Ground-to-ground lightning activity path electrical parameter measurement system based on space-time clustering analysis

Abstract

The invention discloses a measurement system for electrical parameters of a ground-to-ground lightning activity path based on spatial-temporal clustering analysis, and relates to the technical field of measurement of the electrical parameters of the ground-to-ground lightning activity path, and the system comprises a data collection module which is used for collecting ground-to-ground lightning data, and the ground-to-ground lightning data comprises the time, longitude and latitude information of the occurrence of the ground-to-ground lightning; the electrical parameter measurement module is used for measuring electrical parameters on the ground lightning arrester activity path, and the electrical parameters comprise electric field intensity, current intensity and voltage; a sensor monitors an area on a ground-to-ground lightning activity path in real time to obtain accurate electrical parameter data, the sensor is used for measuring electric field intensity, a current transformer is used for measuring current intensity, and a voltage sensor is used for measuring voltage, so that the problem of poor electrical parameter measurement accuracy is solved, the measurement accuracy is greatly improved, and the measurement cost is reduced. When the heavy current at the moment of ground lightning is measured, the current transformer can accurately record the change of the current intensity by virtue of a proper measuring range and quick response, thereby providing reliable data for researching the current characteristics of the ground lightning.

Description

Technical Field

[0001] The present invention relates to the technical field of measuring electrical parameters of the path of cloud-to-ground lightning activity, and specifically provides a system for measuring electrical parameters of the path of cloud-to-ground lightning activity based on spatio-temporal clustering analysis. Background Art

[0002] In lightning research, it is crucial to accurately measure the electrical parameters on the path of cloud-to-ground lightning activity. As a powerful natural discharge phenomenon, cloud-to-ground lightning, with electrical parameters such as electric field intensity, current intensity, and voltage it carries, plays a key role in deeply understanding the physical mechanism of lightning and evaluating the impact of lightning on various facilities and the environment.

[0003] In this regard, the patent of CN107271793B discloses an automated lightning warning system. Taking a thunderstorm cluster as the lightning warning unit, according to the historical path and time of the movement of the thunderstorm cluster, the next occurrence location and time of the thunderstorm cluster are predicted, which is beneficial to avoiding the cumbersome process of separately calculating the movement of lightning points, simplifying the calculation workload, and improving the prediction efficiency. By setting up a coordinate system and combining it for the statistics of the historical path of the thunderstorm cluster and the prediction of the next occurrence location, it is beneficial to improve the accuracy and intuitiveness of the predicted path and location.

[0004] However, there are obvious deficiencies in the existing technologies. Some electric field intensity measurement methods are vulnerable to environmental interference, and it is difficult to guarantee the accuracy of measurement results. External factors will seriously affect the performance of measurement sensors, resulting in large deviations in measurement data. In terms of current intensity measurement, when traditional equipment measures large currents, it may be unable to accurately capture the current changes at the moment of cloud-to-ground lightning due to range limitations or insufficient response speed. For voltage measurement, due to the instantaneousness and complexity of cloud-to-ground lightning voltage, existing voltage measurement means are difficult to accurately measure its peak value and change process.

[0005] To address the above problems, a system for measuring electrical parameters of the path of cloud-to-ground lightning activity based on spatio-temporal clustering analysis is proposed. Summary of the Invention

[0006] The purpose of the present invention is to provide a system for measuring electrical parameters of the path of cloud-to-ground lightning activity based on spatio-temporal clustering analysis, which solves the problem of poor accuracy in measuring electrical parameters in the background art.

[0007] To achieve the above purpose, the present invention provides the following technical solution: A system for measuring electrical parameters of the path of cloud-to-ground lightning activity based on spatio-temporal clustering analysis, including A data acquisition module for collecting cloud-to-ground lightning data, where the cloud-to-ground lightning data includes the time, longitude, and latitude information of the occurrence of cloud-to-ground lightning; An electrical parameter measurement module for measuring the electrical parameters on the path of cloud-to-ground lightning activity, where the electrical parameters include electric field intensity, current intensity, and voltage; The sensor monitors the area on the path of cloud-to-ground lightning activity in real time, obtains accurate electrical parameter data, measures the electric field strength using a sensor, measures the current strength using a current transformer, and measures the voltage using a voltage sensor; In the electrical parameter measurement module, a data fusion algorithm is used to measure the electric field strength, current strength, and voltage value.

