CN114254692A - Multiscale thunderstorm intelligent classification and identification method based on multisource lightning data - Google Patents

Multiscale thunderstorm intelligent classification and identification method based on multisource lightning data Download PDF

Info

Publication number
CN114254692A
CN114254692A CN202111316840.5A CN202111316840A CN114254692A CN 114254692 A CN114254692 A CN 114254692A CN 202111316840 A CN202111316840 A CN 202111316840A CN 114254692 A CN114254692 A CN 114254692A
Authority
CN
China
Prior art keywords
thunderstorm
lightning data
cluster
lightning
cur
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202111316840.5A
Other languages
Chinese (zh)
Inventor
孙世军
郭禹琛
李增伟
何晓凤
庄杰
武正天
朱坤双
李军
韩洪
付以贤
杜远
冯雨晴
綦浩楠
宋香涛
吕守国
韩正新
乔耀华
周洋
蔡俊鹏
贾明亮
冯迎春
王蔚
李冰冰
李永明
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Jiutian Jiutian Meteorological Technology Co ltd
Emergency Management Center Of State Grid Shandong Electric Power Co
Maintenance Branch of State Grid Shandong Electric Power Co Ltd
Original Assignee
Beijing Jiutian Jiutian Meteorological Technology Co ltd
Emergency Management Center Of State Grid Shandong Electric Power Co
Maintenance Branch of State Grid Shandong Electric Power Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Jiutian Jiutian Meteorological Technology Co ltd, Emergency Management Center Of State Grid Shandong Electric Power Co, Maintenance Branch of State Grid Shandong Electric Power Co Ltd filed Critical Beijing Jiutian Jiutian Meteorological Technology Co ltd
Priority to CN202111316840.5A priority Critical patent/CN114254692A/en
Publication of CN114254692A publication Critical patent/CN114254692A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques

Landscapes

  • Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Geophysics And Detection Of Objects (AREA)

Abstract

A multiscale thunderstorm intelligent classification and identification method based on multisource lightning data relates to the technical field of thunderstorm classification and identification, solves the problem that the speed of the existing thunderstorm single body identification and calculation is low and the accuracy is low, and comprises the following processes: processing cut lightning data D ═ x by adopting density clustering method with density peak value and noisejSetting neighborhood parameters and scale information; by applying to all xjThe core objects are found out, the core object set is updated in sequence, the classification sample set which is not visited is updated in sequence, cluster division can be obtained, and then the thunderstorm single body can be obtained. The thunderstorm detection method can rapidly and accurately process the complex thunderstorm conditions with different scales, has excellent identification capability on the thunderstorm monomers, can simultaneously and efficiently identify multiple targets, can provide the thunderstorm density center which is more in line with the physical form of the thunderstorm, and can simultaneously remove the noise of single points.

