, monitoring center maintenance personnel can quickly identify "Medium-risk lightning strikes are expected on towers 85-90 in the northern tropical rainforest area in the next 24 hours" and "Power outage has occurred on tower 215 in the south." Clicking on the tower 215 marker will bring up a details window displaying "Power outage time: 2024-07-12 10:30, associated lightning current: 7500A, line insulation level: 110kV, grounding resistance: 10Ω," providing complete information for maintenance decisions.

[0046] The analysis results from the monitoring center are synchronized in real time to the accompanying mobile app via an encrypted transmission protocol (compatible with overseas 4G communication standards). This ensures that personnel on the field can access the latest fault information. The synchronized content must be tailored to the needs of the emergency repair scenario, highlighting key information such as "location, environment, and materials." Fault location synchronization involves synchronizing the BeiDou meter-level coordinates of the faulty tower (e.g., tower number 215) to the app. Combined with offline geographic information data, this provides navigation routes for repair personnel, while also marking the surrounding terrain (e.g., grassland areas, no obstacles within 50 meters). Emergency repair auxiliary information synchronization includes synchronizing fault-related data (e.g., cause of power outage: lightning strike causing insulator breakdown, type of insulator to be replaced). When the southern repair team receives the power outage synchronization information for tower number 215 on the mobile app, the app automatically loads the tower's coordinates (11.23°S, 14.56°E) and generates a navigation route, while also prompting that "insulating gloves and insulator replacement tools are required; materials can be retrieved from the southern warehouse, 15 kilometers away." Repair personnel can obtain complete repair guidance without returning to the monitoring center.

[0047] During the presentation and synchronization process, a two-way data interaction and verification mechanism between the monitoring center and the mobile app is established to ensure that the analysis results are transmitted without loss or deviation, while also supporting emergency repair personnel to report on the on-site situation, forming a data closed loop. After receiving data, the app automatically compares it with the verification code sent by the monitoring center. If any inconsistency is found (e.g., the lightning current received by the app is 7000A, while the monitoring center sends 7500A), resynchronization is immediately triggered. After emergency repair personnel confirm the fault situation on-site (e.g., "the insulator of tower No. 215 has been replaced, and the line has been restored to power"), they can submit feedback information through the app. This information is synchronized to the monitoring center in real time, updating the status of the visualization interface (e.g., the marker for tower No. 215 changes from red fault to green normal). After the southern emergency repair team completed the emergency repair of tower No. 215, they clicked "Power Supply Restored" on the APP and submitted on-site photos. The monitoring center received feedback in real time. The red fault mark on tower No. 215 in the visualization interface was immediately updated to a green normal mark. At the same time, the "Number of Power Outage Lines in Real Time" on the data dashboard decreased from 3 to 2, ensuring that the information of the maintenance personnel and the emergency repair team was consistent.

[0048] The S105 manages system time and model, achieving time reference synchronization through 1PPS pulse signal + NMEA data parsing, and designs a transmission delay compensation algorithm; it is configured with GPS / GLONASS dual-mode backup, automatically switching to the backup source when the BeiDou signal is lost, ensuring synchronization continuity; it periodically triggers calibration and maintenance commands, records synchronization accuracy deviation values, and generates system maintenance data.

[0049] In one implementation, the BeiDou time synchronization program is used to uniformly manage the system time. A dual mechanism of "1PPS pulse signal + NMEA data parsing" is employed to synchronize the time base between each monitoring terminal and the monitoring center. Simultaneously, a transmission delay compensation algorithm is designed to calibrate signal transmission errors and generate initial time synchronization results, ensuring that the time base accuracy meets the microsecond-level data integration requirements. The 1PPS pulse signal provides a precise trigger point for time synchronization, outputting one pulse signal per second to align the time counting starting points of the monitoring terminals and the monitoring center, avoiding time counting deviations. NMEA data parsing extracts UTC time information and terminal location data from the NMEA protocol data output by the BeiDou positioning module, compensating for the deficiency of the 1PPS signal only triggering without providing time information. The combination of these two methods ensures basic time synchronization accuracy. Transmission delay compensation addresses the differences in signal transmission distances in Angola's complex terrain (e.g., 50 km between the terminal and the monitoring center in rainforest areas and 80 km in grassland areas). The algorithm calculates the signal transmission time in real time (based on the electromagnetic wave propagation speed 3 × 10⁻⁶). 8 (m / s, transmission delay of approximately 267 microseconds over 80 kilometers), dynamically adjusting the compensation value to eliminate synchronization deviations caused by distance.

