Lightning protection monitoring system and method based on big data analysis

Abstract

The invention discloses a lightning protection monitoring system and method based on big data analysis, and relates to the technical field of lightning protection safety. Comprising a parameter acquisition module, a parameter processing module, a dynamic adjustment module, an update verification module and an iterative optimization module. Through multi-dimensional data acquisition and standardization processing, a dynamic early warning threshold is constructed in combination with a parameter correlation degree and a historical sample, the limitation of a traditional fixed threshold is broken through, and environmental fluctuation is adapted to improve lightning early warning accuracy; by means of a double-track mechanism of real-time sample updating, dynamic safety factor adjustment and regular iterative weight optimization, self-evolution of the model is achieved, stiffness is avoided, the false alarm rate is reduced, and long-term stability and reliability of the system are guaranteed; and meanwhile, big data is taken as a core, so that the timeliness and accuracy of risk pre-judgment are improved, false starting of equipment is reduced, safety protection and operation and maintenance cost control are considered, and the method is adaptive to various complex outdoor lightning protection scenes.

Description

Technical Field

[0001] This invention relates to the field of lightning protection safety technology, specifically to a lightning protection monitoring system and method based on big data analysis. Background Technology

[0002] Lightning disasters, as one of the most common natural disasters, pose a serious threat to the safe production and equipment operation of power systems, communication base stations, construction projects, aerospace, and other fields. With the rapid development of modern industry and information technology, the application of various electronic devices and precision instruments is becoming increasingly widespread, significantly increasing their sensitivity to lightning interference. Traditional lightning protection measures are no longer sufficient to meet the precise protection needs in complex environments, necessitating the construction of an efficient and intelligent lightning protection monitoring and early warning system.

[0003] Current mainstream lightning protection monitoring systems mostly rely on fixed threshold early warning mechanisms. These systems collect single or a few environmental parameters, such as atmospheric electric field strength, through sensors, and trigger an early warning when the parameters reach a preset fixed threshold. However, these systems have significant limitations: First, lightning occurrence is closely related to multiple environmental parameters, such as atmospheric electric field strength, ambient humidity, air pressure, and electromagnetic interference. Monitoring a single parameter cannot fully reflect the complex causes of lightning occurrence, easily leading to insufficient early warning accuracy. Second, fixed thresholds do not consider the dynamic changes in environmental parameters. Meteorological conditions vary significantly at different times and in different regions, and static thresholds cannot adapt to complex and changing environments, often resulting in high false alarm rates and a high risk of missed alarms.

[0004] Meanwhile, the parameter weights in existing systems are mostly based on empirical values, lacking quantitative analysis and dynamic adjustment based on historical data, making it difficult to accurately characterize the correlation between each parameter and lightning risk. Furthermore, the utilization efficiency of historical data is low, lacking a closed-loop sample update and model optimization mechanism, resulting in the system's early warning performance failing to continuously improve over time. Although some systems incorporate data processing technology, they lack real-time verification and dynamic correction mechanisms for threshold effectiveness. When environmental conditions change abruptly or data samples accumulate to a certain scale, the accuracy of early warnings decreases significantly, failing to meet the actual needs of high-precision lightning protection monitoring. Summary of the Invention

[0005] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides a lightning protection monitoring system and method based on big data analysis, which solves the problems mentioned in the background section.

[0006] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: a lightning protection monitoring system based on big data analysis, comprising: The parameter acquisition module collects multi-dimensional real-time environmental parameters and lightning-related data through a distributed sensor array. The real-time environmental parameters include the background value of atmospheric electric field intensity, ambient humidity, ambient air pressure, and electromagnetic interference intensity. The lightning-related data is historical data retrieved from the database for the past T specified periods, which includes the background value of atmospheric electric field intensity, ambient humidity, ambient air pressure, electromagnetic interference intensity, and the actual lightning risk results within the specified period. The parameter processing module is used to standardize real-time environmental parameters and lightning-related data. The dynamic adjustment module determines the correlation between various environmental parameters and lightning risk results based on historical data. It selects several samples with the highest similarity to the standardized real-time environmental parameters from the historical data, and then constructs a dynamic early warning threshold based on these samples. The update verification module is used to update the dynamic early warning threshold in real time within a pre-set update cycle; The iterative optimization module is used to periodically retrain the historical data sample set and update the correlation of each parameter.

