CN112684454A - Track cross target association method based on sub-frequency bands - Google Patents

Track cross target association method based on sub-frequency bands Download PDF

Info

Publication number
CN112684454A
CN112684454A CN202011402804.6A CN202011402804A CN112684454A CN 112684454 A CN112684454 A CN 112684454A CN 202011402804 A CN202011402804 A CN 202011402804A CN 112684454 A CN112684454 A CN 112684454A
Authority
CN
China
Prior art keywords
sub
target
band
frequency
bands
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.)
Granted
Application number
CN202011402804.6A
Other languages
Chinese (zh)
Other versions
CN112684454B (en
Inventor
楼万翔
冯凯
孙伟平
傅仁琦
侯觉
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.)
715th Research Institute of CSIC
Original Assignee
715th Research Institute of CSIC
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 715th Research Institute of CSIC filed Critical 715th Research Institute of CSIC
Priority to CN202011402804.6A priority Critical patent/CN112684454B/en
Publication of CN112684454A publication Critical patent/CN112684454A/en
Application granted granted Critical
Publication of CN112684454B publication Critical patent/CN112684454B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)

Abstract

The invention discloses a track cross target association method based on sub-bands, which comprises the steps of firstly decomposing the whole bandwidth of a broadband into a plurality of sub-bands, then respectively carrying out beam forming processing on each sub-band, then screening out target characteristic sub-bands by utilizing acoustic energy information and continuous spectrum information of different sub-bands of each target, and carrying out track association before and after target cross by utilizing the characteristic sub-band information. The invention has the beneficial effects that: the method aims to comprehensively utilize angle information, energy information and spectrum characteristics of the target to solve the problem that the target loses or is mistakenly followed after track crossing. And screening the characteristic sub-frequency bands of the target by using the acoustic energy information and the continuous spectrum information of the target, and performing track association before and after cross tracking by using the characteristic sub-frequency band information, thereby solving the problem of cross tracking loss of the multi-target tracking track.