[0050] The monitoring terminal on tower No. 215 in the southern grassland region of Angola is 80 kilometers away from the monitoring center. After the BeiDou time synchronization program outputs a 1PPS signal, the NMEA data parsing obtains the UTC time as "2024-07-12 10:00:00.000000". The transmission delay compensation algorithm calculates the signal transmission delay to be 267 microseconds. Finally, the terminal time is synchronized to "2024-07-12 10:00:00.000267", with a time deviation of ≤50 microseconds from the monitoring center, meeting the microsecond-level synchronization requirement.

[0051] The BeiDou timing signal status is monitored in real time (signal strength and integrity are checked once per second). A GPS / GLONASS dual-mode backup system is configured. When BeiDou signal loss is detected (e.g., signal strength < -120dBm or data packet loss rate > 10% due to tree obstruction in rainforest areas), the system automatically switches to the GPS or GLONASS backup source, generating a time synchronization continuity guarantee result to ensure uninterrupted time synchronization. Signal status monitoring uses a signal quality detection module to determine in real time whether the BeiDou signal meets the normal thresholds of "signal strength ≥ -110dBm and data packet loss rate ≤ 5%". If the conditions are not met for three consecutive tests, it is determined that the signal is lost. Dual-mode backup switching has a switching response time controlled within 50 milliseconds. After the backup source is activated, the "1PPS + NMEA + delay compensation" mechanism is used to ensure consistent synchronization accuracy (microsecond level) before and after switching, with no significant fluctuations. The monitoring terminal on tower No. 89 in the tropical rainforest region of northern Angola experienced a signal loss due to heavy rain obscuring the BeiDou signal from dense tree canopy, causing the signal strength to drop to -125dBm. The system determined this to be a signal loss and automatically switched to the GPS backup source within 50 milliseconds. After the switch, the synchronization deviation between the terminal time and the monitoring center recovered from 1000 microseconds at the time of the signal loss to ≤50 microseconds, without affecting the accuracy of the timestamp of the lightning strike data.

[0052] The system time and model calibration maintenance instructions are triggered according to a preset cycle (once a month). Operations are conducted during off-peak hours in the Angolan power grid (e.g., 2:00-4:00 AM, when line load is low and monitoring and analysis pressure is low). This process involves two aspects: firstly, verifying the accuracy of the time synchronization system and recording deviation values; and secondly, evaluating the performance of the AI ​​big data model, ultimately generating system maintenance data containing both types of data. Time synchronization accuracy verification: A "reference clock comparison method" is used, comparing the time of each terminal with the high-precision atomic clock at the monitoring center and recording the synchronization accuracy deviation value. If the deviation is >100 microseconds, the time synchronization process is re-executed. Based on actual lightning strike events and model prediction results from the past month, prediction accuracy and recall are calculated (e.g., the model prediction accuracy was 85% and recall was 82% in July). If the accuracy is below 80%, incremental model training is triggered, using new data to optimize parameters. At 2:00 AM on July 1, 2024, the system triggered a calibration and maintenance command. After verifying the time of 20 monitoring terminals, it was found that the synchronization deviation of terminal No. 15 was 120 microseconds. After resynchronization, the deviation was reduced to 40 microseconds. At the same time, the AI ​​model was evaluated. In July, there were 15 actual lightning strikes. The model accurately predicted 13 of them, with an accuracy rate of 86.7%, which met the threshold requirement of ≥80%. No retraining was required. Finally, system maintenance data containing "synchronization deviation data of 20 terminals and model accuracy rate of 86.7%" was generated.

[0053] S106 processes system maintenance data, receives and stores maintenance data from the solar power supply system and time synchronization system, and generates system maintenance data processing results.

[0054] In one implementation, maintenance data from the solar power supply system and the time synchronization system are received in real time and stored in a structured format to establish an initial maintenance dataset. This ensures that the data covers the core operation and maintenance indicators of both systems, providing complete input for subsequent analysis. The maintenance data for the solar power supply system includes three types of core information: solar panel cleaning records (e.g., cleaning date, photoelectric conversion efficiency after cleaning), lithium iron phosphate battery performance parameters (e.g., remaining capacity, charge / discharge cycle count, current temperature), and MPPT charging circuit operating status (e.g., maximum power point tracking accuracy, charging current / voltage). The maintenance data for the time synchronization system includes three types of core information: BeiDou / GPS / GLONASS dual-mode switching records (e.g., switching time, switching reason, signal strength after switching), synchronization accuracy deviation values ​​(e.g., daily time deviation between the terminal and the monitoring center), and transmission delay compensation algorithm calibration results (e.g., compensation value adjustment range, synchronization accuracy after adjustment).

