CN-122024427-A - Weather radar-based power transmission line bird strike risk early warning method

CN122024427ACN 122024427 ACN122024427 ACN 122024427ACN-122024427-A

Abstract

The invention discloses a weather radar-based power transmission line bird strike risk early warning method which comprises the steps of denoising, standardizing and space-time matching processing of weather radar original data, power transmission line geographic data and auxiliary environment data, extracting radar echo characteristic parameters to construct a bird target recognition model, accurately recognizing and positioning bird targets, tracking bird tracks, calculating motion parameters to obtain a predicted path, verifying the position relationship between the path and the power transmission line, constructing a bird strike risk assessment model by combining multiple types of data, calculating real-time risk values of all sections, classifying and generating early warning signals, and pushing the early warning signals to an operation and maintenance platform and related personnel terminals. According to the invention, the deep fusion of the meteorological radar resource and the operation and maintenance requirements of the power transmission line is realized, so that the operation and maintenance fusion of the meteorological radar and the power transmission line is realized, and the interference targets such as birds, precipitation, insects and the like are accurately distinguished.

Inventors

KUANG YANJUN
WU QIONG
WANG HUAIQING
TONG JUNXIN
HUANG KUN
LIU YANAN
HE MENG

Assignees

国网江西省电力有限公司电力科学研究院
江西省气象服务中心(江西省专业气象台、江西省气象宣传与科普中心)

Dates

Publication Date: 20260512
Application Date: 20260416

Claims (10)