Description

Multiscale thunderstorm intelligent classification and identification method based on multisource lightning data
Technical Field
The invention relates to the technical field of thunderstorm classification and identification, in particular to a multiscale thunderstorm intelligent classification and identification method based on multisource lightning data.
Background
The existing thunderstorm single body identification is mostly applied to satellites, radars and lightning instruments; the thunderstorm single body identification in the satellite channel data and the radar echo data is based on image identification technology, threshold value method and other means, so that the thunderstorm single body is indirectly judged and identified. In direct lightning detection, lightning instruments can be divided into foundation lightning position indicators and satellite-borne lightning imagers according to different detection principles.
The foundation lightning locator collects electromagnetic signals of a lightning radiation source from a long distance by using an antenna system, and measures characteristic parameters of the lightning radiation source through data processing; a single lightning bolt can detect the occurrence of lightning, but cannot determine the position and time of the occurrence of lightning, and a lightning positioning system formed by networking a plurality of lightning bolts is required for positioning calculation. The lightning position finder system can be further divided into a low frequency or very low frequency (LF/VLF) lightning position system, such as a National Lightning Detection Network (NLDN) in the united states, which can detect cloud-to-ground flashes and partial cloud flashes; still another is a high or very high frequency (HF/VHF) lightning location system, such as the lightning graphic system (LMA) developed by the new mexico mine technical college in the united states, which is currently the most accurate three-dimensional lightning location system internationally and can detect almost all cloud flashes.
The satellite-borne lightning imager detects the light radiation characteristic when lightning occurs through a CCD area array, and the position, time and the like of the lightning are measured. For example, a lightning imager for a wind cloud satellite number four.
The format of the lightning data record is generally strip by strip, records one-time lightning, generally contains geographic information, strength information, gradient information and the like, and thunderstorm monomers obtained by identifying and Clustering lightning distribution, some existing Clustering methods such as K-means (K-means Clustering algorithm), CFSFDP (Clustering by fast cluster algorithm and find of dense peak), and the like need to know the number of the thunderstorm monomers, or need manual assistance and semi-supervised learning to realize the identification and differentiation of different thunderstorm monomers, DBSCAN (Density-Based Clustering method with Noise) has more excellent capability in identifying a plurality of monomers, and has very good identification capability and can remove some single-point Noise when the thunderstorm monomers are in a concave shape, however, the DBSCAN cannot identify the thunderstorm center after identifying the thunderstorm cell, the pure geometric center calculation may deviate from the actual thunderstorm intensity center, and for some concave-shaped cells, the geometric center may deviate from the cell. Based on this, it is very important to find a relatively fast and accurate thunderstorm monomer automatic identification algorithm technology.
Disclosure of Invention
In order to solve the problems of low speed and low accuracy of thunderstorm individual body identification calculation of the existing thunderstorm individual body automatic identification algorithm, the invention provides a multiscale thunderstorm intelligent classification and identification method based on multisource lightning data.
The technical scheme adopted by the invention for solving the technical problem is as follows:
a multiscale thunderstorm intelligent classification and identification method based on multisource lightning data comprises the following steps:
s1, acquiring the lightning data, and unifying the lightning data in a preset format;
s2, cutting the unified lightning data according to a certain time interval;
s3, recording the cut lightning data as D ═ xj}＝(x1,x2,x3.....xm) J is 1,2,3,. m, m is a positive integer; setting neighborhood parameters E and scale information;
s4, initializing a core object set
Figure BDA0003343926680000021
Initializing cluster number k as 0, initializing classification sample set Γ as D, initializing cluster division
Figure BDA0003343926680000022
S5, finding out D ═ { x ] in S3jAll core objects in the } and update Ω accordingly;
s6, judging whether omega is an empty set or not, if so, judging whether omega is an empty set or not
Figure BDA0003343926680000024
Proceed to S7 if
Figure BDA0003343926680000023
Then proceed to S10;