[0008] Preferably, the electrical parameter measurement system further includes a region division module for region division. The region division analyzes the altitude and topographic and geomorphic features to determine the region boundaries and divides the region into three regions: mountainous areas, grasslands, and deserts; The region division module further divides the three regions according to the slope and aspect; The slope region division is divided into extremely steep regions, steep regions, medium slope regions, and gentle slope regions; The spatio-temporal database establishment module is used to construct a three-dimensional spatio-temporal database from the collected cloud-to-ground lightning data; determine that the dimensions of the three-dimensional spatio-temporal database are time, longitude, and latitude, convert each cloud-to-ground lightning data record into an object point form, and enter all object points into the database in chronological order; The spatio-temporal clustering analysis module is used to perform spatio-temporal clustering analysis on the data in the spatio-temporal database, including a threshold setting unit, a core object judgment unit, an adjacent point search unit, and a cluster refinement unit; The threshold setting unit is used to set the time distance threshold, space distance threshold, and spatio-temporal object quantity threshold; Obtain the region feature information from the region division module, extract the cloud-to-ground lightning data features from the spatio-temporal database, and use an adaptive algorithm to adjust the threshold based on the feedback dynamic adjustment algorithm. The adjustment weights are respectively , and the formula is as follows: ; where represents the threshold weight of the time distance for the cloud-to-ground lightning data point, represents the threshold weight of the space distance factor for the cloud-to-ground lightning data point, represents the threshold weight of the number of spatio-temporal objects in the cloud-to-ground lightning data point, represents the starting point of the clustering analysis, and C represents the target value of the spatio-temporal clustering analysis; The core object judgment unit is used to select an object point from all cloud-to-ground lightning data and judge the cloud-to-ground lightning path; The selected object point is: , where is the unique identifier of the object point, x and y are the longitude information and latitude information respectively, and t is the time information; determine whether the object point already belongs to an existing cluster. If it already belongs to an existing cluster, select the next object point; otherwise, determine whether the object point is a spatio-temporal core object. If it is not a spatio-temporal core object, return to the object point selection step to select the next object point; if it is a spatio-temporal core object, perform adjacent point search. An adjacent point search unit for searching all spatio-temporal adjacent points of the spatio-temporal core object point. For a point determined to be a spatio-temporal core object, based on the time, longitude, and latitude information, in the time dimension, if the time difference between another point and the core object point satisfies not being greater than the set time distance threshold; in the space dimension, if the spatial distance between the two points satisfies not being greater than the set spatial distance threshold, then this point is identified as a spatio-temporal adjacent point. A cluster improvement unit for determining whether the cloud-to-ground flash data is a spatio-temporal core object, and judging whether the number of neighbors within a given range is not less than the spatio-temporal object quantity threshold according to the time distance threshold, spatial distance threshold, and spatio-temporal object quantity threshold. The given range is a time period with a time difference not greater than the time distance threshold in the time dimension, and in the space dimension, it is an area with a spatial distance not greater than the spatial distance threshold.

[0009] Preferably, in the data acquisition module, a distributed sensor network is used to collect cloud-to-ground flash data. For the collected data, real-time data preprocessing is performed, including outlier detection and elimination. The method of judging outliers by the 3-sigma rule is adopted, that is: if the difference between the data point and the average value is greater than three times the standard deviation, then it is identified as an outlier and eliminated.

[0010] Preferably, in the mountain area, a slope greater than 60° is an extremely steep area, a slope of 45° - 60° is a steep slope area, a slope of 30° - 45° is a medium slope area; a slope less than 20° is a gentle slope area; the slope aspect division is used to analyze the influence of the vegetation coverage and soil moisture difference of the slope aspect on the cloud-to-ground flash activity path, and the slope aspect is divided into sunny slope and shady slope. In the grassland, the slope division is that a slope greater than 10° is a steep slope area, and a slope less than 10° is a gentle slope area; the slope aspect division is used to analyze the influence of the vegetation coverage and soil moisture difference of the slope aspect on the cloud-to-ground flash activity path, and the slope aspect is divided into sunny slope and shady slope. In the desert, a dune slope greater than 30° is a steep slope area, and a dune slope less than 30° is a gentle slope area; for the slope aspect division, the windward side of the dune is the sunny slope, and the leeward side of the dune is the shady slope, which is used to analyze the influence of the sand compactness and humidity difference of different slope aspects on the cloud-to-ground flash activity path.