Description

Track cross target association method based on sub-frequency bands
Technical Field
The invention relates to the field of passive target cross tracking in sonar array signal processing, in particular to a track cross target association method based on a sub-frequency band.
Background
If the target is missed or mistracked, the functions of target positioning, movement element settlement, feature analysis and recognition, listening and the like of the sonar system are realized.
The key point of effectively tracking the track crossed target is mainly the correlation method of the tracks before and after crossing and the signal to noise ratio of the weak target in the track is improved as much as possible. If the weak target can be successfully detected and the signal-to-noise ratio of the weak target is higher after detection, the probability of successful association with the track is high; if the trajectories can be correctly correlated, the target is also tracked smoothly. Therefore, how to improve the signal-to-noise ratio of weak targets in the crossed targets and how to improve the correlation accuracy before and after the track crossing are particularly important. For passive sonar, finding a method with strong target detection capability and improving the accuracy of association can also solve part or all of the problems.
The conventional detection tracking processing of the existing passive sonar mainly has the following two problems:
conventional passive detection processing: the entire frequency band is typically divided into several frequency bands. Once the design is complete, the bin width is fixed. In practice, however, for weak targets there are often limited frequency bands (but not necessarily narrow bands). If the processing frequency band is not consistent with the target frequency band, either the signal energy is not fully utilized (the processing frequency band is smaller than the target frequency band) or more noise is introduced to cause the detection capability to be reduced (the processing frequency band is larger than the target frequency band).
Conventional passive tracking: when multi-target tracking is carried out, strong targets interfere weak targets to influence the tracking of the weak targets; when the target tracks are crossed, the conventional algorithm only uses the angle of the target for tracking, and the track is often associated with a measurement error, so that the target is missed or mistakenly tracked.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provides a track cross target association method based on a sub-frequency band.
The object of the present invention is achieved by the following technical means. A track cross target association method based on sub-bands comprises the steps of firstly decomposing the whole bandwidth of a broadband into a plurality of sub-bands, then respectively carrying out beam forming processing on each sub-band, then screening out target characteristic sub-bands by utilizing acoustic energy information and continuous spectrum information of different sub-bands of each target, and carrying out track association before and after target cross by utilizing characteristic sub-band information, so that the problem of multi-target tracking track cross tracking loss is solved.
The invention has the beneficial effects that: the method aims to comprehensively utilize angle information, energy information and spectrum characteristics of the target to solve the problem that the target loses or is mistakenly followed after track crossing. And screening the characteristic sub-frequency bands of the target by using the acoustic energy information and the continuous spectrum information of the target, and performing track association before and after cross tracking by using the characteristic sub-frequency band information, thereby solving the problem of cross tracking loss of the multi-target tracking track.
Drawings
FIG. 1 is a schematic cross-track diagram;
FIG. 2 is an azimuth history map;
FIG. 3 is a schematic diagram of target sub-band energy spillover;
the target of fig. 4 is the energy fraction in the different sub-bands.
Fig. 5 shows a azimuth history of sub-band 4.
Detailed Description
The invention will be described in detail below with reference to the following drawings:
the invention discloses a track crossing target association method based on sub-bands, which comprises the steps of firstly decomposing the whole bandwidth of a broadband into a plurality of sub-bands, then respectively carrying out beam forming processing on each sub-band, then screening out target characteristic sub-bands by utilizing sound energy information and continuous spectrum information of different sub-bands of each target, and carrying out track association before and after target crossing by utilizing the characteristic sub-band information, thereby solving the problem of multi-target tracking track crossing and tracking loss. The algorithm principle is as follows:
(1) sub-band spectral formation
The time snapshot received by the sonar array element is assumed to be x (n) ═ x1(n),…,xM(n)]FFT-converting it to frequency snapshot X (f) ═ X1(f),…,XM(f)]Wherein n is a time snapshot point, f is an FFT frequency point, and M is the number of array elements. The beam output obtained after frequency domain beamforming is B (f, θ)i)：
B(f,θ)＝w(f,θ)·X(f) (1)
Where θ is the beam number and w (f, θ) is its corresponding response vector.
Suppose the entire processing band is fmin fmax]Is divided into L sub-bands, the first sub-band having a frequency range of [ f [ ]l.min fl.max]. Performing beam forming on each sub-band to obtain beam output of each sub-band, i.e.
Figure BDA0002817489270000021
For each Bl(f, theta) square detection is carried out to finally obtain the output P of the first sub-band at the time tl(θ,t)。
All sub-frequency bands at the time t are synthesized to obtain the output of the whole processing frequency band as PFULL(θ, t), i.e.
Figure BDA0002817489270000022
(2) Characteristic sub-band screening