s7, in Ω, randomly selecting a first core object o, initializing the current cluster core object queue Ω cur ═ { o }, k ═ k +1, and making k ═ k', initializing the current cluster CkUpdating Γ, namely Γ '═ Γ - { o }, Γ ═ Γ';
s8, judging whether the omega cur is an empty set, if so, determining whether the omega cur is an empty set
Figure BDA0003343926680000025
Then C iskAfter generation, C is updated, i.e. C is made to be { C {1,C2,...,CkGet up toOmega, i.e. omega' ═ omega-CkΩ ═ Ω', return to S6; if it is
Figure BDA0003343926680000026
Then omega is updated, i.e. omega' ═ omega-CkS9 is performed, Ω ═ Ω';
s9, in Ω cur, randomly taking out a second core object o ', finding out all neighborhood subsamples N e (o ') by e, making Δ ═ N e (o ') # Γ, updating CkI.e. Ck'＝Ck∪Δ、Ck＝Ck', update Γ, i.e. Γ' ═ Γ - Δ, Γ ═ Γ ', update Ω cur, i.e. Ω cur' ═ Ω cur ═ u (Δ ═ Ω) -o ', Ω cur ═ Ω cur', return to S8;
s10, calculating the density center point of C to obtain { C rho1,Cρ2,...,Cρk}。
The invention has the beneficial effects that:
the multiscale thunderstorm intelligent classification and identification method based on multisource lightning data establishes a unique thunderstorm monomer identification method, has very good identification capability on thunderstorm monomers, can simultaneously and efficiently identify multiple targets, can process thunderstorm conditions with different scales and complexity, provides a thunderstorm density center which is relatively in line with the physical form of the thunderstorm, and can remove noise of a single point, so the multiscale thunderstorm intelligent classification and identification method based on multisource lightning data can quickly and accurately automatically identify the thunderstorm monomers and provide the thunderstorm density center. The method provides a method for rapidly clustering and identifying the thunderstorm monomers with different scales at the same time, provides a data base for predicting the movement track of the thunderstorm monomers, has an obvious effect on identifying the multi-thunderstorm monomers, and is widely applied to the fields of power grids, oil storage and the like which need lightning early warning prediction.
Drawings
Fig. 1 is a flow chart of the multiscale thunderstorm intelligent classification and identification method based on multisource lightning data.
Fig. 2 is a flowchart of S5 of the multi-scale thunderstorm intelligent classification and identification method based on multi-source lightning data of the present invention.
Fig. 3 is a graph of the recording result of lightning by a lightning instrument.
Fig. 4A is a thunderstorm individual classification and thunderstorm density central map obtained by classifying and identifying 19:35 lightning data according to the multi-scale thunderstorm intelligent classification and identification method based on multi-source lightning data.
Fig. 4B is a thunderstorm individual classification and thunderstorm density central map obtained by classifying and identifying 19:40 lightning data according to the multi-scale thunderstorm intelligent classification and identification method based on multi-source lightning data.
Fig. 4C is a thunderstorm individual classification and thunderstorm density central map obtained by classifying and identifying 19:55 lightning data according to the multi-scale thunderstorm intelligent classification and identification method based on multi-source lightning data of the present invention.
Fig. 4D is a thunderstorm individual classification and thunderstorm density central map obtained by classifying and identifying 20:05 lightning data according to the multi-scale thunderstorm intelligent classification and identification method based on multi-source lightning data of the present invention.
Fig. 5 is a central diagram of thunderstorm density and the nationwide regional thunderstorm individual classification identified by the multi-scale thunderstorm intelligent classification and identification method based on multi-source lightning data.
Detailed Description
In order that the above objects, features and advantages of the present invention can be more clearly understood, a more particular description of the invention will be rendered by reference to the appended drawings.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, however, the present invention may be practiced in other ways than those specifically described herein, and therefore the scope of the present invention is not limited by the specific embodiments disclosed below.
A multiscale thunderstorm intelligent classification and identification method based on multisource lightning data, as shown in figure 1, comprises the following steps:
and S1, acquiring the lightning data and unifying the lightning data in a preset format.
Lightning data of lightning instruments are collected, wherein the data comprises lightning data of a foundation lightning locator and lightning data of a satellite-borne lightning imager, and specifically comprises lightning data of a three-dimensional lightning instrument and lightning data of an FY4 satellite lightning imager. And performing multi-source data fusion on all lightning data, specifically, unifying the longitudes and latitudes of all lightning data, and analyzing the longitudes and latitudes, the heights and the strengths into four effective information.