[0011] Preferably, in the area division module, for the further division of the mountain area, a terrain undulation index is introduced. The formula for calculating the terrain undulation is , where is the altitude of the highest point in the area, $h_{min}$ is the elevation of the lowest point within the region, $D$ is the difference between the average elevation within the region and the average elevation of the neighborhood, $S$ is the area of the region. According to the terrain undulation degree, the mountainous area is subdivided into high-undulation mountainous areas, medium-undulation mountainous areas, and low-undulation mountainous areas. The lightning flash activity paths in mountainous areas with different undulation degrees may be affected by the terrain to different extents.

[0012] Preferably, in the core object judgment unit, for the judgment of the lightning flash path, time series analysis is added. The autoregressive moving average model is adopted to model and predict the time series data. If the deviation between the predicted value and the actual value is within a certain range, it is considered that this point may belong to the same lightning flash path. Let the time series of the selected object point be The model formula is: where is the autoregressive coefficient, is the moving average coefficient, is the white noise sequence.

[0013] Preferably, in the adjacent point search unit, the K-D tree data structure is adopted for spatial proximity point search. For the given spatio-temporal core object point, adjacent points within the range of the spatial distance threshold are found in the K-D tree.

[0014] Preferably, in the cluster refinement unit, the fuzzy clustering algorithm is introduced. For the newly added object point, its belonging is judged according to its fuzzy membership degree to the existing cluster center. Let the object point be and the existing cluster center be The fuzzy membership degree calculation formula is: where is the model membership degree of the object point belonging to the cluster , is the distance from the object point to the cluster center , is the number of clusters, is the fuzzy index. The object point is determined whether it is a core object and the cluster it belongs to according to the fuzzy membership degree.

[0015] Preferably, in the region division module, the effectiveness of the region division result is evaluated, and the clustering evaluation index coefficient is adopted for evaluation. The data point is the average distance from the point to other points within the same region. The silhouette coefficient is , where is the silhouette coefficient of the lightning flash data point , is the average distance from the lightning flash data point to the lightning flash data points within the same terrain region. Ground flash data points The average distance to the ground flash data points in the nearest different terrain areas. The average value is obtained according to the silhouette coefficient of all data points, and the evaluation value of the entire area division result is obtained.

[0016] Preferably, for the measured value of electric field strength, its information entropy calculation formula is , where Similarly, the information entropy of the measured values of current intensity and voltage can be obtained, where represents the information entropy of the measured value of electric field strength, is the number of measured values of electric field strength, is the measured value of electric field strength for the th indicating the proportion of the th measured value of electric field strength in the total sum of all measured values, is the measured value of electric field strength for the th , where represents the corresponding weight coefficient of the measured value of electric field strength, is the information entropy of the measured value of electric field strength , is the number of measured values of electric field strength, is the index for summation, from 1 to traverse all measured values of electric field strength to calculate the value of the denominator. The calculation of the weight coefficients for current intensity and voltage is the same. The determined weight coefficients are weighted and averaged for fusion to obtain the fusion values of electric field strength, current intensity, and voltage.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: A ground flash activity path electrical parameter measurement system based on spatio-temporal clustering analysis provided by the present invention measures electric field strength through sensors, current intensity through current transformers, and voltage through voltage sensors. Different types of sensors are adjusted according to the characteristics of the corresponding electrical parameters, and can capture the electric field, current, and voltage signals on the ground flash activity path more accurately. Compared with traditional single or poorly general-purpose measurement devices, the measurement accuracy is greatly improved. When measuring the large current at the moment of ground flash, the current transformer can accurately record the change of current intensity with its appropriate range and fast response, providing reliable data for studying the characteristics of ground flash current. Secondly, the electrical parameter measurement module adopts a data fusion algorithm, combines the information entropy of the measured values of electric field strength, current intensity, and voltage to calculate the weight coefficients, and can effectively eliminate the errors and uncertainties existing in each measured value alone, improving the reliability and stability of the measurement results. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 Schematic diagram of the measurement system process of the present invention; Figure 2 Schematic diagram of the process of the spatio-temporal clustering analysis module of the present invention; Figure 3 Decision block diagram of the core object judgment logic of the present invention. Specific implementation manners

[0019] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0020] To further understand the content of the present invention, the present invention will be described in detail in conjunction with the accompanying drawings.