1.A weather radar-based power transmission line bird strike risk early warning method is characterized by comprising the following steps: step S1, collecting multi-source basic data of a target area, and processing the multi-source basic data of the target area to obtain processed real-time weather radar data; S2, radar echo data are acquired, characteristic ranking is carried out through a Relief-F algorithm, 6-8-dimensional radar echo characteristic parameters are screened, the 6-8-dimensional radar echo characteristic parameters are input into a random forest model for training, and an improved random forest bird target recognition model is obtained; S3, tracking a track chain and calculating motion parameters of the identified bird target to obtain a three-dimensional speed vector, and performing error calibration by combining the three-dimensional speed vector with the self-adaptive weight parameters to obtain a bird prediction path through fitting; S4, constructing a bird strike risk assessment model by combining the bird predicted path, the target area multisource basic data and the conflict influence degree of the bird predicted path and the power transmission line, and calculating real-time bird strike risk values of all sections of the power transmission line; And S5, dividing risk grades according to the real-time bird strike risk values, generating corresponding early warning signals, and pushing the early warning signals to the transmission line operation and maintenance management platform and related personnel terminals.
2. The weather radar-based power transmission line bird strike risk early warning method according to claim 1 is characterized in that the specific process of the step S1 is as follows: acquiring weather radar original data and transmission line space geographic data in a target transmission line; Collecting real-time meteorological data, seasonal information, bird migration rule data and historical bird strike accident data of a target power transmission line, and taking the collected real-time meteorological data, seasonal information, bird migration rule data and historical bird strike accident data as auxiliary environment data; Taking the original data of the weather radar, the spatial geographic data of the transmission line and the auxiliary environment data as target area multisource basic data; The method comprises the steps of carrying out denoising treatment on weather radar original data by adopting a self-adaptive wavelet threshold denoising algorithm to remove radar system noise, ground fixed clutter and electromagnetic interference, carrying out elimination by combining terrain elevation data with threshold segmentation and region growing algorithm to obtain denoised weather radar data; Unifying spherical coordinate data of the denoised weather radar data on a Cartesian three-dimensional grid of the same space-time reference through coordinate conversion and interpolation; On a Cartesian three-dimensional grid, performing time synchronization on the denoised weather radar data and the real-time meteorological data based on a time stamp, and performing prediction processing on the real-time meteorological data by using a meteorological prediction model to obtain predicted meteorological data; And (3) correlating the grid data of the denoised weather radar data with the spatial geographic data of the power transmission line and the predicted meteorological data by using a spatial interpolation algorithm, establishing a spatial mapping relation among a radar detection area, the spatial geographic data of the power transmission line and the predicted meteorological data, and carrying out time-space alignment fusion on the denoised weather radar data, the spatial geographic data of the power transmission line and the predicted meteorological data to obtain time-space aligned multi-source fusion data, namely the processed real-time weather radar data.
3. The weather radar-based power transmission line bird strike risk early warning method according to claim 2, wherein the specific process of the step S2 is as follows: The radar echo data comprises a reflectivity factor Z, a differential reflectivity ZDR, a correlation coefficient rhohv, a differential phase shift phidp, an echo area, a shape factor, compactness and radial speed texture characteristics; The reflectivity factor Z, the differential reflectivity ZDR, the correlation coefficient rho hv, the differential phase shift phidp, the area, the shape factor, the compactness and the radial speed texture characteristics of the echo are subjected to importance sorting through a Relief-F algorithm, and 6-8-dimensional radar echo characteristic parameters are screened; initializing a random forest model, and presetting a super-parameter candidate range of the number of decision trees, the maximum depth and the minimum number of split nodes; The super parameter candidate range of the number of the decision trees is {50,100,150,200}, the super parameter candidate range of the maximum depth is {5,10,15}, and the super parameter candidate range of the minimum number of the split nodes is {2,4,6}; All the super-parameter combinations are generated by fully arranging and combining {50,100,150,200}, {5,10,15}, and {2,4,6 }; Taking 6-8-dimensional radar echo characteristic parameters as input, traversing all the super-parameter combinations by a grid search method, calculating bird target recognition accuracy corresponding to each group of super-parameter combinations, selecting the combination with the highest recognition accuracy as an optimal super-parameter, and obtaining a random forest model after super-parameter optimization based on the optimal super-parameter; Introducing an AdaBoost algorithm, carrying out weighted iterative training by taking a random forest model after super parameter optimization as a base classifier and taking 6-8-dimensional radar echo characteristic parameters as input to finally obtain an improved random forest bird target recognition model; extracting azimuth angles, elevation angles and detection distances corresponding to the weather radar identified as birds as polar coordinate data, converting the polar coordinate data into Cartesian rectangular coordinates, obtaining geographic coordinates of bird targets