And S2, cutting the unified lightning data according to a certain time interval.
In the embodiment, the lightning data is cut at intervals of 5 minutes, and usually, the lightning data can be cut into 6 minutes for matching with radar according to the requirements of actual services. And performing space projection operation on the cut lightning data, namely converting longitude and latitude coordinates into geodetic coordinates.
The lightning data obtained in S2 is put into an automatic classifier of DBSCADP & N (Density-Based Spatial Clustering of Applications with Density Peak and Noise Clustering method), which is S3-S11 as follows.
S3, the cut lightning data (lightning data cut and converted into geodetic coordinates at S2) is D ═ xj}＝(x1,x2,x3.....xm) J is 1,2,3,. m, m is a positive integer; and setting neighborhood parameters belonging to E and setting scale information.
In DBSCADP&Inputting cut lightning data in N, D ═ xj}＝(x1,x2,x3.....xm) And inputting neighborhood parameters belonging to the range (namely the minimum sample number and the neighborhood distance threshold value), and inputting scale information. It also includes inputting \ setting the minimum lightning number of thunderstorm MinPts, usually to be 2 or 3.
S4, initializing a core object set omega, and enabling the core object set
Figure BDA0003343926680000041
Initializing cluster number k, and making k equal to 0. Initializing a classification sample set Γ, and making Γ equal to D. Initialize cluster partition C, order
Figure BDA0003343926680000042
S5, finding out the cut lightning data D ═ x in S3j}＝(x1,x2,x3.....xm) And updating the core object set omega according to all the core objects. The specific process is shown in fig. 2.
S5.1, let j equal 1;
s5.2, calculating xjIs x is foundjIs a neighborhood subsample set N e (x) of neighborhood parametersj)；
S5.3, if the neighborhood subsample set N belongs to (x)j) Satisfies the number | N ∈ (x)j) | is not less than MinPts, xjAdding a core object set omega, namely updating omega: Ω' ═ Ω ═ u { x }jH, S5.4, q ═ q'; if | N ∈ (x)j) If | MinPts, then xjAnd S5.4 is carried out without adding the core object set omega.
S5.4, determining whether j is equal to m, if j is not equal to m, j '═ j +1, and then j ═ j', and returning to S5.2 with a new j; if j is m, S6 is performed.
S6, judging whether the core object set omega is an empty set, if so, judging whether the core object set omega is an empty set
Figure BDA0003343926680000053
Then proceed to S7; if it is
Figure BDA0003343926680000054
Then S10 is performed.
S7, randomly selecting a first core object o from the core object set Ω, and initializing a current cluster core object queue Ω cur ═ o }; k +1, let k'; initializing the Current Cluster Ck(Current Cluster is also referred to as Current Cluster sample set), Ck-o }; updating the unvisited classified sample set Γ, i.e. Γ ' ═ Γ - { o }, Γ ' ═ Γ '; proceed to S8.
S8, judging whether the core object queue omega cur is an empty set. If it is
Figure BDA0003343926680000051
Then the current cluster C is clusteredkFinishing the generation; updating clustersDivision, i.e. making C ═ C1,C2,...,Ck}; updating the core object set omega, i.e. omega' ═ omega-CkMaking omega be omega'; returning to S6. If it is
Figure BDA0003343926680000052
The core object set omega, i.e. omega' ═ omega-C, is updatedkAnd S9 is performed with Ω ═ Ω'.
S9, randomly taking out a second core object o ' from the current cluster core object queue omega cur, finding out all neighborhood subsample sets N e (o ') through neighborhood parameters e, enabling delta to be N e (o ') #, and updating the current cluster CkI.e. Ck'＝Ck∪Δ、Ck＝Ck'; updating the unvisited classification sample set Γ, i.e. Γ '═ Γ - Δ, Γ ═ Γ'; updating Ω cur, i.e. Ω cur ═ Ω ═ u — (Δ ═ Ω) — o ', let Ω cur ═ Ω cur'; returning to S8.
S10, calculating cluster division C ═ C1,C2,...,CkThe density center point of { C ρ } is obtained1,Cρ2,...,Cρk}。
Cluster division C ═ { Ci}＝{C1,C2,...,CkK, k is an integer of 0 or more. Calculating each cluster C in the cluster partition CiA density center point of (C), cluster CiCluster C for shortiCluster CiThe density center point of (A) is a cluster CiThe maximum density of all points in the region is denoted as CpiCluster CiThere are q points (q is a positive integer).
CiThe density calculation formula of any point in the image is as follows:
Figure BDA0003343926680000061
where ρ is a cluster CiDensity value of any one point within, dqIs the point to the cluster CiDistance to other points in the cluster, Dmax being CiThe largest distance between any two points in the distance.
Cρi＝Max{ρ1,ρ2,...,ρq}＝ρmax，ρmaxIs the cluster CiThe density center point of (a).
Sequentially finding out the density centers C of all clusters in the cluster division Cρ，Cρ＝{Cρ1,Cρ2,...,Cρk}。
S11, output cluster division C ═ C1,C2,...,Ck}, output density center point { C ρ1,Cρ2,...,Cρk}。