In the passive target detection and tracking process, the frequency spectrum characteristics of the target are not continuously stable in the whole processing frequency band. But the spectral characteristics of the target in certain sub-bands, which are defined as characteristic sub-bands, may be considered to remain stationary for a limited period of time. In a typical sub-frequency, the target may achieve a greater signal-to-noise ratio than the entire processing bin. The characteristic sub-band screening method comprises the following steps:
normalizing each sub-frequency band and the synthesized result respectively to obtain
Pl'(θ,t)＝Pl(θ,t)/max(Pl(θ,t)) (4)
P'FULL(θ,t)＝PFULL(θ,t)/max(PFULL(θ,t)) (5)
Defining the energy overflow of the ith target in the ith sub-band relative to the whole band as
Figure BDA0002817489270000031
Figure BDA0002817489270000032
Figure BDA0002817489270000033
Indicating an energy spill over for a certain period of time,
Figure BDA0002817489270000034
energy spillover representing accumulation time, where the kth time period is denoted as [ t ]i,start ti,end]Length of NtThe time length can be set according to the actual condition, thetaiIndicates that the i-th object is at tiThe beam number of the time instant. They measure the relative strength of a certain sub-band of the target in the whole processing band, and a larger value indicates that the energy of the target in the sub-band is higher in the whole processing band.
the energy ratio of each sub-band wave beam theta at the time t in the whole processing frequency band is
Figure BDA0002817489270000035
Calculating the average energy ratio of the target in each sub-frequency band in a certain time period as
Figure BDA0002817489270000036
Wherein the content of the first and second substances,
Figure BDA0002817489270000037
representing the ith sub-frequencyThe average energy ratio of the i target in the section in the k time section. The cumulative energy ratio of the target I in the ith sub-band is expressed as
Figure BDA0002817489270000038
Figure BDA0002817489270000039
Representing the energy intensity of the target in different sub-bands.
Defining the energy intensity difference of the target I and the target j in the l sub-band as
Figure BDA0002817489270000041
Namely, it is
Figure BDA0002817489270000042
When a certain sub-band satisfies the following condition
Figure BDA0002817489270000043
This sub-band is called the characteristic sub-band, where E0、Δ0And (4) selecting. Assuming that L' characteristic sub-bands are obtained by sub-band screening, the target cross-tracking process is performed in these characteristic sub-bands.
(3) Sub-band target cross-tracking processing
Assume that the time period before the intersection of two target tracks is kbeforeThe time period after the track crossing is kafterThe average energy ratio of the two target tracking front and back in the characteristic sub-frequency band is respectively
Figure BDA0002817489270000044
Figure BDA0002817489270000045
ComputingCross correlation coefficient of the target before and after tracking:
Figure BDA0002817489270000046
Figure BDA0002817489270000047
Figure BDA0002817489270000048
Figure BDA0002817489270000049
wherein-1 is not more than rhoij≤1,ρii＝1。
Respectively calculating rho according to the cross track schematic diagram13、ρ24、ρ14、ρ23Defining the correlation coefficient as eta
η＝(ρ13+ρ24)-(ρ14+ρ23) (17)
When in use
Figure BDA00028174892700000410
Wherein eta0Greater than or equal to 0, general eta0And taking zero. The greater the value of η, the greater the confidence of the association. And completing track cross tracking association.
An even linear array is supposed to be arranged, the number of array elements is 128, the array element spacing is 0.1m, the working frequency range is 1000Hz-4000Hz, and the sampling frequency is 15 kHz. For 3 existing targets, the frequency of a target 1 is 1000Hz-3800Hz, the signal-to-noise ratio is-5 dB, the initial angle is 16 degrees, and the end angle is-11 degrees; the frequency of the target 2 is 1000Hz-3800Hz, the signal-to-noise ratio is-16 dB, the starting angle is 7 degrees, and the ending angle is 7 degrees; the frequency of target 3 is 1000Hz-4000Hz, the signal-to-noise ratio is 0dB, the starting angle is-45 degrees, and the ending angle is-63 degrees. Where target 1 and target 2 begin to intersect at t-23 and end at t-47, the azimuth history is shown in fig. 2.
The process divides the whole band into 8 sub-bands and the cumulative energy spill over for each sub-band for target 1 and target 2 is shown in figure 3.
The different sub-band energies before the target 1 and target 2 trajectories cross are shown in fig. 4.
According to fig. 3 and 4, the characteristic sub-band is selected as [ 13456 ]. The azimuth course of the target in the 4 th sub-band is shown in FIG. 5
Fig. 5 compares with fig. 2, and the signal-to-noise ratio of the weak target 2 in the sub-band 4 is significantly higher than that in the full band. Firstly, calculating the association coefficients before and after the track intersection of the target 1 and the target 2 of 8 sub-frequency bands:
Figure BDA0002817489270000051
and performing association processing by using the characteristic sub-frequency band to obtain an association coefficient:
Figure BDA0002817489270000052
it can be seen from the above that, the cross processing performed by using 8 sub-bands and the characteristic sub-band can be associated successfully, but the association reliability can be improved by using the characteristic sub-band.
The calculation process is as follows:
(1) performing beam forming on the L sub-bands at the time t to obtain a sub-band spectrum Pl(theta, t) and full band beam output PFULL(θ,t)；
(2) Calculating the target before track crossing
Figure BDA0002817489270000053
And
Figure BDA0002817489270000054
screening to obtain a characteristic sub-frequency band;
(3) and calculating the correlation coefficient eta before and after the target track is crossed by utilizing the characteristic sub-frequency bands to carry out correlation judgment.
It should be understood that equivalent substitutions and changes to the technical solution and the inventive concept of the present invention should be made by those skilled in the art to the protection scope of the appended claims.