[0055] On July 10, 2024, the system received maintenance data from the monitoring terminal of tower No. 215 in the southern grassland area: the solar panel cleaning record was "cleaned on 2024-07-09, after cleaning the photoelectric conversion efficiency increased from 75% to 88%"; the lithium iron phosphate battery parameters were "remaining capacity 90%, cycle count 350, temperature 32℃"; the MPPT circuit status was "tracking accuracy 95%, charging current 1.2A"; the time synchronization data was "no dual-mode switching occurred on July 10, synchronization accuracy deviation was 45 microseconds, and the transmission delay compensation value remained at 267 microseconds". All data were classified and stored according to terminal number and acquisition time to generate the initial maintenance dataset.

[0056] The initial maintenance dataset is categorized into three dimensions: "equipment type - maintenance time - fault level," and a structured data index is constructed to facilitate rapid retrieval and correlation analysis, thereby improving the efficiency of maintenance data utilization. The equipment type dimension distinguishes between "solar power supply systems (subdivided into solar panels, lithium batteries, and MPPT circuits)" and "time synchronization systems (subdivided into BeiDou modules, GPS modules, and compensation algorithms)." The maintenance time dimension is divided into "daily / weekly / monthly" time granularity, with the specific data collection time marked (accurate to the minute). The fault level dimension is divided into "normal (parameters within the threshold), warning (parameters close to the threshold), and fault (parameters exceeding the threshold)," for example, a lithium battery remaining capacity ≥80% is considered normal, 60%-80% is a warning, and <60% is a fault. After classifying the initial maintenance data from July 1st to 10th, an index was constructed: "Equipment Type - Solar Power System - Lithium Battery; Maintenance Time - 2024-07-10 14:30; Fault Level - Normal", corresponding to the data for terminal lithium battery No. 215: "Remaining Capacity 90%, Cycle Count 350"; "Equipment Type - Time Synchronization System - Compensation Algorithm; Maintenance Time - 2024-07-10 02:00; Fault Level - Normal", corresponding to the data: "Compensation Value 267 microseconds, Synchronization Deviation 45 microseconds". The structured index can support quick data filtering by "a certain area, a certain equipment, a certain time period".

[0057] Based on the local environmental characteristics of Angola (such as high temperatures in the dry season, insufficient sunlight in the rainy season, and strong winds and sandstorms in the grasslands), maintenance data anomaly judgment rules were established. Structured maintenance data was checked item by item to identify abnormal data and generate maintenance data anomaly reports. For solar power systems, a warning was set for "lithium battery temperature > 40℃" during the dry season (average daily temperature ≥ 35℃), and for "solar panel cleaning cycle exceeding 15 days" during the rainy season (monthly rainfall ≥ 200mm). For time synchronization systems, a warning was set for "BeiDou signal strength < -115dBm" in tropical rainforest areas and "synchronization accuracy deviation > 80 microseconds" in grassland areas. The anomaly report is generated as follows, recording the abnormal equipment number, abnormal parameters, abnormal time, related environmental factors, and preliminary handling suggestions. Analysis of structured data from July 10th revealed an anomaly in the time synchronization data of terminal No. 89 in the northern tropical rainforest region: "BeiDou signal strength -120dBm, synchronization accuracy deviation 95 microseconds". Considering the local heavy rain (50mm of precipitation) on that day, it was determined that "the BeiDou signal was obstructed by tree canopy". An anomaly report was generated: "Device No. JT-089, Anomaly Parameters: BeiDou signal strength -120dBm (threshold ≥ -115dBm), synchronization deviation 95 microseconds (threshold ≤ 80 microseconds), Anomaly Time 2024-07-10 09:30, Related Environment: Heavy rain caused tree canopy obstruction, Recommendation: Check the terminal installation location and adjust the antenna angle if necessary".

[0058] Structured maintenance data is linked and integrated with maintenance data anomaly reports. Key information is summarized by "region-equipment-risk level" to generate system maintenance data processing results containing operation and maintenance suggestions, providing a basis for operation and maintenance personnel to formulate targeted maintenance plans. The structured data of each terminal is divided into northern / southern Angola regions, and the risk level (low / medium / high) of each device is marked in conjunction with anomaly reports. Operation and maintenance suggestions are given first for high-risk devices (such as fault status), and the attention cycle is indicated for devices with early warning.