Cluster division C ═ { C1,C2,...,CkIs k different thunderstorm monomers with density center point { C rho }1,Cρ2,...,CρkIs the thunderstorm density center. Therefore, information such as thunderstorm comprehensive strength can be further obtained.
The following exemplifies the application of the present invention. For one strong precipitation process of Beijing in 2 evening of 8 months and 2 days in 2020, two thunderstorm processes pass through Beijing from 19 hours to 21 hours, and the process is analyzed by adopting a multi-scale thunderstorm intelligent classification and identification method based on multi-source lightning data to perform thunderstorm classification and central identification. Fig. 3 is a record of lightning of the strong precipitation by the lightning instrument, wherein "+" in fig. 3 represents lightning with positive point charge, and "-" represents lightning with negative point charge.
Two kinds of different scale information are respectively adopted for the strong precipitation process, the first kind is to cut off the multi-source lightning data flow into a data set of 5 minutes, and set the scale information of 20km (namely, more than 20km is a new thunderstorm lightning monomer), and the minimum lightning frequency of the thunderstorm is set to be 3; secondly, the multi-source lightning data stream is cut into a data set of 1 minute, 100km of scale information (namely more than 100km is a new thunderstorm lightning monomer) is set, and the minimum lightning frequency of the thunderstorm is set to be 2. The first type of thunderstorm identification is to perform thunderstorm classification and center identification every 5 minutes from 25 minutes to 21 minutes from 19 minutes to finally obtain two single thunderstorms obvious from the northwest image in the southeast direction, a short small thunderstorm monomer is formed in the middle, the thunderstorm monomer classification and the thunderstorm density center show the thunderstorm density center (including the historical thunderstorm density center) of the single thunderstorm, such as figure 4A (corresponding cut lightning data at 19 minutes and 35 minutes in Beijing time), figure 4B (corresponding cut lightning data at 19 minutes and 40 minutes in Beijing time), figure 4C (corresponding cut lightning data at 55 minutes at 19 minutes in Beijing time) and figure 4D (corresponding cut lightning data at 20 minutes in Beijing time and 05 minutes in 20 minutes), two single thunderstorms can be seen in figures 4A to 4D, a light color area in the upper right side area is a single thunderstorm, a point of a dark point connecting line is the point of the single thunderstorm density center of the single thunderstorm (including the thunderstorm density center), the lower left area, dark, is another individual thunderstorm, the dots connecting the lines of light dots indicate the center of thunderstorm density for that individual thunderstorm (including the historical center of thunderstorm density). Aiming at the process, lightning can be effectively divided into different thunderstorm monomers in terms of overall effect. Second type of thunderstorm identification, the identified thunderstorm monomer classification and thunderstorm density center in the whole chinese area at 13 minutes of 19/8/2/2020 is shown in fig. 5, wherein "+" in fig. 5 represents the position of a thunderstorm density center, the large circle near each "+" represents the member in the cluster, the very small black circle in the figure represents the noise less than or equal to MinPts, and as can be seen from fig. 5, through the multiscale thunderstorm intelligent classification and identification method based on multisource lightning data of the present invention, the thunderstorm is identified as a large monomer, and can simultaneously and efficiently identify multiple targets, and give out the thunderstorm density center, and classify some lightning detection times less than 2 as noise.
The multiscale thunderstorm intelligent classification and identification method based on multisource lightning data establishes a unique thunderstorm monomer identification method, has very good identification capability on thunderstorm monomers, can simultaneously and efficiently identify multiple targets, shows relatively excellent classification and identification results under multiscale of a large area (such as all China) and a small area (such as Beijing), provides a thunderstorm density center which is relatively in line with the physical form of the thunderstorm, hardly deviates from the thunderstorm monomers, and can remove noise of a single point, so the multiscale thunderstorm intelligent classification and identification method based on multisource lightning data can rapidly and accurately identify the thunderstorm monomers and provide the thunderstorm density center Under the complex thunderstorm condition, the method also provides a data basis for the prediction of the movement locus of the thunderstorm monomer, has an obvious effect on the identification of multiple thunderstorm monomers, and is widely applied to the fields of power grids, oil storage and the like which need lightning early warning prediction.