based on the Cartesian rectangular coordinates, and realizing bird target positioning based on the geographic coordinates of the bird targets; Orderly recording continuous body scanning frames of weather radar The body scanning frame is a complete weather radar data frame generated by the weather radar completing one-time full airspace three-dimensional body scanning observation, and the time interval between adjacent body scanning frames is as follows ; Sweep data for the kth frame; Sweep data for the k+1st frame; screening out bird targets based on target types and establishing a bird target point set, wherein the bird target point set is represented by: ; ; wherein, birds target point contains longitude coordinate, latitude coordinate, altitude and timestamp, represents: 、 ; Wherein, the A set of all bird target points identified in the kth frame; An ith bird target point identified in a kth frame; A total number of bird target points identified in the kth frame body scan data; longitude coordinates of an ith bird target point in a kth frame; Latitude coordinates of an ith bird target point in a kth frame; Altitude for the ith bird target point in the kth frame; a time stamp for the kth swipe; a set of all bird target points identified in frame k+1; a j-th bird target point identified in the k+1-th frame; total number of bird target points identified in the k+1st frame body scan data; longitude coordinates of the jth bird target point in the k+1th frame; latitude coordinates of a jth bird target point in a k+1th frame; Altitude for the jth bird target point in the k+1 frame; Is the timestamp of the k+1th swipe.
4. The weather radar-based power transmission line bird strike risk early warning method according to claim 3, wherein the specific process of the step S3 is as follows: Space speed constraint, namely setting the maximum safe flying speed vmax of birds; Based on Bird target point in (a) And (3) with Bird target point in (a) Calculating a space distance d; Judging conditions when Then the space velocity constraint is satisfied, otherwise, the space velocity constraint is directly satisfied And (3) with Exclusion, recalculation ; Timing association constraint that only adjacent frames are allowed when the spatial speed constraint is satisfied And (3) with Is a bird target point of (2) And (3) with Establishing an association; When the space velocity constraint and the time sequence association constraint are simultaneously satisfied, then And (3) with The method is characterized by comprising the following steps of: For the next frame Each bird target point in (a) In the previous frame Is a bird target point of (2) Searching all associable candidate point pairs meeting space velocity constraint and time sequence association constraint; When (when) Only 1 of the associable candidate point pairs, the associable candidate point pairs are directly associated with Matching to obtain And (3) with An inter-frame matching result; let the track chain of each bird target be Wherein Start frame for track chain An mth bird target point; Start frame for track chain An mth bird target point in +1; terminating frames for track chains An mth bird target point; a track chain set for the mth bird target point; Wherein the track chain Comprises the following parameters of numbering Target point sequence of effective birds Up-to-date motion speed in track chain Up-to-date motion acceleration in track chain Status of ; According to And (3) with Inter-frame matching results for track chain Updating in real time to create a new track chain of bird target points ; A new track chain for the mth bird target point; comprising the following parameters of new number New and effective bird target point sequence New track chain up to date motion speed New motion acceleration in the track chain New state ; From the slave Extract the latest Effective bird target point ; ≥3; Is the (k+1) th frame An effective bird target point; fitting the latest by least square method Effective bird target point Obtaining a three-dimensional velocity vector ; A three-dimensional velocity vector for the mth bird target point; Longitudinal direction of target point for the mth bird Components of (2); latitudinally for the mth bird target point Components of (2); for the mth bird target point in the altitude direction Components of (2); Calculation of Is the m-th bird target point feature vector : Comprising 、、 ; Azimuth angle characteristics of the mth bird target point; Pitch angle characteristic for the mth bird target point; Radial velocity characteristics for the mth bird target point; According to Through a preset scoring function Calculating to obtain self-adaptive weight parameters ; The adaptive weight parameter of the target point of the mth bird; Weather disturbance weight Establishing weather disturbance weights according to the predicted meteorological data acquired in the step S1 ; Comprehensive weight Comprehensive of And Calculating final weight parameters of the mth bird target point Wherein Is a balance coefficient; Adaptive speed correction according to Calculating the fitting speed of the mth bird target point ; Calculating a final three-dimensional velocity vector When (1) Vmax, pair of The velocity vector is normalized and corrected to obtain a corrected velocity vector ; The final three-dimensional velocity vector is the m-th bird target point; the final three-dimensional velocity vector corrected for the mth bird target point; Future position prediction setting Middle (f) The latest effective bird target points are ; Is the first Frame No The latest effective bird target points; is the first The latest effective bird target point is at the first The longitudinal direction of the frame is known with coordinates; is the first The latest effective bird target point is at the first The latitude direction of the frame is known with coordinates; is the first The latest effective bird target point is at the first The altitude direction of the frame is known with coordinates; Acquisition of the first The real-time timestamp corresponding to the frame is , Is a preset future prediction time stamp, and the future prediction time interval is ; Future predicted location ; Target point for the mth bird Future three-dimensional predicted positions of (a); ; Target