[0021] Combined with Figures 1 - 3 , a lightning flash activity path electrical parameter measurement system based on spatio-temporal clustering analysis of the present invention includes A data acquisition module for collecting lightning flash data, where the lightning flash data includes the time, longitude, and latitude information of the lightning flash occurrence; In the data acquisition module, a distributed sensor network is used to collect lightning flash data. For the collected data, real-time data preprocessing is performed, including outlier detection and elimination. The method of using the 3-sigma rule is adopted to judge outliers, that is: if the difference between a data point and the average value is greater than three times the standard deviation, it is determined as an outlier and eliminated.

[0022] The problems of incomplete and inaccurate collection of lightning flash data are solved. At the same time, outlier detection and elimination ensure the accuracy of the data, reduce the interference of abnormal data on subsequent analysis, and provide a high-quality data basis for subsequent area division, spatio-temporal clustering analysis, and electrical parameter measurement. Accurate and comprehensive data helps to analyze the lightning flash activity path more precisely, improve the understanding of the laws of lightning activities, and thus provide a more reliable basis for the formulation of lightning protection and disaster warning measures.

[0023] An electrical parameter measurement module for measuring the electrical parameters on the lightning flash activity path, where the electrical parameters include electric field strength, current strength, and voltage; The sensor monitors the area on the lightning flash activity path in real time to obtain accurate electrical parameter data, and uses the sensor to measure the electric field strength, measures the current strength through a current transformer, and measures the voltage using a voltage sensor. In the electrical parameter measurement module, a data fusion algorithm is used to measure the electric field strength, current strength, and voltage values. For the measured value of the electric field strength, its information entropy calculation formula is , where Similarly, the information entropy of the current intensity and voltage measurement values can be obtained, where represents the information entropy of the electric field strength measurement value, is the number of electric field strength measurement values, is the electric field strength measurement value of the th represents the proportion of the th electric field strength measurement value in the total sum of all measurement values. is the electric field strength measurement value corresponding weight coefficient, is the electric field strength measurement value information entropy, is the number of electric field strength measurement values, is the index for summation, ranging from 1 to traversing all electric field strength measurement values to calculate the denominator value. Similarly, the weight coefficients for current intensity and voltage are calculated, and the determined weight coefficients are weighted and averaged for fusion to obtain the fusion values of electric field strength, current intensity, and voltage.

[0024] The electrical parameter measurement system of the present invention further includes: a region division module for region division. The region division analyzes altitude, topographic and geomorphic features to determine the region boundaries and divides the region into three regions: mountainous areas, grasslands, and deserts; The region division module further divides the three regions according to slope and aspect; The slope region division is divided into extremely steep regions, steep regions, moderately sloped regions, and gently sloped regions; Combined with geographic information system technology, using the high-resolution terrain data, vegetation cover data, and soil type information it provides, the region boundaries are further accurately determined and the region division is refined, solving the problem in the prior art that the region division is rough and cannot fully consider the influence of different terrain conditions on cloud-to-ground lightning activities. Through detailed region division, the laws of cloud-to-ground lightning activities and the changes in electrical parameters in different terrain regions can be better studied.

[0025] In mountainous areas, slopes greater than 60° are extremely steep regions, slopes of 45° - 60° are steep regions, slopes of 30° - 45° are moderately sloped regions; slopes less than 20° are gently sloped regions; the aspect division is used to analyze the influence of vegetation cover and soil moisture differences in different aspects on the cloud-to-ground lightning activity path, and the aspect is divided into sunny slopes and shady slopes; In the grassland, the slope is divided as follows: areas with a slope greater than 10° are steep slope regions, and areas with a slope less than 10° are gentle slope regions; the aspect is divided to analyze the influence of vegetation cover and soil moisture differences in different aspects on the path of cloud-to-ground lightning activities. The aspects are divided into sunny slopes and shady slopes; In the desert, areas with a dune slope greater than 30° are steep slope regions, and areas with a dune slope less than 30° are gentle slope regions; for the aspect division, the windward side of the dune is the sunny slope, and the leeward side of the dune is the shady slope, which is used to analyze the influence of the sand compactness and moisture differences in different aspects on the path of cloud-to-ground lightning activities; In the regional division module, for the further division of mountainous areas, a terrain undulation index is introduced. The formula for calculating the terrain undulation is , where is the altitude of the highest point in the region, is the altitude of the lowest point in the region, D is the difference between the average elevation in the region and the average elevation of the neighborhood, S is the area of the region. According to the terrain undulation, mountainous areas are subdivided into high-undulation mountainous areas, medium-undulation mountainous areas, and low-undulation mountainous areas. The path of cloud-to-ground lightning activities in mountainous areas with different undulations may be affected by the terrain to different degrees.