Claims (5)

1. A multiscale thunderstorm intelligent classification and identification method based on multisource lightning data is characterized by comprising the following steps:
s1, acquiring the lightning data, and unifying the lightning data in a preset format;
s2, cutting the unified lightning data according to a certain time interval;
s3, recording the cut lightning data as D ═ xj}＝(x1,x2,x3.....xm) J is 1,2,3,. m, m is a positive integer; setting neighborhood parameters E and scale information;
s4, initializing a core object set
Figure FDA0003343926670000011
Initializing cluster number k as 0, initializing classification sample set Γ as D, initializing cluster division
Figure FDA0003343926670000012
S5, finding out D ═ { x ] in S3jAll core objects in the } and update Ω accordingly;
s6, judging whether omega is an empty set or not, if so, judging whether omega is an empty set or not
Figure FDA0003343926670000013
Proceed to S7 if
Figure FDA0003343926670000014
Then proceed to S10;
s7, in Ω, randomly selecting a first core object o, and initializing the current cluster core object queue Ω cur ═ { o }, k ═ c+1, let k equal to k', initialize the current cluster CkUpdating Γ, namely Γ '═ Γ - { o }, Γ ═ Γ';
s8, judging whether the omega cur is an empty set, if so, determining whether the omega cur is an empty set
Figure FDA0003343926670000015
Then C iskAfter generation, C is updated, i.e. C is made to be { C {1,C2,...,CkH, update Ω, i.e., Ω' ═ Ω -CkΩ ═ Ω', return to S6; if it is
Figure FDA0003343926670000016
Then omega is updated, i.e. omega' ═ omega-CkS9 is performed, Ω ═ Ω';
s9, in Ω cur, randomly taking out a second core object o ', finding out all neighborhood subsamples N e (o ') by e, making Δ ═ N e (o ') # Γ, updating CkI.e. Ck'＝Ck∪Δ、Ck＝Ck', update Γ, i.e. Γ' ═ Γ - Δ, Γ ═ Γ ', update Ω cur, i.e. Ω cur' ═ Ω cur ═ u (Δ ═ Ω) -o ', Ω cur ═ Ω cur', return to S8;
s10, calculating the density center point of C to obtain { C rho1,Cρ2,...,Cρk}。
2. The multi-scale thunderstorm intelligent classification and identification method based on multi-source lightning data according to claim 1, wherein the step of setting MinPts is further included in S3, and the step of S5 specifically comprises the steps of:
s5.1, let j equal 1;
s5.2, calculating xjIs x is foundjIs in the neighborhood subsample set of (x)j)；
S5.3, if | N ∈ (x)j) | is not less than MinPts, xjAdding a core object set Ω, i.e., Ω' ═ Ω & { x }jH, q ═ q'; s5.4 is carried out;
s5.4, determining whether j is equal to m, if j is not equal to m, j' ═ j +1, and if j ≠ m, returning to S5.2; if j is m, S6 is performed.
3. The multi-scale thunderstorm intelligent classification and identification method based on multi-source lightning data according to claim 1, wherein the S10 is specifically: c ═ Ci}＝{C1,C2,...,CkK, calculating each cluster C in the cluster partitions CiA density center point of (C), cluster CiThe density center point of (A) is a cluster CiMaximum value of density C rho of all points in the interiori，CiQ points are total, q is a positive integer, CiThe density calculation formula of any point in the image is as follows:
Figure FDA0003343926670000021
where ρ is a cluster CiDensity value of any one point within, dqIs the point to the cluster CiDistance to other points in the cluster, Dmax being CiThe largest distance between any two points in the distance.
4. The multi-scale thunderstorm intelligent classification and identification method based on multi-source lightning data according to claim 1, characterized by further comprising S11, S11, output cluster division C ═ C1, C21,Cρ2,...,Cρk}。
5. The multi-scale thunderstorm intelligent classification and identification method based on multi-source lightning data according to claim 1, wherein the S2 is specifically: and cutting the unified lightning data according to a certain time interval, performing space projection operation on the cut lightning data, and performing S3 on the lightning data after the space projection operation as the cut lightning data.
CN202111316840.5A 2021-11-09 2021-11-09 Multiscale thunderstorm intelligent classification and identification method based on multisource lightning data Pending CN114254692A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111316840.5A CN114254692A (en) 2021-11-09 2021-11-09 Multiscale thunderstorm intelligent classification and identification method based on multisource lightning data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111316840.5A CN114254692A (en) 2021-11-09 2021-11-09 Multiscale thunderstorm intelligent classification and identification method based on multisource lightning data