[0059] The output includes a regional maintenance overview, a list of abnormal devices, maintenance priority ranking, and environmental compatibility recommendations. After integrating data from July 1st to 10th, the following processing results were generated: In the northern tropical rainforest area (towers 80-100), out of 10 terminals, 2 had time synchronization anomalies (medium risk level), and it is recommended to check the antenna location within one week; In the southern grassland area (towers 200-220), out of 10 terminals, 1 solar panel had a cleaning cycle exceeding 20 days (low risk level), and it is recommended to focus on it during next month's cleaning; All terminal lithium batteries, MPPT circuits, and time synchronization compensation algorithms are in normal condition, and the overall system maintenance risk is low. The note also states, "With the rainy season approaching, it is recommended to shorten the solar panel cleaning cycle to 10 days / time."

[0060] This application focuses on the monitoring needs of Angolan transmission lines, using "data-driven + intelligent analysis" as its core to construct a complete technical system encompassing "data acquisition, processing, analysis, and maintenance." In the data acquisition phase, it obtains seven types of data from multiple dimensions, including monitoring data (such as 20,000A range lightning current and electric field strength), historical lightning strike data, and real-time weather data. It utilizes BeiDou positioning to achieve meter-level positioning and microsecond-level clock synchronization for the terminal, adapting to the complex local terrain. In the signal preprocessing stage, invalid signals are eliminated through instrument amplification, wavelet transform denoising, 3σ criterion anomaly correction, spatiotemporal alignment, and microcontroller integration, generating a high-quality dataset to ensure data accuracy.

[0061] At the analysis and judgment level, the AI ​​big data model (LSTM architecture) mines lightning strike patterns, achieving a prediction accuracy of 85%. By combining the electric field strength formula with grid parameters, outage judgment rules are constructed, and verified with real-time grid data, the judgment accuracy exceeds 90%. Structured results containing lightning strike predictions and outage risks are generated. In terms of system support and maintenance, the BeiDou + GPS / GLONASS dual-mode system ensures continuous time synchronization, with monthly off-peak calibration. Solar power supply coupled with MPPT technology achieves self-sufficiency. Maintenance data is categorized and anomaly analyzed to generate operation and maintenance suggestions adapted to Angola's climate and terrain, providing technical support for stable grid operation.

[0062] In one implementation, such as Figure 2 As shown, this application also provides a lightning strike location device with big data intelligent analysis, comprising:

[0063] The acquisition module 201 is used to acquire monitoring data and historical lightning strike data, real-time meteorological data, power grid operation data, geographic information data, power grid equipment parameter data, and historical power outage correlation data. The monitoring data includes lightning current data, electric field strength change data, monitoring terminal location data, and microsecond-level clock synchronization time data processed by the microcontroller. The power grid equipment parameter data includes line insulation level and tower grounding resistance.

[0064] Processing module 202 is used to preprocess the original lightning strike signal based on signal processing algorithms to generate a preprocessed dataset; process the preprocessed dataset, use AI big data models combined with neural network architecture to perform deep analysis of the fused data, mine potential patterns of lightning strikes to achieve lightning event prediction, run power outage judgment algorithms and models to construct judgment rules based on electric field intensity change data and power grid equipment parameters, verify the model output using real-time power grid operation data, determine the line power outage status, and generate analysis results; present and synchronize the analysis results to generate data presentation and synchronization results; manage system time and model, achieve time reference synchronization through 1PPS pulse signal + NMEA data parsing, and design a transmission delay compensation algorithm; configure GPS / GLONASS dual-mode backup, automatically switch to the backup source when the BeiDou signal is lost to ensure synchronization continuity; periodically trigger calibration and maintenance commands, record synchronization accuracy deviation values, and generate system maintenance data; process the system maintenance data, receive and store maintenance data from the solar power supply system and time synchronization system, and generate system maintenance data processing results.

[0065] The various embodiments in this application are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments for evaluating the lightning strike location method, electronic device, electronic device, and readable storage medium with big data intelligent analysis are basically similar to the above-described embodiments of the lightning strike location method with big data intelligent analysis, so the description is relatively simple. Relevant parts can be referred to in the description of the above-described embodiments of the lightning strike location method with big data intelligent analysis.