[0007] As a further aspect of the present invention: the atmospheric electric field intensity background value, ambient humidity, ambient air pressure, and electromagnetic interference intensity in the real-time environmental parameters are respectively denoted as E. t H t P t and I t Where t is the index of the real-time acquisition time; Simultaneously, historical data from the past T specified periods will be retrieved from the database, denoted as the historical sample set S={S1,S2,……S}. T}; Among them, each historical sample S k ={E k H k ,P k ,I k ,R k}, k=1,2,……T,E k H represents the background value of the atmospheric electric field intensity during the specified period. k P represents the ambient humidity during the specified period. k I represents the ambient air pressure during the specified period. k R represents the electromagnetic interference intensity within the specified period. k R represents the actual lightning risk results for the specified period; k A value of 1 indicates that lightning has occurred. k A value of 0 indicates that no lightning occurred.

[0008] As a further aspect of the present invention: the standardization process is achieved through:

[0009] The standardized parameter value X1 is obtained. t ; In the formula, X t X represents the original parameter value at time t; P This is the average value of the parameter in the historical sample set, i.e. ;X max and X min These are the maximum and minimum values ​​of this parameter in the historical sample set; X t Background value of atmospheric electric field intensity E t Ambient humidity H t Ambient air pressure P t Electromagnetic interference intensity I t The variable to be substituted, namely X t ∈[E t H t P t I t ].

[0010] As a further aspect of the present invention, the dynamic early warning threshold is constructed as follows: Step 1: Calculation of the weights of environmental parameters: Based on the historical sample set S, calculate the weights of each environmental parameter relative to the lightning risk result R. k correlation W X X represents the background value of atmospheric electric field intensity, ambient humidity, ambient air pressure, and electromagnetic interference intensity, i.e., X∈[E, H, P, I]. The formula is:

[0011] Where: |X k -X P | represents the absolute value of the deviation between parameter X and the mean in the historical sample; W X ∈[W E W H W P W I That is, the correlation degree W of atmospheric electric field intensity is obtained. E Humidity correlation W H Pressure correlation W P Electromagnetic interference correlation W I And satisfy W E +W H +W P +W I =1; If not satisfied, then proceed as follows: Perform normalization adjustments, W1 X After normalization adjustment, the environmental parameters and the lightning risk result R are... k correlation WX ; Step 2: Screening of historical matching samples: Select the M samples with the highest similarity to the standardized environmental parameters at time t from the historical sample set S. The similarity is Sim. k The calculation formula is:

[0012] in: X1 k For historical sample S k The standardized value of the parameter X; |X1 t -X1 k | represents the standardized deviation between real-time environmental parameters and historical sample parameters; Filter out Sim k The top M samples are denoted as the matching sample set SU={SU1,SU2,……SU}. M}; Step 3, Determining the dynamic early warning threshold: Through: Calculate the statistical values ​​of the atmospheric electric field intensity corresponding to the matching sample set SU, and use them as the dynamic warning threshold Th at time t. t ; in: , which is the average atmospheric electric field intensity in the matched sample set, EU m To match the atmospheric electric field intensity in the sample set; The standard deviation of the atmospheric electric field intensity in the matched sample set; K is the pre-set safety factor.

[0013] As a further aspect of the present invention, the real-time update method for the dynamic early warning threshold is as follows: Within a pre-defined update cycle, environmental parameters at time t+Δt are re-acquired every Δt and denoted as {E}. t+Δt H t+Δt ,P t+Δt ,I t+Δt}; Then, repeat steps 2 and 3 to obtain the new dynamic early warning threshold Th. t+Δt ; Add the actual lightning risk result at time t to the historical sample set S and update the sample set size; If the sample set exceeds the maximum storage size T max If so, then delete the earliest sample; Simultaneously, the system will verify the effectiveness of the dynamic early warning threshold, with the verification metric being the "false alarm rate F," calculated using the following formula:

[0014] Where: N falseN represents the number of times a lightning strike did not occur after a warning was issued. alert Total number of warnings; If F exceeds the preset threshold F0, then proceed as follows: The new safety factor K is obtained. new ; This process continues until F returns to the preset range; Where e is a preset value.

