CN-121983928-A - Intelligent dynamic voltage enhancement type lightning driving device and method based on lightning early warning

CN121983928ACN 121983928 ACN121983928 ACN 121983928ACN-121983928-A

Abstract

The application relates to the technical field of lightning disaster protection, and discloses an intelligent dynamic voltage enhancement type lightning driving device and method based on lightning early warning. The extraction module is used for collecting electric fields and charge parameters to extract evolution characteristics, the prediction module is used for predicting a pilot path to obtain a threat degree scoring matrix, the identification module is used for differently configuring electrode voltages to construct predictive trap distribution, and the monitoring module is used for tracking pilot deviation and dynamically migrating trap positions to realize intelligent lightning protection. The application improves the accuracy, the energy efficiency ratio and the deflection resistance robustness of the lightning protection.

Inventors

YUAN LIANG
LU XIN
GUO ZHIDUAN
LIU ZHENG
LI YUANHENG
ZHANG LEI
WANG KE
DONG HUIFENG
Ren fuli
CHEN BINGXIA

Assignees

国网河南综合能源服务有限公司

Dates

Publication Date: 20260505
Application Date: 20260127

Claims (10)

1. An intelligent dynamic voltage enhancement type lightning driving device based on lightning early warning, which is characterized by comprising: the extraction module is used for collecting electric field intensity and thunderstorm cloud charge parameters, extracting electric field evolution characteristics and obtaining an electric field evolution characteristic sequence; the prediction module is used for performing pilot path prediction processing on the electric field evolution characteristic sequence to obtain a space threat degree scoring matrix; The identification module comprises an identification unit, an output unit and a construction unit, wherein the identification unit is used for identifying space regions with different threat levels according to the space threat degree scoring matrix, the output unit is used for outputting an enhanced voltage to an electrode group in a high threat degree region to form a local strong electric field so as to induce a precursor to develop according to a preset path, the construction unit is used for outputting a reference voltage to an electrode group in a low threat degree region so as to reduce energy consumption, and a high-strength electric field trap is constructed in advance at a possible arrival position of the precursor to obtain space differentiation predictive electric field trap distribution; The monitoring module is used for monitoring the actual development track of the guide in real time, when the spatial deviation of the guide deviated from the predicted path exceeds a set threshold value, a new electric field trap unit is activated at the predicted interception position in front of the deviated guide movement, meanwhile, the output voltage is reduced for the electrode group at the original trap position to realize the trap migration, and the trap position is continuously adjusted by tracking the dynamic change of the guide, so that the tracking type dynamic migration electric field trap chain is obtained.
2. The lightning early warning based intelligent dynamic voltage enhancement type lightning driving device according to claim 1, wherein the extracting module is configured to: Acquiring the electric field intensity in the X-axis direction, the electric field intensity in the Y-axis direction and the electric field intensity in the Z-axis direction through an atmospheric electric field three-component sensor to form an electric field intensity three-dimensional matrix; Acquiring three-dimensional coordinates and electric charge quantity of a thunderstorm cloud charge center through a thunderstorm cloud charge positioning radar; performing time sequence processing on the electric field intensity three-dimensional matrix based on a sliding time window, and calculating an electric field time derivative and a spatial gradient to obtain an electric field change rate characteristic sequence; And calculating a charge center displacement speed vector and a charge quantity change rate based on the three-dimensional coordinates of the thunderstorm cloud charge center, and fusing the electric field change rate feature sequence, the charge center displacement speed vector and the charge quantity change rate to obtain an electric field evolution feature sequence.
3. The lightning early warning based intelligent dynamic voltage enhancement lightning driving device according to claim 1, wherein the prediction module is configured to: inputting the electric field evolution characteristic sequence into a long-period memory layer and a short-period memory layer of a hybrid neural network to extract a time sequence evolution rule, so as to obtain lightning lead development acceleration and an initial development direction angle; Dividing a protection area into three-dimensional space grids, constructing node feature vectors of a graph neural network based on electric field gradients of all grid points and distances from a thunderstorm cloud charge center, establishing an edge connection relation according to electric field gradient direction included angles and space distances among nodes, and calculating to obtain the probability that all the space grid points are penetrated by a leader through a graph neural network aggregation function; Based on the lightning leading development acceleration, the initial development direction angle and the probability that each space grid point is penetrated by a leading, adopting a Monte Carlo tree search algorithm to perform path expansion from a thunderstorm cloud charge center position, calculating the leading extension speed and the candidate position selection probability according to the local electric field intensity at each time step, and generating a plurality of leading evolution path samples; And calculating the minimum distance, the pilot arrival time and the path occurrence probability of each path and the protection target for the plurality of pilot evolution path samples, and counting the number of paths passing through each space grid point and the average minimum distance and average arrival time of the paths to obtain a space threat degree scoring matrix.
4. The lightning early warning based intelligent dynamic voltage enhancement type lightning driving device according to claim 1, wherein the identification unit is configured to: Performing threshold segmentation processing on the space threat degree scoring matrix, marking grids with threat degree scores larger than threat degree thresholds as high threat grids, wherein the threat degree thresholds are determined by counting percentiles of threat degrees of all grids; Carrying out neighborhood aggregation on the high threat grids by adopting a spatial clustering algorithm, merging adjacent high threat grids into threat area clusters, and calculating centroid coordinates and spatial range radiuses of the threat area clusters; marking a grid with threat level scores between one-half of the threat level threshold and the threat level threshold as a secondary threat area; Grids with threat level scores less than one-half the threat level threshold are marked as low threat areas.
5. The lightning early warning based intelligent dynamic voltage enhancement type lightning driving device according to claim 4, wherein the output unit is configured to: Selecting an electrode group closest to the centroid coordinates of the threat area cluster as a main response unit; Calculating a basic enhancement coefficient based on the ratio of the threat degree scoring maximum value to the nominal threat degree in the threat region cluster; calculating the enhancement voltage amplitude value which is required to be output by the main response unit according to the basic enhancement coefficient and the space range radius of the threat area cluster; and calculating the voltage pulse frequency which the main response unit should output according to the predicted arrival time of the lead.