point for the mth bird Longitude coordinate prediction values of (a); Target point for the mth bird Latitude coordinate prediction values of (a); Target point for the mth bird Altitude predictions for (a); Set the standard deviation of the position noise observed by the weather radar Fitting error standard deviation with bird target recognition model The total error standard deviation is ; Based on the total error standard deviation For a pair of 、、 Performing correction to obtain corrected 、、 For a pair of 、、 Performing cubic spline interpolation fitting to obtain a bird prediction path; For the modified mth bird target point Longitude coordinate prediction values of (a); For the modified mth bird target point Latitude coordinate prediction values of (a); For the modified mth bird target point Altitude predictions for (a); Checking longitude and latitude coordinates of the bird prediction path and the power transmission line to obtain conflict influence degree of the bird prediction path and the power transmission line; the collision influence degree of the bird predicted path and the power transmission line comprises analyzing the longitude and latitude coordinate distance between the bird position and the power transmission line And analyzing the predicted time for the bird position to reach the transmission line from the current position 。
5. The method for early warning bird strike risk of power transmission line based on weather radar according to claim 4, wherein the track chain of the mth bird target is characterized in that The updating process comprises the following specific steps: track chain for mth bird target Setting 4 mutually exclusive track states, namely, new generation, updating, temporary storage and extinction; Track state new generation when the next frame Bird target point in (a) Not in line with the previous frame Any bird target point in (a) Successful match and the next frame Bird target point in (a) Meeting the neighborhood density requirement, the target point of the bird which is not successfully matched is Assigning unique track chain numbers And creating a new track chain of avian targets At the same time will Marking as new born; Track chain update-current frame Target point of birds in China With the next frame Bird target point in (a) Successful matching to obtain effective bird target point Target point of effective birds Adding in A kind of electronic device In accordance with Updating And At the same time link track state Marking as updating; track chain temporary storage, namely the current frame Target point of birds in China In the next frame Target point of birds in China No matching point in the case A kind of electronic device Marked as temporary storage and reserved Is a data set of the data set; track chain death when In a temporary storage state, when the next frame Target point of birds in China Not yet match to the corresponding bird target point, or The target point of the birds exceeds the effective coverage of the weather radar 200Km, will A kind of electronic device Marked as extinction and stop pairing Is provided.
6. The method for early warning bird strike risk of power transmission line based on weather radar according to claim 5, wherein the method is characterized by comprising the following steps of 、、 The specific process of (2) is as follows: By calculating the latest Effective bird target point The standard deviation of the speed is normalized to the (0, 1) interval to obtain the speed stability index of the mth bird target point ; By calculating the latest Effective bird target point The angle change mean value of the continuous displacement vector is normalized to the [0,1] interval, and the direction change rate of the mth bird target point is obtained ; By calculating the latest Effective bird target point The standard deviation of the altitude is normalized to the (0, 1) interval to obtain the m-th bird target point altitude fluctuation index 。
7. The weather radar-based power transmission line bird strike risk early warning method of claim 6 is characterized in that the specific process of constructing the bird strike risk assessment model is as follows: Establishing a risk assessment index system, wherein the primary index comprises a bird prediction path, power transmission line space geographic data and auxiliary environment data, and the secondary index is the conflict influence degree of the bird prediction path and the power transmission line; And determining weight coefficients of the primary index and the secondary index by an analytic hierarchy process, and constructing the bird strike risk assessment model based on the weight coefficients of the primary index and the secondary index by a linear weighting method.
8. The method for early warning bird strike risk of power transmission line based on weather radar according to claim 7, wherein the method is characterized in that Through a preset scoring function Calculating to obtain self-adaptive weight parameters ; When birds perform regular long-distance migration, 、、 Are all smaller than the first threshold value of 0.2, and the scoring function =1.0, Calculate adaptive weight parameters Obtaining a predicted path of birds for high weight; When the birds are in the form of a spiral foraging, 、 Greater than a second threshold of 0.5, scoring function =0.3, Calculate adaptive weight parameters Is low weight; When birds encounter severe weather, the long-distance migration mode parameter condition of the birds is not met, the spiral foraging mode parameter condition of the birds is not met, and the scoring function is not met =0.8, Calculate adaptive weight parameters Is a medium weight.
9. An electronic device, comprising a processor, a memory and a bus, wherein the processor and the memory are connected through the bus, the memory is used for storing a set of program codes, and the processor is used for calling the program codes stored in the memory to execute the weather radar-based power transmission line bird strike risk early warning method according to any one of claims 1-8.
10. A non-volatile computer storage medium storing computer executable instructions for performing a weather radar based power transmission line bird strike risk warning method according to any one of claims 1 to 8.