[0026] Analyze the cloud-to-ground lightning return stroke data of the three regions separately. The analysis parameters include the number of cloud-to-ground lightning return strokes, return stroke interval, duration, return stroke spacing, return stroke intensity, and the number of channel grounds. Through comparative analysis, determine the typical characteristics of the cloud-to-ground lightning return stroke data in each region. Use statistical analysis methods to conduct separate statistics and comparisons on the data of each region, and calculate statistical indicators such as the average value and standard deviation of each parameter to determine the typical characteristics; Number of cloud-to-ground lightning return strokes: directly count the total number of cloud-to-ground lightning return strokes in the region, denoted as N; Return stroke interval: calculate the time interval between two adjacent cloud-to-ground lightning return strokes, and then find the average value of all return stroke intervals and standard deviation ; Duration: count the duration of each cloud-to-ground lightning return stroke and find its average value and standard deviation ; Return stroke spacing: measure the spatial distance between two adjacent cloud-to-ground lightning return strokes and calculate the average value and standard deviation ; Return stroke intensity: obtain the intensity value of the cloud-to-ground lightning return stroke through a measuring device and find its average value and standard deviation ; Number of channel grounds: count the number of channel grounds in each cloud-to-ground lightning return stroke and find its average value and standard deviation .

[0027] Compare the parameter statistical indicators of the three regions, create a table to visually display the differences in each parameter among the three regions, and analyze the relationship between the topographic characteristics of different regions and the typical characteristics of cloud-to-ground lightning strike data; Average value calculation formula: For parameter (such as strike interval, duration, etc.), its average value , where is the number of data points, is the value of the -th data point; Standard deviation calculation formula: For parameter , its standard deviation , through statistical analysis steps and algorithm formulas, the typical characteristics of cloud-to-ground lightning strike data in each region can be determined, providing a basis for further studying the laws of cloud-to-ground lightning activities in different topographic regions.

[0028] Space-time database establishment module, used to construct the collected cloud-to-ground lightning data into a three-dimensional space-time database; determine that the dimensions of the three-dimensional space-time database are time, longitude, and latitude, convert each cloud-to-ground lightning data record into an object point form, and enter all object points into the database in chronological order; Space-time clustering analysis module, used to perform space-time clustering analysis on the data in the space-time database, including a threshold setting unit, a core object judgment unit, an adjacent point search unit, and a cluster improvement unit; The threshold setting unit is used to set the time distance threshold, the space distance threshold, and the space-time object quantity threshold; Obtain the regional characteristic information from the regional division module, extract the cloud-to-ground lightning data characteristics from the space-time database, adopt an adaptive algorithm, and adjust the threshold based on the feedback dynamic adjustment algorithm. The adjustment weights are respectively , and the formula is as follows: ; Where represents the threshold weight of the time distance for the cloud-to-ground lightning data point, represents the threshold weight of the space distance factor for the cloud-to-ground lightning data point, represents the threshold weight of the number of space-time objects in the cloud-to-ground lightning data point, represents the starting point of the clustering analysis, and C represents the target value of the space-time clustering analysis; The core object judgment unit is used to select an object point from all cloud-to-ground lightning data and judge the cloud-to-ground lightning path; In the core object determination unit, time series analysis is added to the determination of the cloud-to-ground flash path. The autoregressive moving average model is used to model and predict the time series data. If the deviation between the predicted value and the actual value is within a certain range, it is considered that this point may belong to the same cloud-to-ground flash path. Let the time series of the selected object point be The model formula is: where is the autoregressive coefficient, is the moving average coefficient, is the white noise sequence.

[0029] The selected object point is: where is the unique identifier of the object point, x and y are the longitude information and latitude information respectively, and t is the time information; it is judged whether the object point already belongs to the existing cluster. If it already belongs to the existing cluster, the next object point is selected again; otherwise, it is judged whether the object point is a spatio-temporal core object. If it is not a spatio-temporal core object, return to the object point selection step to select the next object point; if it is a spatio-temporal core object, adjacent point search is performed; The adjacent point search unit is used to search for all spatio-temporal adjacent points of the spatio-temporal core object point. For the point determined to be a spatio-temporal core object, based on the time, longitude, and latitude information, in the time dimension, if the time difference between another point and the core object point satisfies not being greater than the set time distance threshold; in the space dimension, if the spatial distance between the two points satisfies not being greater than the set spatial distance threshold, then this point is identified as a spatio-temporal adjacent point; In the adjacent point search unit, the K-D tree data structure is used for spatial proximity point search. For the given spatio-temporal core object point, adjacent points within the range of the spatial distance threshold are found in the K-D tree.