6. The lightning early warning based intelligent dynamic voltage enhancement type lightning driving device according to claim 5, wherein the construction unit is configured to: Setting a reduction coefficient for the electrode group allocated to the secondary threat area, calculating a secondary enhancement voltage amplitude according to the basic enhancement coefficient, the reduction coefficient and the space range radius, and setting a voltage pulse frequency as a reference frequency; Setting the voltage amplitude value as a reference voltage for the electrode group in the low threat area, and setting the voltage pulse frequency as a standby frequency; And forming gradient electric field intensity spatial distribution according to the enhanced voltage amplitude and the voltage pulse frequency of the high threat level region, the secondary enhanced voltage amplitude and the reference frequency of the secondary threat level region, and the reference voltage and the standby frequency of the low threat level region, so as to obtain spatial differential predictive electric field trap distribution.
7. The lightning early warning based intelligent dynamic voltage enhancement type lightning driving device according to claim 1, wherein the monitoring module is configured to: Collecting a pilot channel luminous signal through an ultra-high-speed photoelectric detection array, recording a pilot starting position coordinate and a pilot position coordinate at each subsequent moment, and constructing an actual pilot track sequence; comparing the actual pilot track sequence with a predicted pilot path, and calculating the space deviation amount and the direction deviation angle at each moment; when the spatial deviation amount of a plurality of continuous sampling points exceeds a spatial deviation threshold value or the direction deviation angle exceeds an angle deviation threshold value, judging that the pilot is significantly deviated, and starting a dynamic trap migration program; Based on the current actual position and the speed vector of the lead, predicting a predicted interception position in a future time period of the lead by adopting linear extrapolation, recalculating threat degree distribution in a radius range of the predicted interception position, identifying a new high threat cluster, selecting an inactive electrode group closest to the centroid of the new high threat cluster as a secondary trap unit, increasing the voltage of the inactive electrode group to a secondary enhanced voltage in a set time, and reducing the voltage of the electrode group in the original trap position to a reference voltage in the set time to obtain a tracking type dynamic migration electric field trap chain.
8. An intelligent dynamic voltage enhanced lightning protection method based on lightning early warning realized by the intelligent dynamic voltage enhanced lightning protection device based on lightning early warning according to any one of claims 1 to 7, characterized in that the method comprises: S1, acquiring electric field intensity and thunderstorm cloud charge parameters, and extracting electric field evolution characteristics to obtain an electric field evolution characteristic sequence; s2, carrying out pilot path prediction processing on the electric field evolution characteristic sequence to obtain a space threat degree scoring matrix; Step S3, identifying space regions with different threat levels according to the space threat degree scoring matrix, outputting enhanced voltages to electrode groups in the high threat degree region to form a local strong electric field so as to induce a pilot to develop according to a preset path, outputting reference voltages to electrode groups in the low threat degree region to reduce energy consumption, and constructing high-strength electric field traps in advance at possible arrival positions of the pilot to obtain space differentiation predictive electric field trap distribution; And S4, monitoring the actual development track of the guide in real time, when the spatial deviation of the guide deviated from the predicted path exceeds a set threshold value, activating a new electric field trap unit at the predicted interception position in front of the deviated guide movement, reducing the output voltage of an electrode group at the original trap position to realize trap migration, and continuously adjusting the trap position by tracking the dynamic change of the guide to obtain a tracking type dynamic migration electric field trap chain.
9. The method according to claim 8, wherein step S1 comprises: Acquiring the electric field intensity in the X-axis direction, the electric field intensity in the Y-axis direction and the electric field intensity in the Z-axis direction through an atmospheric electric field three-component sensor to form an electric field intensity three-dimensional matrix; Acquiring three-dimensional coordinates and electric charge quantity of a thunderstorm cloud charge center through a thunderstorm cloud charge positioning radar; performing time sequence processing on the electric field intensity three-dimensional matrix based on a sliding time window, and calculating an electric field time derivative and a spatial gradient to obtain an electric field change rate characteristic sequence; And calculating a charge center displacement speed vector and a charge quantity change rate based on the three-dimensional coordinates of the thunderstorm cloud charge center, and fusing the electric field change rate feature sequence, the charge center displacement speed vector and the charge quantity change rate to obtain an electric field evolution feature sequence.
10. The method according to claim 8, wherein step S2 comprises: inputting the electric field evolution characteristic sequence into a long-period memory layer and a short-period memory layer of a hybrid neural network to extract a time sequence evolution rule, so as to obtain lightning lead development acceleration and an initial development direction angle; Dividing a protection area into three-dimensional space grids, constructing node feature vectors of a graph neural network based on electric field gradients of all grid points and distances from a thunderstorm cloud charge center, establishing an edge connection relation according to electric field gradient direction included angles and space distances among nodes, and calculating to obtain the probability that all the space grid points are penetrated by a leader through a graph neural network aggregation function; Based on the lightning leading development acceleration, the initial development direction angle and the probability that each space grid point is penetrated by a leading, adopting a Monte Carlo tree search algorithm to perform path expansion from a thunderstorm cloud charge center position, calculating the leading extension speed and the candidate position selection probability according to the local electric field intensity at each time step, and generating a plurality of leading evolution path samples; And calculating the minimum distance, the pilot arrival time and the path occurrence probability of each path and the protection target for the plurality of pilot evolution path samples, and counting the number of paths passing through each space grid point and the average minimum distance and average arrival time of the paths to obtain a space threat degree scoring matrix.