[0015] As a further aspect of the present invention, the update method for the correlation degree of each parameter is as follows: Retrieve historical data sample set S within a specified time period L ={S1,S2,……S L×g}; The specified time period contains L standard time periods, and each standard time period contains g sub-time periods; pass:

[0016] Recalculate the correlation degree W0 of each parameter. X ; Where X P，L This is the average value of the parameter over L standard time periods; Then use the new correlation degree W0 X Replace the original weight W X This enables iterative optimization of the model.

[0017] A lightning protection monitoring method based on big data analysis, implemented through a lightning protection monitoring system based on big data analysis, includes the following steps: Step 1, Parameter Acquisition: Collect multi-dimensional real-time environmental parameters and lightning-related data through a distributed sensor array; Step 2, Parameter Processing: Standardize the real-time environmental parameters and lightning-related data. The third step, dynamic adjustment, is used to construct dynamic early warning thresholds, namely atmospheric electric field intensity early warning thresholds. Step 4: Update and verify: Update the dynamic early warning threshold in real time; Step 5: Iterative optimization: Periodically retrain the historical data sample set and update the correlation of each parameter.

[0018] (III) Beneficial Effects This invention provides a lightning protection monitoring system and method based on big data analysis. Compared with existing technologies, it has the following advantages: This invention overcomes the limitations of traditional lightning protection monitoring, which relies on "single parameters and fixed thresholds," by collecting multi-dimensional environmental and lightning data through a distributed sensor array and standardizing it using historical samples. The dynamic warning threshold, combined with parameter correlation and statistical values ​​from historical matching samples, adapts to the electric field fluctuation characteristics under different environmental conditions, significantly improving the accuracy of lightning warnings, reducing warning deviations caused by environmental differences, and making lightning protection responses more closely aligned with actual scenario changes.

[0019] This invention introduces an update verification and iterative optimization mechanism. It reduces the false alarm rate by updating the sample set in real time and dynamically adjusting the safety coefficient, while periodically retraining historical data to optimize parameter correlation, thus achieving model self-evolution. This dual-track optimization mode of "real-time update + periodic iteration" allows lightning protection monitoring capabilities to continuously improve with data accumulation, effectively avoiding the problem of model stagnation during long-term operation and ensuring the stability and reliability of the system in different time periods and regions.

[0020] This invention, compared to traditional lightning protection systems, uses big data analysis as its core, fully transforming the value of historical data into dynamic decision-making basis, while precisely adjusting early warning strategies through a false alarm rate verification mechanism. This not only improves the timeliness and accuracy of lightning risk prediction, but also reduces unnecessary false activation of lightning protection equipment, lowers equipment wear and maintenance costs, and balances safety protection with operational efficiency, making it suitable for lightning protection needs in various complex outdoor scenarios. Attached Figure Description

[0021] Figure 1 This is a system block diagram of a lightning protection monitoring system based on big data analysis according to the present invention.

[0022] Figure 2 This is a flowchart illustrating a lightning protection monitoring method based on big data analysis according to the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0024] Please see Figure 1 and Figure 2 As shown, the embodiments of the present invention provide the following technical solutions: As an embodiment of the present invention: This invention relates to a lightning protection monitoring system and method based on big data analysis, comprising: The parameter acquisition module collects multi-dimensional real-time environmental parameters and lightning-related data through a distributed sensor array; The specific targets for data collection include: Real-time environmental parameters, including: Atmospheric electric field intensity background value E t The atmospheric electric field intensity detected by the sensor at time t; Ambient humidity H t The relative humidity at time t; Ambient air pressure P t Atmospheric pressure at time t; Electromagnetic interference intensity I t The intensity of electromagnetic interference signals within the monitored area at time t.

[0025] Lightning-related data: Retrieve historical data from the database for the past T specified periods, denoted as the historical sample set S = {S1, S2, ..., S...} T}; Among them, each historical sample S k ={E k H k ,P k ,I k ,R k}, k=1,2,……T,E k H represents the background value of the atmospheric electric field intensity during the specified period. k P represents the ambient humidity during the specified period. k I represents the ambient air pressure during the specified period. k R represents the electromagnetic interference intensity within the specified period. k R represents the actual lightning risk results for the specified period; k A value of 1 indicates that lightning has occurred. k A value of 0 indicates that no lightning occurred; In this embodiment, each specified period is 1 hour; The parameter processing module is used to standardize real-time environmental parameters and historical environmental samples. Standardization process passed:

[0026] The standardized parameter value X1 is obtained. t ; In the formula, X t X represents the original parameter value at time t; P This is the average value of the parameter in the historical sample set, i.e. ;X max and X min These are the maximum and minimum values ​​of this parameter in the historical sample set; X t Background value of atmospheric electric field intensity E t Ambient humidity H t Ambient air pressure P t Electromagnetic interference intensity I t The variable to be substituted, namely X t ∈[E t H t P t I t ]; For example, the real-time atmospheric electric field strength E t The standardization results are as follows:

[0027] Among them, the average real-time atmospheric electric field intensity in the historical sample set, E max and E min These represent the maximum and minimum values ​​of atmospheric electric field intensity in the historical samples, respectively. The dynamic adjustment module is used to construct the dynamic early warning threshold Th. t That is, the atmospheric electric field intensity warning threshold at time t; The specific steps are as follows: Step 1: Calculation of the weights of environmental parameters: Based on the historical sample set S, the correlation W between each environmental parameter and the lightning risk result Rk is calculated. X ; X represents the background value of atmospheric electric field intensity, ambient humidity, ambient air pressure, and electromagnetic interference intensity as input variables, i.e., X∈[E, H, P, I]. The formula is:

[0028] in: |X k -X P | represents the absolute value of the deviation between parameter X and the mean in the historical samples; The numerator represents the cumulative value of "parameter deviation and lightning risk occurring simultaneously", and the denominator represents the total cumulative value of parameter deviation; In this embodiment, W X The value range of W is [0,1]. X The larger the value, the higher the correlation between the parameter and lightning risk; W X ∈[W E W H W P W I That is, the correlation degree W of atmospheric electric field intensity is obtained. E Humidity correlation W H Pressure correlation W P Electromagnetic interference correlation WI And satisfy W E +W H +W P +W I =1; If not satisfied, then proceed as follows: Perform normalization adjustments, W1 X After normalization adjustment, the environmental parameters and the lightning risk result R are... k correlation W X ; Step 2: Screening of historical matching samples: From the historical sample set S, M samples with the highest similarity to the standardized environmental parameters at time t are selected. In this embodiment, M is a value between 5 and 10, which can be adjusted according to the database size. The similarity Sim k The calculation formula is:

[0029] in: X1 k For historical sample S k The standardized value of the parameter X; |X1 t -X1 k | represents the standardized deviation between real-time environmental parameters and historical sample parameters; In this embodiment, Sim k The value range is [0,1], and the larger the value, the higher the similarity between the historical sample and the current environment; Filter out Sim k The top M samples are denoted as the matching sample set SU={SU1,SU2,……SU}. M}; Step 3: Determining the dynamic early warning threshold: pass:

[0030] Calculate the statistical values ​​of the atmospheric electric field intensity corresponding to the matching sample set SU, and use them as the dynamic warning threshold Th at time t. t ; in: , which is the average value of the atmospheric electric field intensity in the matched sample set; The standard deviation of the atmospheric electric field intensity in the matched sample set; K is a pre-set safety factor, ranging from 1.2 to 1.5. In this embodiment, it is adjusted according to the false alarm tolerance of the actual scenario, and a larger value is taken if the false alarm tolerance is low. This embodiment collects multi-dimensional environmental parameters and lightning-related data through a distributed sensor array, and standardizes the parameters by combining them with historical sample sets to ensure data comparability and accuracy. The impact weight of each parameter on lightning risk is quantified through correlation calculation, accurately focusing on key influencing factors. Historical samples are matched based on similarity, and the warning threshold is dynamically determined by combining statistical values ​​and a safety factor, avoiding the limitations of fixed thresholds. This scheme achieves personalized adaptation of warning thresholds, considering both the comprehensive impact of environmental parameters and the tolerance for false alarms in different scenarios through a safety factor. This significantly improves the accuracy and reliability of lightning warnings, providing a scientific basis for lightning protection decisions in various scenarios and reducing losses caused by lightning disasters.