[0030] The cluster refinement unit is used to judge whether the cloud-to-ground flash data is a spatio-temporal core object. According to the time distance threshold, spatial distance threshold, and spatio-temporal object quantity threshold, it is judged whether the number of neighbors within the given range is not less than the spatio-temporal object quantity threshold. The given range is a time period with a time difference not greater than the time distance threshold in the time dimension, and in the space dimension, it is an area with a spatial distance not greater than the spatial distance threshold. In the cluster refinement unit, the fuzzy clustering algorithm is introduced. For the newly added object point, its belonging is judged according to its fuzzy membership degree to the existing cluster center. Let the object point and the existing cluster center be The fuzzy membership degree calculation formula is: where is the model membership degree of the object point belonging to the cluster , is the object point Distance to the cluster center , is the number of clusters, is the fuzzy exponent, which determines whether an object point is a core object and the cluster it belongs to according to the fuzzy membership degree; In the area division module, the effectiveness of the area division result is evaluated using the clustering evaluation index coefficient. The data point is the point The average distance to other points in the same area. The silhouette coefficient is , where is the silhouette coefficient of the cloud-to-ground flash data point , is the average distance of the cloud-to-ground flash data point to the cloud-to-ground flash data points in the same terrain area, The average distance of the cloud-to-ground flash data point to the cloud-to-ground flash data points in the nearest different terrain area. The average value of the silhouette coefficients of all data points is calculated to obtain the evaluation value of the entire area division result.

[0031] Working principle: First, the data acquisition module uses a distributed sensor network to reasonably layout and collect cloud flash data according to different terrain characteristics, and preprocesses the collected data through the 3σ criterion to eliminate outliers, providing an accurate data basis for subsequent analysis. Next, the regional division module finely divides the region based on factors such as altitude, topographic and geomorphic features, slope, and aspect, combined with geographic information system technology to determine different terrain regions, providing regional characteristic information for spatio-temporal clustering analysis. Then, the spatio-temporal database establishment module constructs a three-dimensional spatio-temporal database from the collected cloud flash data, converts the data into object point form and enters it in chronological order, while performing annotation classification and using data compression technology for efficient storage and retrieval of data, providing data support for spatio-temporal clustering analysis. After that, the spatio-temporal clustering analysis module starts to work. The threshold setting unit obtains information from the regional division module and the spatio-temporal database, and uses an adaptive algorithm combined with machine learning technology to adjust the threshold; the core object judgment unit selects object points from the data and judges whether they belong to the existing clusters. If not, it judges whether they are spatio-temporal core objects, and time series analysis is added during the process. If they are spatio-temporal core objects, the adjacent point search unit uses them as a benchmark to search for adjacent points that meet the conditions in the time and space dimensions using the K-D tree data structure. The cluster refinement unit judges the belonging of newly added object points according to the threshold and combined with the fuzzy clustering algorithm, so as to accurately judge the cloud flash path and core objects. Finally, the electrical parameter measurement module uses sensors to monitor the area on the cloud flash activity path in real time, obtains electrical parameter data, measures different electrical parameters through specific sensors, introduces a data fusion algorithm to improve the measurement accuracy, and combines wireless sensor network technology to achieve real-time remote monitoring and data transmission, providing accurate and timely electrical parameter information for lightning protection and disaster warning; Through comprehensive data acquisition, using a distributed sensor network and performing real-time preprocessing to ensure accurate and reliable data, fine regional division, subdividing the region according to altitude and topographic and geomorphic features, considering the influence of different terrain conditions on the cloud flash activity path and electrical parameters, accurate spatio-temporal clustering analysis, each unit collaboratively uses an adaptive algorithm to adjust the threshold to judge the cloud flash path and core objects; precise electrical parameter measurement, using sensors for real-time monitoring and introducing a data fusion algorithm to improve the accuracy. At the same time, innovations are made in aspects such as regional division and core object judgment, with obvious advantages in accuracy, reliability, and practicality compared with the existing technology.

[0032] It should be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variation thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, article or device.