Description

Intelligent dynamic voltage enhancement type lightning driving device and method based on lightning early warning Technical Field The application relates to the technical field of lightning disaster protection, in particular to an intelligent dynamic voltage enhancement type lightning driving device and method based on lightning early warning. Background Lightning is taken as a serious natural disaster phenomenon, and forms a great threat to important infrastructures such as power transmission lines, communication base stations, petrochemical facilities and the like. The traditional lightning protection technology mainly relies on passive protection devices such as lightning rods, lightning strips and the like, lightning current is led into the ground in a lightning leading mode, or a fixed-position lightning driving device is adopted to continuously apply constant voltage to a space to form a static electric field barrier. With the development of the atmospheric electric field monitoring technology and the thunderstorm cloud charge positioning technology in recent years, partial lightning early warning systems can detect abnormal changes of electric fields and send early warning signals in a few minutes before lightning occurs, but the early warning information is mainly used for personnel evacuation and equipment power-off protection, effective linkage is not formed with an active protection device, the existing active lightning driving device generally applies high-voltage pulses with the same amplitude to all electrode arrays in a protection area after the thunderstorm cloud is detected to be close, and attempts are made to change the development path of lightning leads through evenly distributed strong electric fields. However, the prior art has the following defects that firstly, a high-intensity electric field is applied to the whole protection space indiscriminately by a traditional fixed voltage lightning driving method, so that the system energy consumption is huge, energy waste is caused in a non-threat area, a power supply system is difficult to support continuous high-power output under the condition that the thunderstorm duration is long, secondly, although the existing lightning early warning system can detect the approaching of thunderstorm cloud and the abnormal change of an atmospheric electric field, the specific development path and the space threat distribution of a lightning lead cannot be predicted, only the dangerous grade judgment of the general system can be provided, the lightning driving device cannot be guided to accurately position a high threat area in space and concentrate electric field resources for targeted protection, and thirdly, the existing lightning driving device adopts a preset static electric field trap distribution mode, once the lead is influenced by factors such as wind field disturbance, local electric field nonuniformity or charge distribution change in the development process, the static trap position deviates from a predicted path, space dislocation occurs with an actual track, so that the effective interception of the trap failure cannot be realized. Disclosure of Invention The application provides an intelligent dynamic voltage enhancement type lightning driving device and method based on lightning early warning, which are used for solving the problems that in the prior art, a lightning protection system cannot predict leading space threat distribution to cause unreasonable protection resource allocation, a static electric field trap cannot cope with leading deviation behaviors to cause high protection failure rate, and improving the accuracy, energy efficiency ratio and deviation resistance robustness of lightning protection. In a first aspect, the present application provides an intelligent dynamic voltage enhancement type lightning driving device based on lightning early warning, the intelligent dynamic voltage enhancement type lightning driving device based on lightning early warning includes: the extraction module is used for collecting electric field intensity and thunderstorm cloud charge parameters, extracting electric field evolution characteristics and obtaining an electric field evolution characteristic sequence; the prediction module is used for performing pilot path prediction processing on the electric field evolution characteristic sequence to obtain a space threat degree scoring matrix; The identification module comprises an identification unit, an output unit and a construction unit, wherein the identification unit is used for identifying space regions with different threat levels according to the space threat degree scoring matrix, the output unit is used for outputting an enhanced voltage to an electrode group in a high threat degree region to form a local strong electric field so as to induce a precursor to develop according to a preset path, the construction unit is used for outputting a reference voltage to an electrode group in a low