CN116520311A - GLMB-based adaptive track initiation method - Google Patents

GLMB-based adaptive track initiation method Download PDF

Info

Publication number
CN116520311A
CN116520311A CN202310493470.5A CN202310493470A CN116520311A CN 116520311 A CN116520311 A CN 116520311A CN 202310493470 A CN202310493470 A CN 202310493470A CN 116520311 A CN116520311 A CN 116520311A
Authority
CN
China
Prior art keywords
target
measurement
density
track
new
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
CN202310493470.5A
Other languages
Chinese (zh)
Inventor
国强
卢宇翀
王亚妮
戚连刚
卢芳葳
黄帅
卡柳日内.尼古拉
任海宁
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.)
Harbin Engineering University
Original Assignee
Harbin Engineering University
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 Harbin Engineering University filed Critical Harbin Engineering University
Priority to CN202310493470.5A priority Critical patent/CN116520311A/en
Publication of CN116520311A publication Critical patent/CN116520311A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/66—Radar-tracking systems; Analogous systems
    • G01S13/72—Radar-tracking systems; Analogous systems for two-dimensional tracking, e.g. combination of angle and range tracking, track-while-scan radar
    • G01S13/723—Radar-tracking systems; Analogous systems for two-dimensional tracking, e.g. combination of angle and range tracking, track-while-scan radar by using numerical data
    • G01S13/726—Multiple target tracking
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
    • G01S13/50—Systems of measurement based on relative movement of target
    • G01S13/58—Velocity or trajectory determination systems; Sense-of-movement determination systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • 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
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Mathematical Physics (AREA)
  • Mathematical Optimization (AREA)
  • Theoretical Computer Science (AREA)
  • Computational Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Probability & Statistics with Applications (AREA)
  • Operations Research (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Algebra (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

The invention belongs to the technical field of multi-target tracking, and particularly relates to a self-adaptive track initiation method based on GLMB.