Description

Weather radar-based power transmission line bird strike risk early warning method Technical Field The invention relates to the field of power transmission line risk early warning of weather radar, in particular to a power transmission line bird strike risk early warning method based on weather radar. Background The transmission line is used as a core component of the power system, and the safe and stable operation of the transmission line is directly related to the reliability of power supply. Bird strike is one of the main risks of threatening the safety of the transmission line, and birds may collide with key components such as wires and insulators when moving nearby the transmission line, or nest on towers to cause serious faults such as line short circuit and tripping. According to statistics, the faults of the power transmission line caused by bird strike account for 15% -30% of the total fault number, particularly, the fault occurrence rate is obviously increased in the migration season and the propagation season of birds, and huge economic loss and operation and maintenance pressure are brought to the power department. Traditional bird strike risk early warning means mainly rely on technologies such as manual inspection, optical monitoring (such as a camera), acoustic monitoring and the like. However, the method has obvious limitations that manual inspection efficiency is low, cost is high, all-weather and large-scale coverage is difficult to realize, optical monitoring is easily influenced by illumination conditions (such as night, overcast, rainy, foggy and snowy weather) and environmental shielding, monitoring precision and effective range are limited, acoustic monitoring is easily influenced by environmental noise, and false alarm rate is high. Therefore, a technical method capable of breaking through environmental restrictions, realizing large-scale, high-precision and all-day bird strike risk monitoring and early warning is needed. At present, the application of the weather radar in bird migration law research is mainly focused on the field of bird migration law research, for example, the national weather agency of the United states utilizes a weather radar network to monitor the time, route and density of bird migration, data support is provided for bird protection, and part of universities in China (such as Nanjing information engineering university) develop bird target recognition research based on the weather radar, and the birds and the weather targets are distinguished by extracting characteristics of echo intensity, velocity spectrum width and the like. The existing research is only stopped at the monitoring level of bird macro distribution, a special bird strike risk early warning model is not established according to the specific scene of the power transmission line and the characteristics of the spatial position information and the bird strike risk of the power transmission line, and the special bird strike risk early warning model cannot be directly applied to the safe operation and maintenance of the power transmission line. The invention aims to overcome the defects of limited monitoring range, poor environmental adaptability, high cost, low early warning precision and the like in the existing early warning technology for the bird strike risk of the power transmission line, and provides a weather radar-based early warning method for the bird strike risk of the power transmission line. According to the method, the existing weather radar network resources are fully utilized, the spatial position information of the power transmission line is combined through the accurate processing and analysis of the weather radar echo data, the large-scale, all-weather and high-precision monitoring of bird activities along the power transmission line is realized, a scientific bird strike risk assessment model is built, a grading early warning signal is timely sent out, a reliable decision basis is provided for operation and maintenance personnel of the power transmission line, and the occurrence rate of bird strike accidents is reduced. Disclosure of Invention Aiming at the short plates in the prior art, the invention provides a weather radar-based power transmission line bird strike risk early warning method, which solves the problems mentioned in the background art. In order to achieve the purpose, the invention provides the technical scheme that the method for early warning the risk of bird strike of the power transmission line based on the weather radar comprises the following steps: step S1, collecting multi-source basic data of a target area, and processing the multi-source basic data of the target area to obtain processed real-time weather radar data; S2, radar echo data are acquired, characteristic ranking is carried out through a Relief-F algorithm, 6-8-dimensional radar echo characteristic parameters are screened, the 6-8-dimensional radar echo characteristic parameters are input into a ra