[0033] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A ground flash activity path electrical parameter measurement system based on spatio-temporal clustering analysis, characterized in that, Including, a data acquisition module for acquiring cloud-to-ground flash data, where the cloud-to-ground flash data includes the time, longitude, and latitude information of the occurrence of the cloud-to-ground flash; an electrical parameter measurement module for measuring the electrical parameters on the path of the cloud-to-ground flash activity, where the electrical parameters include electric field strength, current intensity, and voltage; sensors are used to monitor the area on the path of the cloud-to-ground flash activity in real time, obtain accurate electrical parameter data, measure the electric field strength using sensors, measure the current intensity using a current transformer, and measure the voltage using a voltage sensor; In the electrical parameter measurement module, a data fusion algorithm is used to measure the electric field strength, current intensity, and voltage values.

2. The ground flash activity path electrical parameter measurement system based on spatio-temporal clustering analysis according to claim 1, wherein: The electrical parameter measurement system further includes a region division module for region division. The region division analyzes the altitude and topographic and geomorphic features to determine the region boundaries and divides the region into three regions: mountainous areas, grasslands, and deserts; The region division module further divides the three regions according to the slope and aspect; The slope region division is divided into extremely steep regions, steep regions, moderately sloped regions, and gently sloped regions; a spatio-temporal database establishment module for constructing the acquired cloud-to-ground flash data into a three-dimensional spatio-temporal database; determining that the dimensions of the three-dimensional spatio-temporal database are time, longitude, and latitude, converting each cloud-to-ground flash data record into an object point form, and entering all object points into the database in chronological order; a spatio-temporal clustering analysis module for performing spatio-temporal clustering analysis on the data in the spatio-temporal database, including a threshold setting unit, a core object judgment unit, an adjacent point search unit, and a cluster refinement unit; The threshold setting unit is used to set the time distance threshold, space distance threshold, and spatio-temporal object quantity threshold; Obtain regional feature information from the regional division module, extract cloud-to-ground lightning data features from the spatio-temporal database, and use an adaptive algorithm to adjust the threshold based on the feedback dynamic adjustment algorithm. The adjustment weights are respectively , and the formula is as follows: ; wherein represents the threshold weight of the time distance for the cloud-to-ground flash data points, represents the threshold weight of the spatial distance factor for the cloud-to-ground flash data points, represents the threshold weight of the number of spatio-temporal objects in the cloud-to-ground flash data points, represents the starting point of the clustering analysis, and C represents the target value of the spatio-temporal clustering analysis; The core object judgment unit is used to take an object point from all the cloud-to-ground flash data and judge the cloud-to-ground flash path; The selected object point is: , where is the unique identifier of the object point, x and y are the longitude information and latitude information respectively, and t is the time information; determine whether the object point already belongs to an existing cluster. If it already belongs to an existing cluster, select the next object point again; otherwise, determine whether the object point is a spatio-temporal core object. If it is not a spatio-temporal core object, return to the object point selection step to select the next object point again; if it is a spatio-temporal core object, perform an adjacent point search; The adjacent point search unit is used to search for all spatio-temporal adjacent points of the spatio-temporal core object point. For a point determined to be a spatio-temporal core object, based on the time, longitude, and latitude information, in the time dimension, if the time difference between another point and the core object point satisfies not being greater than the set time distance threshold; in the space dimension, if the spatial distance between the two points satisfies not being greater than the set space distance threshold, then this point is considered a spatio-temporal adjacent point; The cluster refinement unit is used to judge whether the cloud-to-ground flash data is a spatio-temporal core object, and based on the time distance threshold, space distance threshold, and spatio-temporal object quantity threshold, judge whether the number of neighbors within a given range is not less than the spatio-temporal object quantity threshold, where the given range is a time period with a time difference not greater than the time distance threshold in the time dimension and a region with a spatial distance not greater than the space distance threshold in the space dimension.

3. The ground flash activity path electrical parameter measurement system based on spatio-temporal clustering analysis according to claim 1, characterized in that: In the data acquisition module, a distributed sensor network is used to acquire cloud-to-ground flash data, and for the acquired data, real-time data preprocessing is performed, including outlier detection and elimination. The method of judging outliers using the 3-sigma rule is adopted, that is: if the difference between the data point and the average value is greater than three times the standard deviation, then it is determined as an outlier and eliminated.