[0031] As a second embodiment of the present invention: In specific implementation, compared with Embodiment 1, the technical solution of this embodiment differs from that of Embodiment 1 only in that this embodiment further includes: Update the verification module to update the dynamic early warning threshold in real time; In this embodiment, the dynamic warning threshold changes with time and environmental parameters, and the update cycle is Δt = 5 minutes. The specific process is as follows: Within a pre-defined update cycle, environmental parameters at time t+Δt are re-acquired every Δt and denoted as {E}. t+Δt H t+Δt ,P t+Δt ,I t+Δt}; Then, repeat steps 2 and 3 to obtain the new dynamic early warning threshold Th. t+Δt ; Add the actual lightning risk result (1 or 0) at time t to the historical sample set S and update the sample set size; If the sample set exceeds the maximum storage size T max If so, then delete the earliest sample; Simultaneously, the system will verify the effectiveness of the dynamic early warning threshold, with the verification metric being the "false alarm rate F," calculated using the following formula:

[0032] in: N false This represents the number of times no lightning strikes occurred after a warning was issued. N alert Total number of warnings; If F exceeds the preset threshold F0, then proceed as follows:

[0033] A new safety factor K is obtained. new ; This process continues until F returns to the preset range; Where e is a preset value, and in this embodiment, e is 5; This embodiment adds an update and verification module to the first embodiment. It updates environmental parameters and recalculates the dynamic warning threshold in real time every 5 minutes, ensuring the threshold keeps pace with environmental changes and improving the timeliness and dynamic adaptability of the warning. Simultaneously, actual lightning risk results are continuously added to the historical sample set, enabling dynamic optimization of the sample library and enhancing the model's adaptability to environmental changes. The false alarm rate is used to verify the warning effect in real time. When the false alarm rate exceeds a preset range, the safety factor is automatically adjusted, forming a closed-loop optimization mechanism of "collection-calculation-verification-adjustment." This effectively controls the false alarm rate, avoiding safety hazards caused by missed alarms and reducing resource waste from false alarms, further improving the practicality and stability of the lightning protection monitoring system.

[0034] As an embodiment of the present invention: In its specific implementation, compared to Embodiment 1 and Embodiment 2, the technical solution of this embodiment is to combine the solutions of Embodiment 1 and Embodiment 2. The difference between the technical solution of this embodiment and Embodiment 1 and Embodiment 2 lies only in that this embodiment further includes: The iterative optimization module is used to periodically retrain on historical data samples and update the correlation W of each parameter. X ; The specific process is as follows: Retrieve historical data sample set S within a specified time period L ={S1,S2,……S L×g}; The specified time period contains L standard time periods, and each standard time period contains g sub-time periods; In this embodiment, it is assumed that 30 days is the time period of the specified period, each standard period is set as 1 day, that is, 30 standard periods, and each day is divided into 24 sub-periods, that is, each hour is a sub-period. Then the data collected in each hour in the historical data sample set is used as the sample. pass:

[0035] Recalculate the correlation degree W0 of each parameter. X ; Where X P，L This is the average value of the parameter over L standard time periods; Then use the new correlation degree W0 X Replace the original weight W X This enables iterative optimization of the model; This embodiment integrates the core advantages of Embodiments 1 and 2, adding an iterative optimization module to retrain the system by retrieving historical data samples over a specified 30-day period. By statistically averaging parameters over different time periods, the correlation between each environmental parameter and lightning risk is recalculated, enabling dynamic updates of weights and avoiding warning biases caused by long-term use of fixed weights. This solution fully leverages the temporal value of historical data, allowing parameter weights to adapt to environmental changes over different periods, further optimizing the calculation accuracy of dynamic warning thresholds. Simultaneously, it continues the real-time update and false alarm rate verification mechanism, forming a dual optimization system of "real-time adjustment + periodic iteration," significantly improving the system's adaptability to complex and changing environments, making lightning warnings more relevant to real-world scenarios, providing technical support for long-term stable operation, and broadening its applicability.

[0036] As an embodiment of the present invention: In specific implementation, compared with Embodiment 1, Embodiment 2 and Embodiment 3, the technical solution of this embodiment is to combine the solutions of Embodiment 1, Embodiment 2 and Embodiment 3.