Publications (1)

Publication Number Publication Date
CN114254692A true CN114254692A (en) 2022-03-29

Family

ID=80790650

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111316840.5A Pending CN114254692A (en) 2021-11-09 2021-11-09 Multiscale thunderstorm intelligent classification and identification method based on multisource lightning data

Country Status (1)

Country Link
CN (1) CN114254692A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116500703A (en) * 2023-06-28 2023-07-28 成都信息工程大学 Thunderstorm monomer identification method and device
CN117809192A (en) * 2024-03-01 2024-04-02 南京信息工程大学 Thunderstorm identification method based on DENCLUE clustering algorithm

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116500703A (en) * 2023-06-28 2023-07-28 成都信息工程大学 Thunderstorm monomer identification method and device
CN116500703B (en) * 2023-06-28 2023-09-01 成都信息工程大学 Thunderstorm monomer identification method and device
CN117809192A (en) * 2024-03-01 2024-04-02 南京信息工程大学 Thunderstorm identification method based on DENCLUE clustering algorithm
CN117809192B (en) * 2024-03-01 2024-04-26 南京信息工程大学 DENCLUE clustering algorithm-based thunderstorm identification method

Similar Documents

Publication Publication Date Title
CN109738970B (en) Method, device and storage medium for realizing lightning early warning based on lightning data mining
EP3680609A1 (en) Antenna downward inclination angle measurement method based on multi-scale deep semantic segmentation network
CN112526513B (en) Millimeter wave radar environment map construction method and device based on clustering algorithm
CN114254692A (en) Multiscale thunderstorm intelligent classification and identification method based on multisource lightning data
EP2927709A1 (en) Identifying and tracking convective weather cells
CN110555841B (en) SAR image change detection method based on self-attention image fusion and DEC
CN109946762A (en) A kind of method and system based on probability distribution Short-term Forecast precipitation
CN110751209B (en) Intelligent typhoon intensity determination method integrating depth image classification and retrieval
CN104899562A (en) Texture segmentation and fusion based radar remote-sensing image artificial building recognition algorithm
CN109668567B (en) Polarized light orientation method and system under multi-cloud condition of unmanned aerial vehicle and polarized light sensor
CN110346800A (en) A kind of thunderstorm recognition methods based on dual-polarization weather radar body total number evidence
WO2018168165A1 (en) Weather forecasting device, weather forecasting method, and program
CN108154271A (en) A kind of surface air temperature method of quality control based on spatial coherence and surface fitting
CN112084871B (en) High-resolution remote sensing target boundary extraction method based on weak supervised learning
CN115761513A (en) Intelligent remote sensing identification method for mountain large landslide based on semi-supervised deep learning
CN109948697B (en) Method for extracting urban built-up area by using multi-source data to assist remote sensing image classification
CN114137637B (en) Thunderstorm center trace aggregate probability forecasting method based on lightning and radar data
CN117029840A (en) Mobile vehicle positioning method and system
CN115828187A (en) Star-base and foundation lightning data fusion method
CN110675363A (en) Automatic calculation method of DNA index for cervical cells
Zha et al. Detecting of overshooting cloud tops via Himawari-8 imagery using dual channel multiscale deep network
CN112633347B (en) Method for analyzing wind characteristics of landing typhoon near-stratum and wind characteristics difference between landing typhoon near-stratum and clear sky
CN112668615B (en) Satellite cloud picture prediction method based on depth cross-scale extrapolation fusion
Hou et al. A new method of characterizing flow patterns of vortices and detecting the centers of vortices in a numerical wind field
CN114019255B (en) Method and system for positioning and predicting thunderstorm center

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination