CN114943252B - Multi-rotor-block combination feature extraction method in multi-rotor-wing unmanned aerial vehicle recognition - Google Patents

Multi-rotor-block combination feature extraction method in multi-rotor-wing unmanned aerial vehicle recognition Download PDF

Info

Publication number
CN114943252B
CN114943252B CN202210548932.4A CN202210548932A CN114943252B CN 114943252 B CN114943252 B CN 114943252B CN 202210548932 A CN202210548932 A CN 202210548932A CN 114943252 B CN114943252 B CN 114943252B
Authority
CN
China
Prior art keywords
unmanned aerial
aerial vehicle
subblock
sub
matrix
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.)
Active
Application number
CN202210548932.4A
Other languages
Chinese (zh)
Other versions
CN114943252A (en
Inventor
周代英
廖阔
沈晓峰
冯健
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.)
University of Electronic Science and Technology of China
Original Assignee
University of Electronic Science and Technology of China
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 University of Electronic Science and Technology of China filed Critical University of Electronic Science and Technology of China
Priority to CN202210548932.4A priority Critical patent/CN114943252B/en
Publication of CN114943252A publication Critical patent/CN114943252A/en
Application granted granted Critical
Publication of CN114943252B publication Critical patent/CN114943252B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/08—Feature extraction
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/12—Classification; Matching
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems

Landscapes

  • Radar Systems Or Details Thereof (AREA)

Abstract

The invention belongs to the technical field of unmanned aerial vehicle identification, and particularly relates to a multi-rotor-wing unmanned aerial vehicle identification multi-rotor-wing combined feature extraction method. The method comprises the steps of firstly utilizing a plurality of windows with different sizes to carry out sub-block division on a time-frequency spectrogram of radar echo data of an unmanned aerial vehicle target, then extracting corresponding sub-block change characteristics of each type of sub-block time-frequency spectrogram, combining to form a characteristic vector, and identifying the unmanned aerial vehicle target. Due to the fact that the change information of the sub-blocks with different sizes is utilized, the characteristics which are more stable to the tiny change of the radar echo data can be extracted, and therefore the recognition rate of the target is improved. The effectiveness of the method is verified according to simulation experiment results of the 4-type multi-rotor unmanned aerial vehicle.

Description

Multi-model sub-block combination feature extraction method in multi-rotor unmanned aerial vehicle recognition
Technical Field
The invention belongs to the technical field of unmanned aerial vehicle identification, and particularly relates to a multi-rotor unmanned aerial vehicle identification multi-rotor block combination feature extraction method.
Background
With the rapid development of the unmanned aerial vehicle technology, great security threats are brought in military and civil aspects, and how to effectively monitor and manage the unmanned aerial vehicle, even to destroy the unmanned aerial vehicle attacking one party, is a major problem to be solved at present. Therefore, accurately identifying the type of the unmanned aerial vehicle has very important significance for national security defense.
At present, a subspace method is a traditional target identification method, a transformation subspace matrix is mainly established by utilizing a training data set according to a determined criterion, classification features of targets are directly extracted from data, for unmanned aerial vehicle targets, difference information of different targets mainly comprises micro-motion information in radar echo data, and due to modeling characteristics of a subspace, the micro-motion information in the echo data cannot be fully utilized, so that the target identification rate of the conventional subspace identification method has room for further improvement.
Disclosure of Invention
The invention aims to provide a method for extracting multi-type sub-block combined features in multi-rotor unmanned aerial vehicle identification.
The technical scheme of the invention is as follows:
a multi-rotor-block combined feature extraction method in multi-rotor unmanned aerial vehicle identification comprises the following steps:
s1, defining the acquired radar echo training data sequence of the multi-rotor unmanned aerial vehicle as an n-dimensional column vector x ij ，i＝1,2,…g，j＝1,2…N i Where i represents the unmanned aerial vehicle class, g represents the number of classes, N i Representing the training sample number of the ith unmanned aerial vehicle target, the total training sample number is
Figure BDA0003653651890000011
S2, training sample data x ij Performing short-time Fourier transform to obtain time-frequency spectrogram A ij ：
A ij ＝[a ij,km ] K×M (1)
Wherein, a ij,km Represents a time-frequency spectrogram A ij The elements in (A) are K =1,2, \8230A, K, M =1,2, \8230A, M, K is a row subscript indicating a frequency change direction, M is a column subscript indicating a time change direction, and K is A ij M is A ij The number of columns;
s3, time setting spectrogram A ij Performing multi-type sub-block division:
map A ij Dividing into sub-blocks of 3x3, 4x4 and 5x5 size to obtain 3 sub-block sets S ij,3×3 、S ij,4×4 And S ij,5×5 ：
Figure BDA0003653651890000021
/>
Figure BDA0003653651890000022
Figure BDA0003653651890000023
Figure BDA0003653651890000024
Figure BDA0003653651890000025
Figure BDA0003653651890000026
Wherein the content of the first and second substances,
Figure BDA0003653651890000027
subblock set S representing a size of 3 × 3 ij,3×3 R of (1) 1 Line l 1 A sub-block matrix of columns,
Figure BDA0003653651890000028
represents a subblock matrix pick>
Figure BDA0003653651890000029
The element of (1), R 1 Denotes S ij,3×3 Number of lines, L 1 Denotes S ij,3×3 The number of columns;
Figure BDA00036536518900000210
subblock set S representing a size of 4x4 ij,4×4 R of (1) 2 Line l 2 The subblock matrix of a column, according to which the value of the subblock is selected>
Figure BDA00036536518900000211
Represents a subblock matrix pick>
Figure BDA00036536518900000212
The element of (1), R 2 Denotes S ij,4×4 Number of lines, L 2 Denotes S ij,4×4 The number of columns; />
Figure BDA00036536518900000213
Subblock set S representing a size of 5 × 5 ij,5×5 R of (1) 3 Line l 3 A subblock matrix of columns, <' >>
Figure BDA0003653651890000031
Represents a subblock matrix pick>
Figure BDA0003653651890000032
The element of (1), R 3 Denotes S ij,5×5 Number of lines, L 3 Denotes S ij,5×5 The number of columns;
s4, calculating the average value of each type of subblock
Figure BDA0003653651890000033
And &>
Figure BDA0003653651890000034
Figure BDA0003653651890000035
Figure BDA0003653651890000036
Figure BDA0003653651890000037
S5, extracting the combination characteristics of the multiple types of subblocks:
by mean of subblocks
Figure BDA0003653651890000038
And &>
Figure BDA0003653651890000039
As pixel values, the amount of change in the lateral and longitudinal directions of each type of sub-block image is calculated: />
Figure BDA00036536518900000310
Figure BDA00036536518900000311
Figure BDA00036536518900000312
Figure BDA00036536518900000313
Figure BDA00036536518900000314
Figure BDA00036536518900000315
Wherein the content of the first and second substances,
Figure BDA00036536518900000316
and &>
Figure BDA00036536518900000317
The amount of change in the lateral and longitudinal directions are represented, respectively, whereby the magnitude and direction angle of the change are calculated:
Figure BDA00036536518900000318
Figure BDA0003653651890000041
Figure BDA0003653651890000042
Figure BDA0003653651890000043
Figure BDA0003653651890000044
Figure BDA0003653651890000045
wherein, the first and the second end of the pipe are connected with each other,
Figure BDA0003653651890000046
and &>
Figure BDA0003653651890000047
Represents the magnitude of the change amount>
Figure BDA0003653651890000048
And &>
Figure BDA0003653651890000049
Representing the direction angle of the variation, and forming the amplitude and direction angle of the variation in each type of sub-block diagram into a variation characteristic matrix H ij,3×3 、H ij,4×4 And H ij,5×5 ：
Figure BDA00036536518900000410
Figure BDA00036536518900000411
/>
Figure BDA00036536518900000412
S6, changing characteristic matrix H of each type of sub-block diagram ij,3×3 、H ij,4×4 And H ij,5×5 Composing a combined feature vector v by rows ij ：
Figure BDA00036536518900000413
Wherein v is ij I.e. the extracted features, h ij,1,3×3 、
Figure BDA00036536518900000414
Are respectively a matrix H ij,3×3 Middle 1 st and R 1 Line, h ij,1,4×4 、/>
Figure BDA0003653651890000051
Are respectively a matrix H ij,4×4 Middle 1 st and R 2 Line, h ij,1,5×5 、/>
Figure BDA0003653651890000052
Are respectively a matrix H ij,5×5 Middle 1 st and R 3 And (6) rows.
After the characteristics are obtained, the target can be classified and identified by using a nearest neighbor classifier.
The method has the advantages that on one hand, the micro-motion information corresponding to the target is utilized explicitly through time-frequency analysis, and in addition, the combined characteristics formed by the multi-type sub-blocks are adopted, so that the influence of micro-changes in data on the identification performance is reduced, the characteristics which are more stable to the micro-changes of radar echo data can be extracted, and the identification rate of the target is improved.
Detailed Description
The following simulations were combined to demonstrate the effectiveness and progress made by the present invention:
adopt 4 types of unmanned aerial vehicle to simulate, including three rotor unmanned aerial vehicle, four rotor unmanned aerial vehicle, six rotor unmanned aerial vehicle, eight rotor unmanned aerial vehicle, its simulation parameter is shown as table 1. The simulated radar parameters comprise: the radar carrier frequency is 24GHz; the pulse repetition frequency is 100KHz; the distance between the target and the radar is 200m; the pitching angle and the azimuth angle of the unmanned aerial vehicle relative to the radar are respectively 10 degrees and 30 degrees.
Each type of target records 10s of radar echo signals and divides the radar echo signals into segments with fixed length of 0.05s (at least including one rotation period), the overlap between the segments is 50%, each segment includes 0.05 × 100000=5000 radar echo sampling data points, and each type has 400 segments. In 400 segments, 200 segments are randomly selected as a training data set, and the remaining 200 segments are used as a test data set, so that the training data set of the 4 types of targets totally comprises 800 segments, and the test data set comprises 800 segments. For the selected training data set, the multi-type sub-blocks are extracted by using the method to form feature vectors, a classification and identification experiment is carried out on the unmanned aerial vehicle by using a nearest neighbor classifier, and the average correct identification rate of the 4 types of multi-rotor unmanned aerial vehicles reaches 98%. Wherein the signal-to-noise ratio is 15dB.
Table 1 simulation parameters of four unmanned aerial vehicles
Figure BDA0003653651890000053
/>

Claims (1)

1. The utility model provides a many types of subblocks combination feature extraction method in many rotor unmanned aerial vehicle discernments which characterized in that includes the following steps:
s1, defining an acquired radar echo training data sequence of the multi-rotor unmanned aerial vehicle as an n-dimensional column vector x ij ，i＝1,2,…g，j＝1,2…N i Where i represents the unmanned aerial vehicle class, g represents the number of classes, N i The training sample number of the ith unmanned aerial vehicle target is represented, and the total training sample number is
Figure FDA0003653651880000011
S2, training sample data x ij Performing short-time Fourier transform to obtain time-frequency spectrogram A ij ：
A ij ＝[a ij,km ] K×M
Wherein, a ij,km Represents a time-frequency spectrogram A ij The elements in (A) are K =1,2, \8230A, K, M =1,2, \8230A, M, K is a row subscript indicating a frequency change direction, M is a column subscript indicating a time change direction, and K is A ij Number of rows of (M) is A ij The number of columns;
s3, time setting spectrogram A ij Performing multi-type subblock division:
map A ij Dividing into sub-blocks of 3x3, 4x4 and 5x5 size to obtain 3 sub-block sets S ij,3×3 、S ij,4×4 And S ij,5×5 ：
Figure FDA0003653651880000012
Figure FDA0003653651880000013
Figure FDA0003653651880000014
Figure FDA0003653651880000015
Figure FDA0003653651880000021
Figure FDA0003653651880000022
Wherein, the first and the second end of the pipe are connected with each other,
Figure FDA0003653651880000023
to representSub-block set with size 3x 3->
Figure FDA0003653651880000024
R of (1) 1 Line l 1 A subblock matrix of columns, <' >>
Figure FDA0003653651880000025
Represents a subblock matrix pick>
Figure FDA0003653651880000026
Element of (1), R 1 Denotes S ij,3×3 Number of lines, L 1 Denotes S ij,3×3 The number of columns; />
Figure FDA0003653651880000027
Subblock set S representing a 4x4 size ij,4×4 R of (1) 2 Line l 2 The subblock matrix of a column, according to which the value of the subblock is selected>
Figure FDA0003653651880000028
Representing a sub-block matrix>
Figure FDA0003653651880000029
The element of (1), R 2 Denotes S ij,4×4 Number of lines, L 2 Denotes S ij,4×4 The number of columns; />
Figure FDA00036536518800000210
Subblock set S representing a size of 5x5 ij,5×5 R of (1) 3 Line l 3 The subblock matrix of a column, according to which the value of the subblock is selected>
Figure FDA00036536518800000211
Represents a subblock matrix pick>
Figure FDA00036536518800000212
The element of (1), R 3 Denotes S ij,5×5 Number of lines, L 3 Denotes S ij,5×5 Number of columns of；
S4, calculating the average value of each type of subblock
Figure FDA00036536518800000213
And &>
Figure FDA00036536518800000214
Figure FDA00036536518800000215
Figure FDA00036536518800000216
Figure FDA00036536518800000217
S5, extracting the combination characteristics of the multiple types of subblocks:
by mean of sub-blocks
Figure FDA00036536518800000218
And &>
Figure FDA00036536518800000219
As pixel values, the amount of change in the lateral and longitudinal directions of each type of sub-block image is calculated:
Figure FDA00036536518800000220
Figure FDA00036536518800000221
Figure FDA0003653651880000031
Figure FDA0003653651880000032
Figure FDA0003653651880000033
Figure FDA0003653651880000034
wherein the content of the first and second substances,
Figure FDA0003653651880000035
and &>
Figure FDA0003653651880000036
The amount of change in the lateral and longitudinal directions are represented, respectively, whereby the magnitude and direction angle of the change are calculated:
Figure FDA0003653651880000037
Figure FDA0003653651880000038
Figure FDA0003653651880000039
Figure FDA00036536518800000310
Figure FDA00036536518800000311
Figure FDA00036536518800000312
/>
wherein, the first and the second end of the pipe are connected with each other,
Figure FDA00036536518800000313
and &>
Figure FDA00036536518800000314
Represents the magnitude of the change, is>
Figure FDA00036536518800000315
And &>
Figure FDA00036536518800000316
Representing the direction angle of the variation, and forming the amplitude and direction angle of the variation in each type of sub-block diagram into a variation characteristic matrix H ij,3×3 、H ij,4×4 And H ij,5×5 ：
Figure FDA00036536518800000317
Figure FDA0003653651880000041
Figure FDA0003653651880000042
S6, changing characteristic matrix H of each type of sub-block diagram ij,3×3 、H ij,4×4 And H ij,5×5 Composing a combined feature vector v by rows ij ：
Figure FDA0003653651880000043
Wherein v is ij I.e. the extracted features, h ij,1,3×3 、
Figure FDA0003653651880000044
Are respectively a matrix H ij,3×3 Middle 1 st and R 1 Line, h ij,1,4×4 、/>
Figure FDA0003653651880000045
Are respectively a matrix H ij,4×4 Middle 1 st and R 2 Line, h ij,1,5×5 、/>
Figure FDA0003653651880000046
Are respectively a matrix H ij,5×5 Middle 1 st and R 3 And (6) a row. />
CN202210548932.4A 2022-05-20 2022-05-20 Multi-rotor-block combination feature extraction method in multi-rotor-wing unmanned aerial vehicle recognition Active CN114943252B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210548932.4A CN114943252B (en) 2022-05-20 2022-05-20 Multi-rotor-block combination feature extraction method in multi-rotor-wing unmanned aerial vehicle recognition

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210548932.4A CN114943252B (en) 2022-05-20 2022-05-20 Multi-rotor-block combination feature extraction method in multi-rotor-wing unmanned aerial vehicle recognition

Publications (2)

Publication Number Publication Date
CN114943252A CN114943252A (en) 2022-08-26
CN114943252B true CN114943252B (en) 2023-04-07

Family

ID=82908524

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210548932.4A Active CN114943252B (en) 2022-05-20 2022-05-20 Multi-rotor-block combination feature extraction method in multi-rotor-wing unmanned aerial vehicle recognition

Country Status (1)

Country Link
CN (1) CN114943252B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111273285A (en) * 2020-03-06 2020-06-12 电子科技大学 Micro Doppler spectrum correlation matrix characteristic extraction method for multi-rotor unmanned aerial vehicle
CN112162273A (en) * 2020-09-25 2021-01-01 电子科技大学 Singular vector-based multi-rotor unmanned aerial vehicle physical parameter extraction method
WO2021036286A1 (en) * 2019-08-30 2021-03-04 华为技术有限公司 Target behavior recognition method, apparatus and radar system
CN113269203A (en) * 2021-05-17 2021-08-17 电子科技大学 Subspace feature extraction method for multi-rotor unmanned aerial vehicle recognition
CN113925495A (en) * 2021-10-20 2022-01-14 福建工程学院 Arteriovenous fistula abnormal tremor signal identification system and method combining statistical learning and time-frequency analysis
CN113988119A (en) * 2021-10-14 2022-01-28 电子科技大学 Block probability distribution feature extraction method in multi-rotor unmanned aerial vehicle recognition