Claims

1. A lightning stroke positioning method with big data intelligent analysis, characterized in that, The method comprises the following steps: Obtain monitoring data and historical lightning data, real-time meteorological data, power grid operation data, geographic information data, power grid equipment parameter data, and historical power outage correlation data, wherein the monitoring data includes lightning current data processed by a single-chip microcomputer, electric field intensity change data, monitoring terminal position data, and microsecond-level clock synchronization time data, and the power grid equipment parameter data includes line insulation level and tower grounding resistance; Preprocess the original lightning signal based on a signal processing algorithm to generate a preprocessed data set; The pre-processed data set is processed, the AI big data model is combined with the neural network architecture to analyze the fusion data in depth, the potential law of lightning stroke is mined to realize lightning stroke event prediction, the power outage research and judgment algorithm and model are run to construct research and judgment rules combined with electric field intensity change data and power grid equipment parameters, real-time power grid operation data is used to verify the model output, the line outage state is judged, and the analysis result is generated, which includes processing the fusion data of the pre-processed data set and historical lightning stroke data, real-time meteorological data, power grid operation data, geographic information data, and historical power outage associated data, mining the potential law of lightning stroke, and generating lightning stroke event prediction results, wherein the fusion data is pre-processed through data cleaning, feature engineering, and data standardization; the electric field intensity change data combined with the power grid equipment parameters are processed based on the power outage research and judgment algorithm to generate the preliminary judgment result of the line outage state, wherein the electric field intensity is calculated based on the formula wherein, is the electric field intensity, which is used to quantify the electric field strength around the high-voltage transmission line; λ is the linear charge density, which represents the amount of charge carried by the transmission line per unit length; is the vacuum permittivity; is the distance parameter, which represents the vertical distance from the calculation point to the transmission line, and the preliminary research and judgment result is determined by whether the electric field intensity change reaches the line outage threshold; the preliminary judgment result of the line outage state is verified and optimized combined with real-time power grid operation data, false positive data is removed, and the target line outage state judgment result is generated, wherein the real-time power grid operation data is used to verify the relevance of the electric field intensity change and the actual operation state of the line, so that the research and judgment result conforms to the actual operation situation of the Angolan power grid; the lightning stroke event prediction result and the line outage state judgment result are integrated to generate a structured analysis result containing lightning prediction time, location, intensity, and line outage risk level. Present and synchronize the analysis results to generate data presentation and synchronization results; Manage system time and models, achieve time reference synchronization through 1PPS second pulse signal+NMEA data analysis, design transmission delay compensation algorithm, configure GPS / GLONASS dual-mode backup, automatically switch to backup source when Beidou signal loses lock to ensure synchronization continuity, periodically trigger calibration maintenance instructions, record synchronization accuracy deviation values, and generate system maintenance data; Process system maintenance data, receive and store solar power supply system and time synchronization system maintenance data, and generate system maintenance data processing results.

2. The method of claim 1, wherein, The method comprises the following steps: Input the original lightning signal into a high-speed, high-precision, and low-distortion instrument amplifier, amplify the original weak signal by the amplifier, and improve the signal-to-noise ratio, wherein the amplifier needs to adapt to the signal characteristics after the sensor sensitivity is improved by 200%, and the original lightning signal includes lightning current signals within 20000A range and electric field intensity signals converted by capacitive principle; Input the lightning signal amplified by the instrument amplifier into a signal noise reduction module, run wavelet transform denoising algorithm and adaptive filtering algorithm, remove noise components containing power grid harmonics and surrounding electromagnetic interference, extract clear lightning characteristic signals, and adapt to different environments such as high electromagnetic interference in Angola tropical rainforest and signal attenuation in southern semiarid grassland; Input the lightning signal after noise reduction into an abnormal value processing module, identify abnormal data caused by sensor drift and transmission packet loss by using 3σ criterion, correct the abnormal values by interpolation repair method, and ensure signal data integrity; Input the lightning signal after abnormal value correction into a time synchronization module, combine the microsecond-level clock synchronization time data of Beidou time service, use time axis alignment algorithm to unify the time granularity of lightning current signals, electric field intensity signals, and associated meteorological / power grid data, and realize multi-source signal space-time alignment; Input the lightning signal after space-time alignment into a high-speed low-power single-chip microcomputer, convert and preliminarily integrate the signal by the single-chip microcomputer to generate structured signals containing purified lightning current waveform, quantized electric field intensity data, and terminal position data; Input the structured signal processed by the single-chip microcomputer into a data verification module, remove invalid signals according to signal quality constraint conditions, and generate a preprocessed data set.

3. The method of claim 1, wherein, The system time and model are managed, the time reference synchronization is realized through 1PPS second pulse signal+NMEA data analysis, and a transmission delay compensation algorithm is designed; a GPS / GLONASS dual-mode backup is configured, and when the Beidou signal is lost, the backup source is automatically switched to, thereby ensuring the synchronization continuity; Periodic calibration and maintenance instructions are triggered, the synchronization accuracy deviation value is recorded, and system maintenance data is generated, including: The system time is managed by using the Beidou time service program, the time reference synchronization of each monitoring terminal and the monitoring center is realized through 1PPS second pulse signal and NMEA data analysis, a transmission delay compensation algorithm is designed to calibrate the delay error in the signal transmission process, and an initial time synchronization result is generated, wherein the 1PPS second pulse signal provides an accurate time trigger point, the NMEA data analysis obtains position and time information, and the two ensure the basic accuracy of time synchronization, and the transmission delay compensation algorithm dynamically adjusts the compensation value according to the signal transmission distance difference under the complex terrain of Angola; The Beidou time service signal state is monitored in real time, a GPS / GLONASS dual-mode backup system is configured, and when the Beidou signal is lost, the backup source is automatically switched to, thereby generating a time synchronization continuity guarantee result, wherein the dual-mode backup switching response time needs to be controlled within milliseconds to ensure that the microsecond-level clock synchronization accuracy is not affected by signal switching; The system time and model calibration and maintenance instructions are triggered according to the preset period, the accuracy of the time synchronization system is checked, the synchronization accuracy deviation value is recorded, and the performance of the AI big data model is evaluated, thereby generating system maintenance data including time synchronization deviation data and model performance evaluation data, wherein the calibration and maintenance need to be carried out in combination with the low load period of the Angola power grid to avoid affecting the normal monitoring and judgment functions.

4. The method of claim 3, wherein, The system maintenance data is processed, the maintenance data of the solar power supply system and the time synchronization system is received and stored, and system maintenance data processing results are generated, including: The maintenance data of the solar power supply system and the time synchronization system is received and stored, and an initial maintenance data set is generated, wherein the solar power supply system maintenance data includes solar panel cleaning records, lithium iron phosphate battery performance parameters, and MPPT charging circuit working status, and the time synchronization system maintenance data includes Beidou / GPS / GLONASS dual-mode switching records, synchronization accuracy deviation values, and transmission delay compensation algorithm calibration results; The initial maintenance data set is classified and processed, a data index is constructed according to the equipment type-maintenance time-fault level dimension, and structured maintenance data is generated; The structured maintenance data is analyzed for abnormalities, maintenance data abnormality determination rules are established in combination with the environmental characteristics of Angola, and a maintenance data abnormality report is generated; The structured maintenance data and the maintenance data abnormality report are processed, and system maintenance data processing results are generated.

5. A lightning stroke positioning device with big data intelligent analysis, characterized in that, The device is used to implement the method of claim 1, and the device comprises: The acquisition module is used for acquiring monitoring data and historical lightning stroke data, real-time meteorological data, power grid operation data, geographic information data, power grid equipment parameter data, and historical power outage correlation data. The monitoring data includes lightning current data processed by a single-chip microcomputer, electric field intensity change data, monitoring terminal position data, and microsecond-level clock synchronization time data. The power grid equipment parameter data includes line insulation level and tower grounding resistance. The processing module is used for preprocessing the original lightning stroke signal based on a signal processing algorithm to generate a preprocessed data set; processing the preprocessed data set, using an AI big data model combined with a neural network architecture to perform deep analysis on the fused data, mining potential rules of lightning stroke occurrence to realize lightning stroke event prediction, running a power outage research algorithm and model to construct research rules in combination with electric field intensity change data and power grid equipment parameters, using real-time power grid operation data to verify model output, judging line outage state, and generating analysis results; presenting and synchronously processing the analysis results to generate data presentation and synchronization results; managing system time and model, realizing time reference synchronization through a 1PPS second pulse signal+NMEA data analysis, designing a transmission delay compensation algorithm; configuring a GPS / GLONASS dual-mode backup, automatically switching to a backup source when a Beidou signal loses lock to ensure synchronization continuity; periodically triggering calibration maintenance instructions, recording synchronization accuracy deviation values, and generating system maintenance data; processing the system maintenance data, receiving and storing solar power supply system and time synchronization system maintenance data, and generating system maintenance data processing results.

6. An electronic device, comprising: It comprises: a first processor; and a memory for storing executable instructions of the first processor; wherein the first processor is configured to execute the executable instructions to perform the lightning stroke positioning method with big data intelligent analysis according to any one of claims 1-4. ​

Images