Claims (2)

1. A track cross target association method based on sub-frequency bands is characterized in that: the method comprises the steps of firstly decomposing the whole bandwidth of a broadband into a plurality of sub-frequency bands, then respectively carrying out beam forming processing on each sub-frequency band, then screening out target characteristic sub-frequency bands by utilizing acoustic energy information and continuous spectrum information of different sub-frequency bands of each target, and carrying out track association before and after target crossing by utilizing the characteristic sub-frequency band information, thereby solving the problem of multi-target tracking track crossing and tracking loss.
2. The track cross target association method based on the sub-band according to claim 1, wherein: the method specifically comprises the following steps:
(1) sub-band spectral formation
The time snapshot received by the sonar array element is assumed to be x (n) ═ x1(n),…,xM(n)]FFT-converting it to frequency snapshot X (f) ═ X1(f),…,XM(f)]Wherein n is a time snapshot point, f is an FFT frequency point, M is the number of array elements, and the output of the wave beam obtained after the wave beam of the frequency domain is formed is B (f, theta)i)：
B(f,θ)＝w(f,θ)·X(f) (1)
Wherein, theta is the beam number, and w (f, theta) is the corresponding response vector;
suppose the entire processing band is fmin fmax]Is divided into L sub-bands, the first sub-band having a frequency range of [ f [ ]l.minfl.max]Forming a beam for each sub-band to obtain a beam output for each sub-band, i.e.
Figure FDA0002817489260000011
For each Bl(f, theta) square detection is carried out to finally obtain the output P of the first sub-band at the time tl(θ,t)；
All sub-frequency bands at the time t are synthesized to obtain the output of the whole processing frequency band as PFULL(θ, t), i.e.
Figure FDA0002817489260000012
(2) Characteristic sub-band screening
The characteristic sub-band screening method comprises the following steps:
normalizing each sub-frequency band and the synthesized result respectively to obtain
Pl'(θ,t)＝Pl(θ,t)/max(Pl(θ,t)) (4)
P'FULL(θ,t)＝PFULL(θ,t)/max(PFULL(θ,t)) (5)
Defining the energy overflow of the ith target in the ith sub-band relative to the whole band as
Figure FDA0002817489260000021
Figure FDA0002817489260000022
Figure FDA0002817489260000023
Indicating an energy spill over for a certain period of time,
Figure FDA0002817489260000024
energy spillover representing accumulation time, where the kth time period is denoted as [ t ]i,start ti,end]Length of NtThe time length can be set according to the actual condition, thetaiIs shown asTarget number i at tiA beam number of a time;
the energy ratio of each sub-band wave beam theta at the time t in the whole processing frequency band is
Figure FDA0002817489260000025
Calculating the average energy ratio of the target in each sub-frequency band in a certain time period as
Figure FDA0002817489260000026
Wherein the content of the first and second substances,
Figure FDA0002817489260000027
the average energy ratio of the i number of targets in the ith sub-band in the kth time period is represented, and the accumulated energy ratio of the i number of targets in the ith sub-band is represented as
Figure FDA0002817489260000028
Figure FDA0002817489260000029
Representing the energy intensity of the target in different sub-frequency bands;
defining the energy intensity difference of the target I and the target j in the l sub-band as
Figure FDA00028174892600000210
Namely, it is
Figure FDA00028174892600000211
When a certain sub-band satisfies the following condition
Figure FDA00028174892600000212
This sub-band is called the characteristic sub-band, where E0、Δ0Selecting; assuming that L' characteristic sub-bands are obtained through sub-band screening, target cross-tracking processing is performed in the characteristic sub-bands;
(3) sub-band target cross-tracking processing
Assume that the time period before the intersection of two target tracks is kbeforeThe time period after the track crossing is kafterThe average energy ratio of the two target tracking front and back in the characteristic sub-frequency band is respectively
Figure FDA0002817489260000031
Figure FDA0002817489260000032
Calculating the cross-correlation coefficient of the target before and after tracking:
Figure FDA0002817489260000033
Figure FDA0002817489260000034
Figure FDA0002817489260000035
Figure FDA0002817489260000036
wherein-1 is not more than rhoij≤1,ρii＝1；
Respectively calculating rho according to the cross track schematic diagram13、ρ24、ρ14、ρ23Defining the correlation coefficientIs eta
η＝(ρ13+ρ24)-(ρ14+ρ23) (17)
When in use
Figure FDA0002817489260000037
Wherein eta0And the greater the eta value is, the greater the association confidence coefficient is, and the track cross tracking association is completed.
CN202011402804.6A 2020-12-04 2020-12-04 Track cross target association method based on sub-frequency bands Active CN112684454B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011402804.6A CN112684454B (en) 2020-12-04 2020-12-04 Track cross target association method based on sub-frequency bands

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011402804.6A CN112684454B (en) 2020-12-04 2020-12-04 Track cross target association method based on sub-frequency bands

Publications (2)

Publication Number Publication Date
CN112684454A true CN112684454A (en) 2021-04-20
CN112684454B CN112684454B (en) 2022-12-06

Family

ID=75447261

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011402804.6A Active CN112684454B (en) 2020-12-04 2020-12-04 Track cross target association method based on sub-frequency bands

Country Status (1)

Country Link
CN (1) CN112684454B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113239775A (en) * 2021-05-09 2021-08-10 西北工业大学 Method for detecting and extracting flight path in azimuth history map based on layered attention depth convolution neural network

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5119432A (en) * 1990-11-09 1992-06-02 Visidyne, Inc. Frequency division, energy comparison signal processing system
US6104345A (en) * 1998-04-09 2000-08-15 State Of Israel, Ministry Of Defense Armament Development Authority-Rafael Direction of arrival tracking of multiple targets
US20040030500A1 (en) * 2002-08-08 2004-02-12 Struzinski William A. Target track crossing prediction/detection
CN109188443A (en) * 2018-06-29 2019-01-11 中国船舶重工集团公司第七〇五研究所 A kind of passive target tracking method based on Interactive Multiple-Model
CN109298413A (en) * 2018-09-01 2019-02-01 哈尔滨工程大学 A kind of method that specific aim solves the problems, such as the multiple target tracking data correlation under complex electromagnetic environment
CN109444897A (en) * 2018-09-13 2019-03-08 中国船舶重工集团公司第七〇五研究所 A kind of more gusts of Data Associations based on multiple features
JP2020034363A (en) * 2018-08-29 2020-03-05 沖電気工業株式会社 Signal tracking device, signal tracking system, signal tracking method, and program