[0037] This embodiment comprehensively integrates the technical solutions of the previous three embodiments, combining multi-dimensional data collection, standardized processing, dynamic early warning threshold calculation, real-time update verification, and regular iterative optimization into a complete end-to-end lightning protection monitoring system. Distributed collection ensures data comprehensiveness, standardized processing guarantees data validity, correlation calculation and similarity matching improve threshold targeting, real-time updates ensure timeliness, false alarm rate verification controls early warning quality, and regular iterative optimization enhances long-term adaptability. This solution achieves closed-loop management across the entire chain of data processing, threshold calculation, and model optimization, possessing both the flexibility to respond to environmental changes in real time and the sustainability of long-term iterative upgrades. It maximizes the accuracy, timeliness, and stability of lightning warnings, making it suitable for various complex scenarios with stringent lightning protection requirements, and providing comprehensive and highly reliable technical support for lightning disaster protection.

[0038] A lightning protection monitoring method based on big data analysis, implemented through a lightning protection monitoring system based on big data analysis, includes the following steps: Step 1, Parameter Acquisition: Collect multi-dimensional real-time environmental parameters and lightning-related data through a distributed sensor array; Step 2, Parameter Processing: Standardize the real-time environmental parameters and lightning-related data. The third step, dynamic adjustment, is used to construct dynamic early warning thresholds, namely atmospheric electric field intensity early warning thresholds. Step 4: Update and verify: Update the dynamic early warning threshold in real time; Step 5: Iterative optimization: Periodically retrain the historical data sample set and update the correlation of each parameter.

[0039] It should be stated that all user data collected in this application was collected with the user's consent and authorization, and the use of user data is legal and compliant, and the use and processing of user data comply with the relevant laws, regulations and standards of the relevant regions.

[0040] Furthermore, any content not described in detail in this specification is existing technology known to those skilled in the art.

[0041] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0042] The above formulas are all dimensionless calculations. Dimensionless calculation involves introducing a reference benchmark, such as the maximum, minimum, standard deviation, or theoretical extreme value of a physical quantity, to transform the original physical quantity into a dimensionless relative value. This value is usually mapped to a specific interval, such as [0,1] or [-1,1], which eliminates the influence of units while preserving the relative size relationship of the physical quantities. The formula is derived from software simulation based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0043] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

[0044] The various embodiments of this application have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A lightning protection monitoring system based on big data analysis, characterized in that, include: The parameter acquisition module collects multi-dimensional real-time environmental parameters and lightning-related data through a distributed sensor array; Real-time environmental parameters include atmospheric electric field intensity background value, ambient humidity, ambient air pressure and electromagnetic interference intensity; lightning-related data are historical data retrieved from the database for the past T specified periods, which include atmospheric electric field intensity background value, ambient humidity, ambient air pressure and electromagnetic interference intensity, as well as the actual lightning risk results within the specified period. The parameter processing module is used to standardize real-time environmental parameters and lightning-related data. The dynamic adjustment module determines the correlation between various environmental parameters and lightning risk results based on historical data. It selects the M samples with the highest similarity to the standardized real-time environmental parameters from the historical data, and then constructs a dynamic early warning threshold based on these M samples. The update verification module is used to update the dynamic early warning threshold in real time within a pre-set update cycle; The iterative optimization module is used to periodically retrain the historical data sample set and update the correlation of each parameter.

2. The lightning protection monitoring system based on big data analysis according to claim 1, characterized in that: Let E represent the background value of atmospheric electric field intensity, ambient humidity, ambient air pressure, and electromagnetic interference intensity in the real-time environmental parameters. t H t P t and I t Where t is the index of the real-time acquisition time; Simultaneously, historical data from the past T specified periods will be retrieved from the database, denoted as the historical sample set S={S1,S2,……S}. T }; Among them, each historical sample S k ={E k H k ,P k ,I k ,R k }, k=1,2,……T,E k H represents the background value of the atmospheric electric field intensity during the specified period. k P represents the ambient humidity during the specified period. k I represents the ambient air pressure during the specified period. k R represents the electromagnetic interference intensity within the specified period. k R represents the actual lightning risk results for the specified period; k A value of 1 indicates that lightning has occurred. k A value of 0 indicates that no lightning occurred.

3. The lightning protection monitoring system based on big data analysis according to claim 2, characterized in that: The standardized processing method is as follows: The standardized parameter value is obtained by subtracting the average value of the parameter in the historical sample set from the real-time raw parameter value, and then dividing by the difference between the maximum value and the minimum value of the parameter in the historical sample set. The real-time raw parameter values ​​are any one of the following: atmospheric electric field intensity background value, ambient humidity, ambient air pressure, and electromagnetic interference intensity.