Description

GLMB-based adaptive track initiation method
Technical Field
The invention belongs to the technical field of multi-target tracking, and particularly relates to a self-adaptive track initiation method based on GLMB (Generalized Labeled Multi-Bernoulli, generalized label Bernoulli).
Background
Through decades of development of Multi-Target tracking (MTT) technology, two sets of algorithm systems are formed at present, one is a traditional MTT algorithm, a Multi-Target tracking problem is decomposed into a plurality of single-Target tracking problems through a Data Association (DA) algorithm, and as the DA algorithm has the characteristic of combined explosion, the matching combination required to enumerate is multiplied along with the increase of the number of targets, so that the algorithm loses real-time performance; the other is a multi-target tracking algorithm based on random finite modeling, by modeling multi-target states and multi-target measurement as a random finite set (Random Finite Set, RFS), mechanisms such as track initiation, suspension and the like can be naturally described, and the measurement-track association process can be completely avoided, so that extensive attention and research of students are brought about due to systematicness and scientificity of RFS theory.
In a multi-target tracking process, metrology is the only source of information, but in RFS-based MTT algorithms, the general nascent target density is given by a priori information, typically the nascent target is fixed in several locations, and the nascent target delivery parameters are given by a priori information. In a real detection scenario, the a priori information required for a new target is unknown, and may occur anywhere within the detection area, against which a standard GLMB filter would not be able to initiate a track correctly.
Aiming at the problems, the invention provides a track self-adaptive starting algorithm suitable for a GLMB filter, and the track starting process can be completed only according to measurement information under the condition that the position of a new target is uncertain.
Disclosure of Invention
In order to overcome the problems in the prior art, the invention provides a self-adaptive track starting method based on GLMB.
The technical scheme for solving the technical problems is as follows:
the invention provides a self-adaptive track starting method based on GLMB, which comprises the following steps:
converting the polar coordinate system measurement data received by the radar into rectangular coordinate system down-conversion measurement data;
according to the prior information, correlating the single target density of the new target with the single target density of the survival target to obtain multi-target prediction probability density;
filtering clutter in the measured data by a speed screening rule and a Doppler information screening rule;
updating the multi-target posterior probability density by adopting a sequential filtering mode;
the probability of target measurement is calculated through posterior probability density in the updating process, the survival target and the new target are distinguished through the probability of target measurement, and the measurement related to the new target is reserved for the new track at the next moment.
Further, according to the prior information, the single target density of the new target and the single target density of the survival target are correlated to obtain the multi-target prediction probability density, and the method specifically comprises the following steps:
according to the new generation component calculated at the previous moment, calculating to obtain the single target density of the new generation target;
according to posterior information transmitted at the previous moment, calculating to obtain single target density of the survival target;
and connecting the single target density of the new target and the single target density of the survival target in parallel to obtain the multi-target prediction probability density.
Further, correlating the single target density of the new target with the single target density of the surviving target to obtain a multi-target prediction probability density, which further includes: and calculating the initial running state of the new target score according to the hidden speed information in the Doppler information.
Further, the multi-objective posterior probability density:
in the method, in the process of the invention,mean, variance and weight of the loss measurements are represented;representing the mean, variance and weight of the position measurements; θ (l) represents the track association map with label l; delta (θ (l)) is a delta function, when θ (l) =0, delta (θ (l))=1 indicates that the measurement is not associated with the track, and indicates that the track has missed detection; conversely, θ (l) noteq0, δ (θ (l))=0 indicates that the track and the measurement information are updated, indicating that the track is normal.
Further, the probability ρ (z) of the target measurement is:
wherein,,representing a measurement set Z at time k k Correlation probability of medium survival target i and measurement, otherwise, 1-p i Representing a measurement set Z at time k k The associated probabilities of the nascent object and clutter; p is p D,k Probability for sensor detection; z k For the measurement point at time k, < >>Gaussian density, H, representing mean m, variance P k For the observation matrix, R is the position measurement noise covariance, h (·) is defined as follows:
in the above formula, (x, y) represents the target position, (x) s ,y s ) Representing the sensor position.
Compared with the prior art, the invention has the following technical effects:
according to the invention, all possible tracks are traversed according to measurement data at two continuous moments, the measurement data is subjected to preliminary screening through a speed screening rule and a Doppler information screening rule to remove most of clutter, a target measurement probability is calculated according to posterior probability to distinguish measurement sources, whether measurement is from a new target or a survival target can be distinguished according to the target measurement probability, the initial running state of the new target is calculated by utilizing the implicit speed information in Doppler measurement, the next prediction process is carried out, and the track starting function is completed.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a block diagram of a self-adaptive track initiation GLMB algorithm of the present invention;
FIG. 2 is a diagram of the actual motion trajectory of the inventive object;
FIG. 3 is a schematic diagram of the inventive sensor measurement;
FIG. 4 is a graph of GLMB tracking effects for the adaptive initiation of the inventive track;
FIG. 5 is a graph of GLMB tracking effects of the present invention;
FIG. 6 is a schematic diagram of OSPA distance according to the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the present invention to achieve the preset purpose, the following detailed description is given below of the specific implementation, structure, features and effects of the technical solution according to the present invention with reference to the accompanying drawings and preferred embodiments. The particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
Referring to fig. 1 to 6, the present invention proposes a GLMB-based adaptive track initiation method, which can estimate the motion state of a target only by measuring information at the previous time, incorporate the motion state into the prediction process of a filtering algorithm, and remove unreasonable new components in the branch clipping process, thereby effectively reducing the occurrence probability of false short and small tracks and improving the performance of a tracker.
The adaptive track starting method based on GLMB specifically comprises the following steps:
and 1, carrying out coordinate transformation on the measurement data of the radar.
The polar coordinate measurement data of the radar is [ r ] z ,θ z ]，r z For measuring distance in polar coordinates, θ z Is the azimuth angle under polar coordinates; measuring distance r z ＝r+v r Azimuth angle θ z ＝θ+v θ Wherein r represents the true distance between the sensor and the tracking target, v r Represents distance noise, θ represents true azimuth, v θ Representing angular noise.
Classical measurement is converted into:
wherein x is z Representing the x-axis coordinate of the converted rectangular coordinate system; y is z Representing the transformed y-axis coordinates.
Due to the presence of noise, an estimation result with errors is given after nonlinear conversion, and the accurate compensation procedure of the errors depends on the angle noise v θ Let the angle noise v θ Can be explained by using a symmetric probability density function, which is symmetric due to the angle noise, a desired value can be obtained：
E(sin v θ )＝0
E represents the expectation that the Sin function is an odd function, theorem: the probability density is an odd function and the expected value is 0. The formula describes the theorem described above.
The desire by the above formula can be obtained:
wherein ε θ Is the azimuth compensation factor, when epsilon θ When ε is equal to ε, a bias is expected to occur θ Not deflected can be given when not 0:
wherein x is m Representing the x-axis coordinate and y-axis coordinate in a rectangular coordinate system m Representing the y-axis coordinate in a rectangular coordinate system.
The covariance of the position component corresponding to the measuring point is as follows:
rm is the covariance of the position components, rm is a matrix of 2 x 2, and for convenience of the calculation process for representing the four position values, specific parameters are as follows:
wherein the compensation factor epsilon θ ＝E(cosv θ )，ε′ θ ＝E(cos2v θ )。
And 2, according to the prior information, correlating the single target density of the newly generated target with the single target density of the survival target to obtain the multi-target prediction probability density.
Since the value of the new component is calculated from the measured value at the previous time, the new component cannot be directly connected in parallel with the survival component, and additional state prediction is required for the new component.
Obtaining multi-target prediction probability density according to prior information:
wherein X represents a set of target states, and delta (X) represents a tag dissimilarity indicator, defined as
I.e. when the labels of the elements in set X differ, Δ (X) =1; conversely, Δ (X) =0. Delta (·) represents the dirac delta function, defined as: />L(X + ) Representing a label projection function, +.>A survival tag->Weight of->Representing the new born tag->Is a weight of (2). />And p +，B (. L) represents viable target density and nascent target density, respectively, for a given tag set L, p S Representing the probability of survival of the target at the next time.
Wherein, the single target density of the survival targets is:
wherein,,
in the above-mentioned method, the step of,gaussian density representing mean m, variance P, +.>Gaussian component weights for surviving targets; f is a state transition matrix, Q is the covariance of noise; />Is a scaling factor.
The single target density of the nascent target is:
in the above-mentioned method, the step of,gaussian density representing mean m, variance P, +.>The gaussian component weights for the nascent object. F is a state transition matrix, Q is the covariance of noise; />The probability of existence is assigned to the new target.
The new target needs to estimate the new target motion state using the new target parameters calculated in step 3 at the previous time.
Step 3, filtering clutter in the measurement data at two continuous moments through a speed screening rule and a Doppler information screening rule; and calculating the initial running state of the new target score according to the hidden speed information in the Doppler information.
Let the target estimation state at k time bem is the target estimated state mean value, and the measurement set at the corresponding moment is Z k ＝{Y k ,D k (in which the position measurement set->Doppler measurement set-> Indicating doppler measurements. Since the position measurement set is in the form of polar coordinates, it is converted into rectangular coordinates, and +.> And->Radial distance and azimuth angle measured by radar, respectively, < >>And->The x and y coordinates after coordinate conversion.
Here, one measurement value is taken from each of the conversion measurement sets of k-1 and k successive timesAnd screening the measurement by using the speed information and the Doppler information respectively, wherein the screening rule is as follows:
the speed information screening rule is as follows:
the Doppler information screening rule is as follows:
wherein I II 2 Representing the 2-norm of the vector, Δt representing the time interval between the two instants k-1 and k, v 1 、v 2 Respectively representing the minimum speed and the maximum speed of the detection target; sigma (sigma) d Is the measurement noise of Doppler information.
The velocity of motion during doppler information screening can be calculated by:
after measurement screening, calculating the state vector and covariance of the new track at the moment k through a subsequent algorithm, wherein the calculation process is as follows:
mean value required by new targetSum of covariance->The method comprises the following steps of:
the position component of the mean value is directly measured by the measurement value:
the corresponding position component covariance is:
the velocity component of the mean value isThe corresponding covariance is:
the Doppler measurement equation is:
the above formula was reconstructed by observation:
wherein,,
in the above, n d,k Representing the noise of the doppler observation,
estimating velocity components using a linear minimum mean square error criterionAnd velocity component covariance +.>The method can obtain the following steps:
the velocity component covariance is:
wherein,, is the position of the Doppler radar.
The new component weights required in the prediction step are set to:
and 4, introducing Doppler measurement, and updating the posterior probability density by adopting a sequential filtering mode.
Since doppler radar has unique doppler measurements (radial velocity) compared to the measurement data of conventional radar, sequential filtering is used to introduce doppler measurements during delta-GLMB update.
The specific implementation steps are as follows:
the delta-GLMB update posterior probability density is:
wherein θ (l) represents a track association map with label l; delta (θ (l)) is a german function for distinguishing between missed detection targets and normal detection, when θ (l) =0, delta (θ (l))=1 indicates that the measurement is not associated with the track, and indicates that the track has missed detection; conversely, θ (l) noteq0, δ (θ (l))=0 indicates track and measurement informationUpdated, indicating that the track is normal.Gaussian density representing mean m, variance P, +.>For the Gaussian component weight of the object, l represents the label information corresponding to the component, +.>Representing the measurement point corresponding to the component.
The updated branch weights are:
the single target normalization constant is:
when θ (l) =0, the track is missed and the measurement is lostThe parameters of (2) are given by:
when θ (l) noteq0, position measurementThe parameters are obtained by processing position measurement and Doppler measurement, and the specific method is as follows:
firstly, using position measurement to update state:
wherein H is c Representing the matrix of observations of the object,representing likelihood functions, p D K is the detection probability of radar yp And S is yp The gain and innovation covariance of the position measurement are respectively represented, and the corresponding calculation process is as follows:
compared with the traditional filter, in order to effectively utilize Doppler information, a sequential filtering mode is adopted, namely, firstly, position information is utilized to update the state, then Doppler measurement is utilized to further update the state, after more accurate state estimation and likelihood function are obtained, finally, position and Doppler measurement information are utilized to calculate weight.
The Doppler information y is used below d The method comprises the following specific steps of:
sequentially updating the state of the target by using Doppler information:
the weights of the location components are:
wherein,,the clutter intensities, denoted as position component and doppler component, respectively. Predictive Doppler measurement->The method comprises the following steps:
the Doppler measurement gain and covariance are:
wherein R is c Sum sigma d The Jacobian matrix for the Doppler measurement is:
H d (m)＝[h 1 ,h 2 ,h 3 ,h 4 ]
the parameters in the formula are respectively as follows:
and 5, calculating the probability of target measurement through posterior probability density in the updating process, distinguishing the survival target from the new target through the probability of target measurement, and keeping the measurement related to the new target for the new track at the next moment.
After the calculation, a birth track is established according to each measurement value, and the measurements comprise clutter which are not removed by the measurement screening rule, the measurement of the current survival target and the measurement generated by the new target of interest. Although most of the clutter is removed by measurement screening before calculation, there is still some clutter that cannot be removed by simple screening rules, and often can be corrected by subsequent filtering algorithms. The last few clutter that remains will fall near the viable target measurement, which will cause the filter to consider this to be the measurement point produced by the current viable target, which in turn will cause divergence in the filtering result. For the above phenomena, we need to group the measurement set, and divide the measurement set into the survival target related point and the new target related point, and the measurement dividing steps are as follows:
at time k, measurement set Z k From the survival target correlation point Z B,k Point Z related to new target S,k Composition, herein defined as measurement set Z k Is from the order ofThe probability ρ (z) of the target is:
wherein,,indicating the time Z of k k Correlation probability of target i of medium survival and measurement, otherwise, 1-p i Indicating the measurement Z at time k k Is derived from the associated probabilities of the new target i and clutter. P is p D,k For the sensor detection probability, z k For the measurement point at time k, < >>Gaussian density, H, representing mean m, variance P k For the observation matrix, R is the position measurement noise covariance, h (·) is defined as follows:
in the above equation, (x, y) represents the state vector of the target, (x) s ,y s ) Representing the sensor position.
If ρ (z) k )>0.5 then considers that the measurement is derived from the current surviving target, whereas when ρ (z k ) And < 0.5 indicates that the measurement is derived from a new target or clutter. Dividing the measurement set into Z according to the above rule B,k ，Z S,k Two parts, and Z B,k For generating new target component at the next moment, Z is S,k For use in subsequent updates of the GLMB.
The experimental conditions in this application are: the new positions (-750, 20), (-750 ), (750, 750), (-750, 750) appear at different positions at different moments in time, the survival probability of the target is p S =0.99, the detection probability of the sensor is p D =0.98, a total of 9 targets, each object moving at constant speed and a state vector ofWherein->And->Representing the location component +_>And->Representing the velocity components, the initial states of the 9 tracked objects and the times of birth and death are shown in table 1. Fig. 2 shows the motion trajectories of the targets, and fig. 3 shows the data measured by the sensors, including the target-generated measurements and clutter.
TABLE 1 tracking target run State Table
The target state transition matrix is set as:
/>
the covariance matrix is:
wherein v represents process noise with covariance of Q and mean value of 0, sigma v =10m/s. The prior speed standard deviation of the target is sigma s ＝17m/s，Standard deviation of Doppler observed noise is sigma d In the experiment, clutter density obeys poisson distribution, clutter point number obeys mean value lambda in each period c Poisson distribution of =20, the position of each clutter point is uniformly distributed in the measuring range, λ c Representing the average number of clutter per unit volume. The Doppler radar position is set asThe observed noise covariance matrix is r=diag ([ (pi/180)) 2 ,100]). For ease of comparison, pruning parameters t=10 for all filters -5 Combining threshold u=4 and maximum number of gaussian components J max =100, the multi-objective extraction threshold is 0.5.
To confirm the effectiveness of the algorithm, it can be seen by comparing fig. 4 and 5 that the GLMB cannot effectively track the target without adding the track adaptive start algorithm, and the target appearing at the unknown position is completely undetectable, even when the target tracks intersect, giving an erroneous tracking result. Fig. 6 shows the performance of the filtering algorithm, with fluctuations occurring at time 0,10,30,40,50, since new targets are now generated, but can remain converged within 2s, illustrating the effectiveness of the track initiation algorithm.
In summary, the method of the embodiment can calculate parameters required for starting a new track in one step according to doppler measurement, and can effectively reduce the influence of clutter and measurement points of survival targets on the new track in a measurement screening and grouping mode.
The above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention, and are intended to be included in the scope of the present invention.

Claims (5)

1. The adaptive track starting method based on GLMB is characterized by comprising the following steps of:
converting the polar coordinate system measurement data received by the radar into rectangular coordinate system down-conversion measurement data;
according to the prior information, correlating the single target density of the new target with the single target density of the survival target to obtain multi-target prediction probability density;
filtering clutter in the measured data by a speed screening rule and a Doppler information screening rule;
updating the multi-target posterior probability density by adopting a sequential filtering mode;
the probability of target measurement is calculated through posterior probability density in the updating process, the survival target and the new target are distinguished through the probability of target measurement, and the measurement related to the new target is reserved for the new track at the next moment.
2. The adaptive track initiation method based on GLMB according to claim 1, wherein the correlation between the single target density of the new target and the single target density of the surviving target is performed according to prior information, so as to obtain a multi-target prediction probability density, and specifically comprising the following steps:
according to the new generation component calculated at the previous moment, calculating to obtain the single target density of the new generation target;
according to posterior information transmitted at the previous moment, calculating to obtain single target density of the survival target;
and connecting the single target density of the new target and the single target density of the survival target in parallel to obtain the multi-target prediction probability density.
3. The adaptive track initiation method according to claim 2, wherein the correlation between the single target density of the new target and the single target density of the surviving target to obtain the multi-target prediction probability density further comprises: and calculating the initial running state of the new target according to the hidden speed information in the Doppler information, wherein the initial running state comprises a new target mean value, a covariance and a weight.
4. A GLMB based adaptive track start method according to claim 3, wherein the multi-objective posterior probability density:
in the method, in the process of the invention,mean, variance and weight of the loss measurements are represented;representing the mean, variance and weight of the position measurements; θ (l) represents the track association map with label l; delta (θ (l)) is a delta function, when θ (l) =0, delta (θ (l))=1 indicates that the measurement is not associated with the track, and indicates that the track has missed detection; conversely, θ (l) noteq0, δ (θ (l))=0 indicates that the track and the measurement information are updated, indicating that the track is normal.
5. The adaptive track start method based on GLMB as set forth in claim 4, wherein the probability ρ (z) of the target measurement is:
wherein,,representing a measurement set Z at time k k The probability of correlation of the medium survival target i with the measurement, and vice versa,1-p i representing a measurement set Z at time k k The associated probabilities of the nascent object and clutter; p is p D,k Probability for sensor detection; z k For the measurement point at time k, < >>Gaussian density, H, representing mean m, variance P k For the observation matrix, R is the position measurement noise covariance, h (·) is defined as follows:
in the above formula, (x, y) represents the target position, (x) s ,y s ) Representing the sensor position.
CN202310493470.5A 2023-05-04 2023-05-04 GLMB-based adaptive track initiation method Pending CN116520311A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310493470.5A CN116520311A (en) 2023-05-04 2023-05-04 GLMB-based adaptive track initiation method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310493470.5A CN116520311A (en) 2023-05-04 2023-05-04 GLMB-based adaptive track initiation method

Publications (1)

Publication Number Publication Date
CN116520311A true CN116520311A (en) 2023-08-01

Family

ID=87402613

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310493470.5A Pending CN116520311A (en) 2023-05-04 2023-05-04 GLMB-based adaptive track initiation method

Country Status (1)

Country Link
CN (1) CN116520311A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117724087A (en) * 2024-02-07 2024-03-19 中国人民解放军海军航空大学 Radar multi-target tracking double-tag Bernoulli filtering algorithm

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117724087A (en) * 2024-02-07 2024-03-19 中国人民解放军海军航空大学 Radar multi-target tracking double-tag Bernoulli filtering algorithm
CN117724087B (en) * 2024-02-07 2024-05-28 中国人民解放军海军航空大学 Radar multi-target tracking double-tag Bernoulli filtering algorithm

Similar Documents

Publication Publication Date Title
CN110503071B (en) Multi-target tracking method based on variational Bayesian label multi-Bernoulli superposition model
CN106407677B (en) A kind of multi-object tracking method in the case of missing measurement
CN114609912B (en) Angle measurement target tracking method based on pseudo-linear maximum correlation entropy Kalman filtering
CN111722214B (en) Method for realizing radar multi-target tracking PHD
CN107688179B (en) Comprehensive probability data interconnection method based on Doppler information assistance
Shi et al. Adaptive UKF for target tracking with unknown process noise statistics
CN110865343B (en) LMB-based particle filter tracking-before-detection method and system
CN111257865B (en) Maneuvering target multi-frame detection tracking method based on linear pseudo-measurement model
CN116520311A (en) GLMB-based adaptive track initiation method
CN111259332B (en) Fuzzy data association method and multi-target tracking method in clutter environment
CN112328959A (en) Multi-target tracking method based on adaptive extended Kalman probability hypothesis density filter
CN111142101B (en) Data association method
CN107561506B (en) Multistage false alarm probability joint recursion target detection method
CN115544425A (en) Robust multi-target tracking method based on target signal-to-noise ratio characteristic estimation
Feng et al. Cardinality Balanced Multi-Target Multi-Bernoulli filter for target tracking with amplitude information
Wang et al. Research on multi-maneuvering target tracking JPDA algorithm
CN112666516A (en) Passive tracking method based on track information field
CN102707278B (en) Multi-target tracking method for singular value decomposition
Sun et al. Real-time detection and tracking of multiple ground targets
CN111504326B (en) Robust GLMB multi-target tracking method based on T distribution
CN114577213B (en) Distributed multi-platform underwater multi-target association and passive positioning method
Sönmez et al. Analysis of performance criteria for optimization based bearing only target tracking algorithms
CN116224320B (en) Radar target tracking method for processing Doppler measurement under polar coordinate system
Wang et al. Adaptive tracking of group targets using random matrix
CN113534131B (en) GM-PHD-based motion parameter estimation combined smooth filtering method

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