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5119432A (en) * 1990-11-09 1992-06-02 Visidyne, Inc. Frequency division, energy comparison signal processing system
US6104345A (en) * 1998-04-09 2000-08-15 State Of Israel, Ministry Of Defense Armament Development Authority-Rafael Direction of arrival tracking of multiple targets
US20040030500A1 (en) * 2002-08-08 2004-02-12 Struzinski William A. Target track crossing prediction/detection
CN109188443A (en) * 2018-06-29 2019-01-11 中国船舶重工集团公司第七〇五研究所 A kind of passive target tracking method based on Interactive Multiple-Model
JP2020034363A (en) * 2018-08-29 2020-03-05 沖電気工業株式会社 Signal tracking device, signal tracking system, signal tracking method, and program
CN109298413A (en) * 2018-09-01 2019-02-01 哈尔滨工程大学 A kind of method that specific aim solves the problems, such as the multiple target tracking data correlation under complex electromagnetic environment
CN109444897A (en) * 2018-09-13 2019-03-08 中国船舶重工集团公司第七〇五研究所 A kind of more gusts of Data Associations based on multiple features

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
SANJAY K. MEHTA等: "A modified Eckart post-beamformer filter for improved detection using broadband features", 《1996 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING CONFERENCE PROCEEDINGS》 *
楼万翔 等: "基于Hough变换的目标交叉跟踪算法", 《舰船电子工程》 *
陈卓: "不同类型水听器布阵阵元域融合可行性分析", 《电子世界》 *
饶柏斌 等: "一种基于调制频段补偿的宽带解调方法", 《声学技术》 *
高运 等: "被动声呐多目标自动检测与跟踪算法研究", 《声学与电子工程》 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113239775A (en) * 2021-05-09 2021-08-10 西北工业大学 Method for detecting and extracting flight path in azimuth history map based on layered attention depth convolution neural network

Also Published As

Publication number Publication date
CN112684454B (en) 2022-12-06

Similar Documents

Publication Publication Date Title
CN108680910B (en) Frequency modulation broadcast external radiation source radar target detection method based on waveform cognition
CN101825707B (en) Monopulse angular measurement method based on Keystone transformation and coherent integration
US9361907B2 (en) Sound signal processing apparatus, sound signal processing method, and program
CN108828566B (en) Underwater pulse signal identification method based on towed linear array
CN109324322B (en) Direction finding and target identification method based on passive phased array antenna
CN104267379A (en) Active and passive radar cooperative anti-interference method based on waveform design
JP6415288B2 (en) Radar equipment
CN111751792A (en) Passive radar signal sorting PDW preprocessing method
CN104777450B (en) A kind of two-stage MUSIC microphone array direction-finding method
CN102141610A (en) Range-Doppler spectrum-based ionized layer clutter region identification method
CN111751797B (en) Method for determining first-order and second-order echo spectrum boundary of high-frequency ground wave radar based on azimuth angle
CN112684454B (en) Track cross target association method based on sub-frequency bands
CN109839623A (en) A kind of objects outside Earth landing instrumentation radar Area Objects echo-signal ranging processing method
CN113835068A (en) Blind source separation real-time main lobe interference resisting method based on independent component analysis
Yang et al. Improved CFAR detection and direction finding on time–frequency plane with high-frequency radar
CN109655794B (en) Detection and identification method for suppressing interference by narrow-band self-defense noise
CN103308910A (en) Method for detecting offshore non-navigational state ship target by using high-frequency ground wave radar
CN109923430A (en) For carrying out the device and method of phase difference expansion
CN110221289B (en) Target detection method for three-coordinate active phased array radar
CN109814065B (en) Beam forming method based on phase factor weighting
CN109581366B (en) Discrete sidelobe clutter identification method based on target steering vector mismatch
CN111796241A (en) Side lobe suppression method for amplitude comparison between multi-beam array beams
CN109061597B (en) Ionospheric clutter suppression method based on blind source separation and time-frequency ridge domain filtering
US8199053B2 (en) Three dimensional RF search system
KR101801325B1 (en) Radar apparatus based on virtual channel and method for generating virtual channel using the same

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