4. The lightning protection monitoring system based on big data analysis according to claim 3, characterized in that: The determination method is as follows: Based on the historical sample set S, multiply the absolute value of the deviation of the environmental parameter from the sample mean in all historical samples by the sum of the corresponding lightning risk results, and then divide by the sum of the absolute values ​​of the deviation of that parameter from the sample mean in all historical samples; calculate the relationship between each environmental parameter and the lightning risk result R. k correlation W X ; Where X represents the background value of atmospheric electric field intensity, ambient humidity, ambient air pressure, and electromagnetic interference intensity, i.e., X∈[E, H, P, I]; W X ∈[W E W H W P W I That is, the correlation degree W of atmospheric electric field intensity is obtained. E Humidity correlation W H Pressure correlation W P Electromagnetic interference correlation W I .

5. The lightning protection monitoring system based on big data analysis according to claim 4, characterized in that: The similarity is determined as follows: Select the M samples from the historical sample set S that have the highest similarity to the standardized environmental parameters at time t; Then subtract 1 the sum of the absolute values ​​of the deviations between the real-time standardized parameter values ​​and the standardized parameter values ​​of the historical samples; Among them, the environmental parameters include the background value of atmospheric electric field intensity, ambient humidity, ambient air pressure, and electromagnetic interference intensity, and the correlation degree of each item is the correlation degree of the corresponding parameters that has been determined. Then select Sim k The top M samples are denoted as the matching sample set SU={SU1,SU2,……SU}. M } 6. The lightning protection monitoring system based on big data analysis according to claim 5, characterized in that: The dynamic early warning threshold is calculated as follows: The pre-set safety factor is multiplied by the standard deviation of the atmospheric electric field intensity in the matching sample set, and then the average atmospheric electric field intensity in the matching sample set is added to calculate the statistical value of the corresponding atmospheric electric field intensity value in the matching sample set SU. This statistical value is used as the dynamic warning threshold Th at time t. t .

7. The lightning protection monitoring system based on big data analysis according to claim 4, characterized in that: in, The final correlation coefficients for atmospheric electric field intensity, humidity, air pressure, and electromagnetic interference all sum to 1. If the condition is not met, normalization is applied to make the sum of the correlations of each item equal to 1. The adjusted correlation is the final correlation used for calculation.

8. A lightning protection monitoring system based on big data analysis according to claim 6, characterized in that: The dynamic early warning threshold is updated in real time as follows: Within a pre-defined update cycle, environmental parameters at time t+Δt are re-acquired every Δt and denoted as {E}. t+Δt H t+Δt ,P t+Δt ,I t+Δt }; The new dynamic early warning threshold Th is obtained by calculating the dynamic early warning threshold accordingly. t+Δt ; Add the actual lightning risk result at time t to the historical sample set S and update the sample set size; If the sample set exceeds the maximum storage size T max If so, then delete the earliest sample; At the same time, the system will verify the effectiveness of the dynamic early warning threshold. The verification index is the false alarm rate F, where the false alarm rate is equal to the number of times that lightning did not occur after the early warning divided by the total number of early warnings. If F exceeds the preset threshold F0, then proceed as follows: The new safety factor K is obtained. new Where e is a preset value; This continues until F returns to the preset range.

9. A lightning protection monitoring system based on big data analysis according to claim 4, characterized in that: The update method for the correlation of each parameter is as follows: First, retrieve the historical data sample set S within the specified past time period. L ={S1,S2,……S L×g }; where the specified time period contains L standard time periods, and each standard time period contains g sub-time periods; Next, the absolute value of the deviation between the parameter and its average value in all historical samples within the specified time period is multiplied by the sum of the lightning risk results for the corresponding samples, and then divided by the sum of the absolute values ​​of the deviation between the parameter and its average value in all historical samples within the specified time period; a new correlation degree for each parameter is obtained and denoted as W0. X ; The specified time period contains L standard time periods, each standard time period contains g sub-time periods, and the average value of the parameter within the specified time period is the arithmetic mean of the parameter within the L standard time periods. Then use the new correlation degree W0 X Replace the original weight W X This enables iterative optimization of the model.

10. A lightning protection monitoring method based on big data analysis, characterized in that, This method is implemented using a lightning protection monitoring system based on big data analysis as described in any one of claims 1-9.

Images