4. A ground flash activity path electrical parameter measurement system based on spatio-temporal clustering analysis according to claim 2, characterized in that: In mountainous areas, slopes greater than 60° are extremely steep areas, slopes of 45° - 60° are steep slopes, slopes of 30° - 45° are moderately sloped areas; slopes less than 20° are gentle slopes; for aspect division, it is used to analyze the influence of vegetation cover and soil moisture differences in different aspects on the path of cloud-to-ground lightning activity, and the aspects are divided into sunny slopes and shady slopes; In grasslands, slope division is that slopes greater than 10° are steep slopes, and slopes less than 10° are gentle slopes; for aspect division, it is used to analyze the influence of vegetation cover and soil moisture differences in different aspects on the path of cloud-to-ground lightning activity, and the aspects are divided into sunny slopes and shady slopes; In deserts, dune slopes greater than 30° are steep slopes, and dune slopes less than 30° are gentle slopes; for aspect division, the windward side of the dune is the sunny slope, and the leeward side of the dune is the shady slope, which is used to analyze the influence of sand compactness and moisture differences in different aspects on the path of cloud-to-ground lightning activity.

5. A ground flash activity path electrical parameter measurement system based on spatio-temporal clustering analysis according to claim 4, characterized in that: In the area division module, for the further division of mountainous areas, the terrain undulation index is introduced. The formula for calculating the terrain undulation is , where is the altitude of the highest point in the area, is the altitude of the lowest point in the area, D is the difference between the average elevation in the area and the average elevation of the neighborhood, S is the area of the area. According to the terrain undulation, the mountainous areas are subdivided into high-undulation mountainous areas, medium-undulation mountainous areas and low-undulation mountainous areas. The lightning activity paths in mountainous areas with different undulations may be affected by the terrain to different degrees.

6. The lightning activity path electrical parameter measurement system based on spatio-temporal clustering analysis according to claim 2, wherein: In the core object judgment unit, time series analysis is added to the judgment of the cloud-to-ground flash path. The autoregressive moving average model is used to model and predict the time series data. If the deviation between the predicted value and the actual value is within a certain range, it is considered that this point may belong to the same cloud-to-ground flash path. Let the time series of the selected object point be , and the model formula is: , where is the autoregressive coefficient, is the moving average coefficient, is the white noise sequence.

7. A ground flash activity path electrical parameter measurement system based on spatio-temporal clustering analysis according to claim 2, characterized in that: In the adjacent point search unit, the K-D tree data structure is used for spatial neighboring point search. For a given spatio-temporal core object point, adjacent points within the range of the spatial distance threshold are found in the K-D tree.

8. The ground flash activity path electrical parameter measurement system based on spatio-temporal clustering analysis according to claim 2, characterized in that: In the cluster refinement unit, a fuzzy clustering algorithm is introduced. For newly added object points, their belonging is determined according to their fuzzy membership degrees to the existing cluster centers. Let the object point , and the existing cluster center be . The calculation formula for the fuzzy membership degree is as follows: , where is the object point belongs to the cluster of the model membership degree, is the object point to the cluster center distance, is the number of clusters, is the fuzzy exponent, which determines whether the object point is a core object and the cluster it belongs to according to the fuzzy membership degree.

9. The ground flash activity path electrical parameter measurement system based on spatio-temporal clustering analysis according to claim 5, characterized in that: In the region division module, the validity of the region division result is evaluated. The clustering evaluation index coefficient is used for evaluation. The data point is the average distance from the point to other points in the same region. The silhouette coefficient is , where is the silhouette coefficient of the cloud-to-ground flash data point , is the average distance from the cloud-to-ground flash data point to the cloud-to-ground flash data points in the same terrain region, is the average distance from the cloud-to-ground flash data point to the cloud-to-ground flash data points in the nearest different terrain region. The average value of the silhouette coefficients of all data points is calculated to obtain the evaluation value of the entire region division result.

10. A ground flash activity path electrical parameter measurement system based on spatio-temporal clustering analysis according to claim 1, characterized in that: For the measured value of electric field strength, the information entropy calculation formula is , where Similarly, the information entropy of the measured value of current intensity and voltage can be obtained, where represents the information entropy of the measured value of electric field strength, is the number of measured values of electric field strength, is the th measured value of electric field strength, where represents the th proportion of the measured value of electric field strength in the total sum of all measured values, is the th measured value of electric field strength; the weight coefficient is calculated according to the information entropy, and the weight coefficient calculation formula is , where represents the weight coefficient corresponding to the measured value of electric field strength , is the information entropy of the measured value of electric field strength , is the number of measured values of electric field strength, is the index for summation, ranging from 1 to traverse all measured values of electric field strength to calculate the value of the denominator. Similarly, for the calculation of the weight coefficients of current intensity and voltage, the determined weight coefficients are weighted and averaged to obtain the fusion values of electric field strength, current intensity, and voltage.

Images