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021036286A1 (en) * 2019-08-30 2021-03-04 华为技术有限公司 Target behavior recognition method, apparatus and radar system
CN111273285A (en) * 2020-03-06 2020-06-12 电子科技大学 Micro Doppler spectrum correlation matrix characteristic extraction method for multi-rotor unmanned aerial vehicle
CN112162273A (en) * 2020-09-25 2021-01-01 电子科技大学 Singular vector-based multi-rotor unmanned aerial vehicle physical parameter extraction method
CN113269203A (en) * 2021-05-17 2021-08-17 电子科技大学 Subspace feature extraction method for multi-rotor unmanned aerial vehicle recognition
CN113988119A (en) * 2021-10-14 2022-01-28 电子科技大学 Block probability distribution feature extraction method in multi-rotor unmanned aerial vehicle recognition
CN113925495A (en) * 2021-10-20 2022-01-14 福建工程学院 Arteriovenous fistula abnormal tremor signal identification system and method combining statistical learning and time-frequency analysis

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
Zhengjia Xu 等.Modeling and Performance Analysis of Opportunistic Link Selection for UAV Communication.Sensors.2021,第21卷(第2期),1-29. *
李粮余.基于微多普勒谱图的多旋翼无人机识别.中国优秀硕士学位论文全文数据库工程科技Ⅱ辑.2022,(第01期),C031-618. *
章鹏飞 ; 李刚 ; 霍超颖 ; 殷红成 ; .基于双雷达微动特征融合的无人机分类识别.雷达学报.2018,第7卷(第05期),557-564. *

Also Published As

Publication number Publication date
CN114943252A (en) 2022-08-26

Similar Documents

Publication Publication Date Title
CN109086700B (en) Radar one-dimensional range profile target identification method based on deep convolutional neural network
CN108229404A (en) A kind of radar echo signal target identification method based on deep learning
CN107169492B (en) Polarized SAR target detection method based on FCN-CRF master-slave network
CN107392122B (en) Polarimetric SAR image target detection method based on multi-polarimetric feature and FCN-CRF fusion network
CN104299232B (en) SAR image segmentation method based on self-adaptive window directionlet domain and improved FCM
CN107563433A (en) A kind of infrared small target detection method based on convolutional neural networks
CN113395280B (en) Anti-confusion network intrusion detection method based on generation countermeasure network
CN104215935A (en) Weighted decision fusion based radar cannonball target recognition method
CN111539488B (en) Imaging distortion resisting target classification method for ultra-narrow pulse radar under complex dynamic track
CN113988119B (en) Block probability distribution feature extraction method in multi-rotor unmanned aerial vehicle recognition
CN105160353A (en) Polarimetric SAR data ground object classification method based on multiple feature sets
CN110232371A (en) High-precision HRRP Radar Multi Target recognition methods based on small sample
CN113269203B (en) Subspace feature extraction method for multi-rotor unmanned aerial vehicle recognition
CN112765767A (en) Rock-soil body parameter random field modeling method considering rotation effect
CN114943252B (en) Multi-rotor-block combination feature extraction method in multi-rotor-wing unmanned aerial vehicle recognition
CN107390187A (en) Clutter suppression method based on Gaussian mixtures
CN104077610B (en) The method of the SAR image target recognition of two-dimension non linearity projection properties
CN109360231A (en) Based on the Sea Ice Remote Sensing image simulation method for dividing shape depth convolution to generate confrontation network
CN112766381A (en) Attribute-guided SAR image generation method under limited sample
CN110516727B (en) Hyperspectral image classification method based on FPGA (field programmable Gate array) depth edge filter
CN111325158A (en) CNN and RFC-based integrated learning polarized SAR image classification method
CN114936576B (en) Subblock differential coding distribution characteristic extraction method in multi-rotor unmanned aerial vehicle identification
CN115496987A (en) Small sample SAR target identification method based on deep double-similarity measurement network
Wang et al. Universal adversarial perturbation for remote sensing images
CN115795299A (en) Coding feature extraction method in multi-rotor unmanned aerial vehicle recognition

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
GR01 Patent grant
GR01 Patent grant