CN102706345A - Maneuvering target tracking method based on fading memory sequential detector - Google Patents

Maneuvering target tracking method based on fading memory sequential detector Download PDF

Info

Publication number
CN102706345A
CN102706345A CN2012101900658A CN201210190065A CN102706345A CN 102706345 A CN102706345 A CN 102706345A CN 2012101900658 A CN2012101900658 A CN 2012101900658A CN 201210190065 A CN201210190065 A CN 201210190065A CN 102706345 A CN102706345 A CN 102706345A
Authority
CN
China
Prior art keywords
motor
driven
state
target
memory sequential
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
CN2012101900658A
Other languages
Chinese (zh)
Other versions
CN102706345B (en
Inventor
彭冬亮
潘宝贵
邵根富
陈华杰
郭云飞
申屠晗
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.)
Hangzhou Dianzi University
Original Assignee
Hangzhou Dianzi 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 Hangzhou Dianzi University filed Critical Hangzhou Dianzi University
Priority to CN201210190065.8A priority Critical patent/CN102706345B/en
Publication of CN102706345A publication Critical patent/CN102706345A/en
Application granted granted Critical
Publication of CN102706345B publication Critical patent/CN102706345B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Feedback Control In General (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

The invention discloses a maneuvering target tracking method based on a fading memory sequential detector. Aiming at advantages and disadvantages of a decision-based single-model tracking method and a multi-model tracking method, the invention provides the maneuvering target tracking method based on the fading memory sequential detector. The method comprises the following steps of: performing Kalman filter-unscented Kalman filter (KF-UKF) combined filtration, wherein Kalman filtration based on a uniform model is adopted for a non-maneuvering target; when a maneuvering target is detected, expanding the dimension of a coefficient of variation (CV) model, and performing state estimation by adopting an unscented Kalman filter (UKF); detecting maneuvering by using the fading memory sequential detector; and after the maneuvering is detected, correcting a dimension expanded state vector and a covariance matrix, so that the matched state is matched with the actual moving mode. By the method, the state of the target can be corrected in real time, and matched filters can be selected according to different moving modes, so that the tracking precision is improved.

Description

A kind of maneuvering target tracking method based on decay memory sequential detector
Technical field
The invention belongs to maneuver detection and maneuvering target tracking field, relate to a kind of maneuvering target tracking method based on decay memory sequential detector.
Background technology
All have a wide range of applications in the military-civil fields such as maneuvering target tracking traffic control aloft (ATC) in recent years,, ground investigation supervision, homing.The difficulty of maneuvering target tracking mainly comes from the uncertainty of target travel and state observation.The uncertain uncertainty that is meant target identities of observation belongs to the category of multiple target tracking; The uncertainty of the uncertain main feeling the pulse with the finger-tip mark motion model of target travel; Mainly comprise: unknown motor-driven input and process noise statistical nature; These factors all can cause trace model and target travel mode mismatch, cause tracking performance to descend, and can cause filtering divergence when serious.
Current, various maneuvering target tracking researchs all are that the motor-driven uncertainty of surrounding target is launched, and become two types of main maneuvering target tracking methods based on the single model method and the multi-model process of making a strategic decision.
Because the maneuvering target motor pattern is represented by model usually, thereby in fact the target travel mode decision in the maneuvering target tracking is exactly the target movement model decision-making.It is one of technology relevant with the model decision-making in the current mobility target following that target maneuver detects.In the tracing process, after the target travel pattern changes, can cause the mismatch of Filtering Model and motor pattern usually, cause filtering error to increase, must make adjustment timely to used Filtering Model by corresponding mode decision process.Maneuver detection is exactly through detecting motor-driven generation, adjusting a kind of decision-making technic that trace model matees corresponding mode of motion then.
Although multi-model process enjoys favor with its good performance for tracking and robustness, its design process is complicated and need great computational resource when realizing.In many resource-constrained occasions, application is widely arranged based on the single model method of making a strategic decision.Research recently shows that the maneuver detection method of no-float can obtain the tracking performance close with multi-model process based on the single model method of making a strategic decision.Thereby timely, correct maneuver detection becomes the key point of these class methods.
To the single model tracking based on decision-making, the experts and scholars and the engineering technical personnel of target tracking domain have carried out a large amount of research both at home and abroad.But their used detecting device mostly is based on the fixing sliding window formula detecting device of length of window, the selection that common issue with of the existence of this type detecting device is a maneuver detection device length of window.Generally, detection probability increases with length of window, but the increase that the length of window increase can cause maneuver detection to postpone, this influence to performance of target tracking is huge; Though and length of window reduces to reduce maneuver detection and postpones, false-alarm probability can increase thereupon, thereby how to select length of window be the difficult point of this type detecting device.
Summary of the invention
The objective of the invention is to deficiency, designed a kind of decay memory sequential detector to existing detecting device.And detecting device merged in the target tracking algorism, for maneuver detection, target tracking domain research provide a kind of maneuvering target tracking method based on decay memory sequential detector.
A kind of maneuvering target tracking method based on decay memory sequential detector may further comprise the steps:
Step 1, state covariance are put initial value and Kalman filtering: when target during at linear uniform motion, moving target adopts two-dimentional Kalman filter, and system model adopts the CV model, carries out Kalman filtering according to state equation and measurement equation.
Step 2, detector design: according to the likelihood ratio function of definition, design a kind of decay memory sequential detector based on Generalized Likelihood Ratio, described decay memory sequential detector is realized by decay memory Sequential Detection Algorithm.
Step 3, input are estimated and state compensation: utilize decay memory sequential detector to detect the time of motor-driven generation, afterwards motor-driven amplitude is estimated.Motor-driven amplitude according to estimating is revised new breath and covariance.
Step 4, expand dimension and revise in real time with state: when kConstantly detect when motor-driven, to [ N, k] state in the time interval compensates, then state vector and covariance matrix are expanded dimension, and according to the size of the motor-driven amplitude that estimates the CV model is revised in real time.
Step 5, withdraw from motor-driven differentiation: under given level of significance,, think that then the generation of acceleration is not the motor-driven generation by target, can think that target is in non-maneuvering condition if discriminant is set up.
Step 6, tasteless Kalman Filter Estimation: motor-driven when the target generation, adopt tasteless Kalman filter that dbjective state is carried out filtering, tasteless Kalman filtering process does not at first have the mark conversion, carries out filtering with the state variable after the conversion then.
Beneficial effect of the present invention: it is the condition that order obtains that the present invention utilizes observation data; Decay sequential detector based on Generalized Likelihood Ratio has been proposed; The benefit of decay memory is to have utilized the data in the different moment these theories of influence difference to motor-driven differentiation, and the feasible moment data of closing on are bigger to the influence of motor-driven differentiation, and the moment far away influences less to motor-driven differentiation; Can improve the detecting device precision, shorten maneuver detection and postpone.Maneuver detection device with design is integrated in the tracker simultaneously; Designed maneuvering target tracking method based on decay memory sequential detector; While method adopts the method for Kalman filter and tasteless Kalman filter associating filtering to carry out filtering, and target adopts Kalman filter when being in non-maneuvering condition, when target is in maneuvering condition; Adopt tasteless Kalman filter to carry out filtering, improved filtering accuracy.This method is for handling maneuvering target track question; Particularly for motor-driven and non-motor-driven alternate motor pattern, this method has good tracking performance, compares simultaneously based on the maneuvering target tracking method of multi-model; This method has reduced calculated amount, has improved the practicality of method.The present invention has stronger maneuver detection performance and practicality.
Description of drawings
Fig. 1 is the inventive method process flow diagram.
Embodiment
Below in conjunction with accompanying drawing the present invention is described further.
As shown in Figure 1, the present invention includes following steps:
Step 1, state covariance are put initial value and Kalman filtering: when target during at linear uniform motion, moving target adopts two-dimentional Kalman filter, and system model adopts the CV model, carries out Kalman filtering according to state equation and measurement equation.
Step 2, detector design:, design a kind of decay memory sequential detector based on Generalized Likelihood Ratio according to the likelihood ratio function of definition.Described decay memory sequential detector is realized by decay memory Sequential Detection Algorithm.In target maneuver detects; Observation data obtains often in proper order; Thereby than the sliding window formula detecting device of batch processing; Sequential detection be fit to use more with maneuver detection in, and sequential detection need not to preestablish detection window length, thereby removed the important difficult problem of sliding window formula in detecting from.
Step 3, input are estimated and state compensation: utilize decay memory sequential detector to detect the time of motor-driven generation, afterwards motor-driven amplitude is estimated.Motor-driven amplitude according to estimating is revised new breath and covariance.
Step 4, expand dimension and revise in real time with state: when detecting constantly at k when motor-driven, to [ N, k] state in the time interval compensates, then state vector and covariance matrix are expanded dimension, and according to the size of the motor-driven amplitude that estimates the CV model is revised in real time.
Step 5, withdraw from motor-driven differentiation: under given level of significance,, think that then the generation of acceleration is not the motor-driven generation by target, can think that target is in non-maneuvering condition if discriminant is set up.
Step 6, tasteless Kalman Filter Estimation: motor-driven when the target generation, adopt tasteless Kalman filter that dbjective state is carried out filtering, tasteless Kalman filtering process does not at first have the mark conversion, carries out filtering with the state variable after the conversion then.
To put the concrete steps of initial value and Kalman filtering following for the state covariance in the step 1:
When target during at linear uniform motion, moving target adopts two-dimentional Kalman filter, and system model adopts the CV model.Carry out Kalman filtering according to state equation and measurement equation.
Initial covariance is:
(1)
In the formula;
Figure 2012101900658100002DEST_PATH_IMAGE004
is measuring noise square difference, and
Figure 2012101900658100002DEST_PATH_IMAGE006
is the sampling period.Because is the covariance of location components, size equals the covariance of measurement noise.
Figure 2012101900658100002DEST_PATH_IMAGE010
,
Figure 2012101900658100002DEST_PATH_IMAGE012
are the covariance of speed component; Relationship by speed and position can obtain,
Figure 390057DEST_PATH_IMAGE010
= = .
Figure 2012101900658100002DEST_PATH_IMAGE016
is the covariance of component of acceleration; Relationship by position and acceleration; Get
Figure 211568DEST_PATH_IMAGE016
=
Figure 2012101900658100002DEST_PATH_IMAGE018
; In like manner can get,
Figure 2012101900658100002DEST_PATH_IMAGE022
Figure 2012101900658100002DEST_PATH_IMAGE024
Figure 2012101900658100002DEST_PATH_IMAGE026
.Just can obtain initial covariance matrix this moment.
The concrete steps of detector design are following in the step 2:
State equation was when hypothetical target was motor-driven:
(2)
Wherein, C is the input gating matrix;
Figure 2012101900658100002DEST_PATH_IMAGE030
is k motor-driven amplitude size constantly, and
Figure 2012101900658100002DEST_PATH_IMAGE032
is process noise.
Figure 858975DEST_PATH_IMAGE030
and
Figure 611030DEST_PATH_IMAGE032
all have Gaussian.
At first do dualism hypothesis:
Figure 2012101900658100002DEST_PATH_IMAGE034
：
Figure 2012101900658100002DEST_PATH_IMAGE036
Figure 2012101900658100002DEST_PATH_IMAGE038
Figure 2012101900658100002DEST_PATH_IMAGE040
：
Figure 2012101900658100002DEST_PATH_IMAGE042
(3)
Wherein hypothesis expression is no motor-driven; The expression appearance is motor-driven to suppose
Figure 949793DEST_PATH_IMAGE040
,
Figure 2012101900658100002DEST_PATH_IMAGE044
the motor-driven amplitude of expression.
Definition likelihood ratio function
Figure 2012101900658100002DEST_PATH_IMAGE046
Figure 2012101900658100002DEST_PATH_IMAGE048
(4)
Wherein,
Figure 2012101900658100002DEST_PATH_IMAGE050
is new breath; Have the white Gaussian noise characteristic,
Figure 2012101900658100002DEST_PATH_IMAGE052
be the joint probability density function of new breath under the motor-driven hypothesis.
The decay memory Sequential Detection Algorithm of recursive form is following:
Figure 2012101900658100002DEST_PATH_IMAGE054
?，
Figure 2012101900658100002DEST_PATH_IMAGE056
(5)
Wherein,
Figure 2012101900658100002DEST_PATH_IMAGE058
is decay factor, and the value of can rule of thumb draw.
Relation according to formula (5) and 0 can be made as 0 with detection threshold, the probability density function Gaussian distributed, and motor-driven decision rule is following:
(1) if ; Then accepting
Figure 2012101900658100002DEST_PATH_IMAGE062
supposes; Promptly exist motor-driven; So motor-driven zero-time is:
Figure 2012101900658100002DEST_PATH_IMAGE064
, carry out state compensation afterwards;
(2)? If
Figure 2012101900658100002DEST_PATH_IMAGE066
? continue the inspection process
Figure 2012101900658100002DEST_PATH_IMAGE068
.
Input is estimated with the concrete steps of state compensation following in the step 3:
Utilize decay memory sequential detector to detect the time of motor-driven generation, afterwards motor-driven amplitude is estimated that it is motor-driven to suppose that n takes place constantly, k detects motor-driven generation constantly, then being estimated as of motor-driven amplitude:
Figure 2012101900658100002DEST_PATH_IMAGE070
(6)
Wherein,
Figure 2012101900658100002DEST_PATH_IMAGE072
is that k state-transition matrix,
Figure 2012101900658100002DEST_PATH_IMAGE074
constantly is motor-driven error covariance matrix, and
Figure 2012101900658100002DEST_PATH_IMAGE076
is new the breath.
Adopt the state estimation and the covariance in a sampling period of Kalman filter filtering to be estimated as:
Figure 2012101900658100002DEST_PATH_IMAGE078
Figure 2012101900658100002DEST_PATH_IMAGE080
Figure 2012101900658100002DEST_PATH_IMAGE084
Figure 2012101900658100002DEST_PATH_IMAGE086
(7)
Wherein,
Figure 2012101900658100002DEST_PATH_IMAGE088
;
Figure 2012101900658100002DEST_PATH_IMAGE090
is filter gain;
Figure 2012101900658100002DEST_PATH_IMAGE092
is state one-step prediction covariance matrix,
Figure 2012101900658100002DEST_PATH_IMAGE094
be new breath covariance matrix.
So [ N, k] state correction in the time interval is following:
(8)
In the formula (8), [ N, k] correction of time interval internal state, comprised the corresponding analytic relationship of the time interval and motor-driven amplitude size, because dbjective state be a 6 DOF, comprise, position, speed, acceleration aliquot, and the size of motor-driven amplitude is exactly the size of acceleration.Therefore the motor-driven amplitude and the time interval are satisfied the analytic relationship of formula (8).
New breath is revised:
Figure 2012101900658100002DEST_PATH_IMAGE098
Figure 2012101900658100002DEST_PATH_IMAGE100
(9)
The covariance correction:
Figure 2012101900658100002DEST_PATH_IMAGE102
(10)
Wherein
Figure 2012101900658100002DEST_PATH_IMAGE104
is the error covariance matrix of motor-driven amplitude.
Figure 2012101900658100002DEST_PATH_IMAGE106
,
Figure 2012101900658100002DEST_PATH_IMAGE108
,
Figure 2012101900658100002DEST_PATH_IMAGE110
Be with [ N ... K] between the relevant matrix of coefficients of state.
It is following with the concrete steps of state correction in real time to expand dimension in the step 4:
Motor-driven when detecting constantly at k, and to [ N, k] after state in the time interval compensates, at last state vector and covariance matrix are expanded dimension, and according to the size of the motor-driven amplitude that estimates model is revised in real time.
Figure 2012101900658100002DEST_PATH_IMAGE112
(11)
(12)
Wherein,
Figure 2012101900658100002DEST_PATH_IMAGE116
is the estimation of motor-driven amplitude.
The concrete steps that withdraw from motor-driven differentiation in the step 5 are following:
The normalization of input amplitude square does
Figure 2012101900658100002DEST_PATH_IMAGE118
(13)
[ N, k] in the time interval, the normalization quadratic sum of amplitude does
(14)
Discriminant:
Figure 2012101900658100002DEST_PATH_IMAGE122
(15)
Under given level of significance
Figure 2012101900658100002DEST_PATH_IMAGE124
; If satisfy following formula; Think that then acceleration is not by the motor-driven generation of target; Do non-motor-driven processing under this kind situation, can think that target is in non-maneuvering condition.
The concrete steps that tasteless Kalman filter is estimated in the step 6 are following:
The UKF wave filter promptly is tasteless Kalman filter among the present invention.
UKF at first carries out tasteless conversion (UT) when the treatment state equation, use the state variable after the tasteless conversion to carry out Filtering Estimation then, to reduce evaluated error.
Calculate 2n+1
Figure 2012101900658100002DEST_PATH_IMAGE126
point
Figure 2012101900658100002DEST_PATH_IMAGE128
; I=1; 2; 2n, wherein n is the state vector dimension.
Figure 2012101900658100002DEST_PATH_IMAGE130
(16)
In the formula
Figure 2012101900658100002DEST_PATH_IMAGE132
;
Figure 2012101900658100002DEST_PATH_IMAGE134
determined
Figure 772834DEST_PATH_IMAGE126
spread of points degree; Usually get one less on the occasion of; As 0.01,
Figure 2012101900658100002DEST_PATH_IMAGE136
gets 0 usually.
Calculate
Figure 2012101900658100002DEST_PATH_IMAGE138
point
Figure 890831DEST_PATH_IMAGE128
; I=1; 2; 2n, through the propagation of state equation:
Figure 2012101900658100002DEST_PATH_IMAGE140
(17)
In the formula
Figure 2012101900658100002DEST_PATH_IMAGE142
.
Calculate
Figure 769794DEST_PATH_IMAGE126
point
Figure 2012101900658100002DEST_PATH_IMAGE144
,
Figure 2012101900658100002DEST_PATH_IMAGE146
through the propagation of measurement equation:
(18)
Calculate the step prediction in advance of output:
Figure 2012101900658100002DEST_PATH_IMAGE150
(19)
After new measurement amount
Figure 2012101900658100002DEST_PATH_IMAGE152
arrives, carrying out filtering upgrades:
Figure 2012101900658100002DEST_PATH_IMAGE154
(20)
Wherein
Figure 2012101900658100002DEST_PATH_IMAGE156
is the filter gain matrix.
The main flow process of the inventive method is: hypothetical target is initial to move with uniform velocity; At first state vector and covariance matrix are put initial value according to measuring value; Employing is estimated state based on the two-dimentional Kalman filter of CV model, utilizes decay memory sequential detector to carry out maneuver detection simultaneously.If do not have motor-drivenly, return the state estimation of proceeding following one-period; If detect motor-drivenly, then the time of motor-driven amplitude and motor-driven generation is estimated.Consider the time-delay of maneuver detection, compensate and correct filter state and covariance this moment.At last the CV model is expanded dimension, obtain a model of revising in real time, carry out UKF filtering according to real-time revised model afterwards according to revised state and motor-driven amplitude.Repeat motor-driven testing process,, then import and estimate and state compensation, simultaneously model is carried out real-time update if detect motor-drivenly again; If do not detect motor-drivenly, then withdraw from motor-driven judgement, if motor-driven amplitude greater than thresholding, then expression is not withdrawed from motor-driven; Repeat UKF filtering,, think that then target is in non-maneuvering condition, at this moment if less than thresholding; Adopt the Kalman filtering of two dimension, repeat above process, constituted maneuvering target tracking method based on decay memory sequential detector.

Claims (2)

1. maneuvering target tracking method based on decay memory sequential detector is characterized in that this method may further comprise the steps:
Step 1, state covariance are put initial value and Kalman filtering: when target during at linear uniform motion, moving target adopts two-dimentional Kalman filter, and system model adopts the CV model, carries out Kalman filtering according to state equation and measurement equation;
Step 2, detector design: according to the likelihood ratio function of definition, design is based on the decay memory sequential detector of Generalized Likelihood Ratio, and described decay memory sequential detector is realized by decay memory Sequential Detection Algorithm;
Step 3, input are estimated and state compensation: utilize decay memory sequential detector to detect the time of motor-driven generation, afterwards motor-driven amplitude is estimated; Motor-driven amplitude according to estimating is revised new breath and covariance;
Step 4, expand dimension and revise in real time with state: when kConstantly detect when motor-driven, to [ N, k] state in the time interval compensates, then state vector and covariance matrix are expanded dimension, and according to the size of the motor-driven amplitude that estimates the CV model is revised in real time;
Step 5, withdraw from motor-driven differentiation: under given level of significance,, think that then the generation of acceleration is not the motor-driven generation by target, can think that target is in non-maneuvering condition if discriminant is set up;
Step 6, tasteless Kalman Filter Estimation: motor-driven when the target generation, adopt tasteless Kalman filter that dbjective state is carried out filtering, tasteless Kalman filtering process does not at first have the mark conversion, carries out filtering with the state variable after the conversion then.
2. according to the described a kind of maneuvering target tracking method of right 1, it is characterized in that based on decay memory sequential detector: step 2 specifically:
At first do dualism hypothesis:
：
Figure 363521DEST_PATH_IMAGE003
Figure 451563DEST_PATH_IMAGE004
：
Figure 843230DEST_PATH_IMAGE005
(1)
Wherein hypothesis
Figure 42130DEST_PATH_IMAGE001
expression is no motor-driven; Expression occurs motor-driven to suppose
Figure 383113DEST_PATH_IMAGE004
; The motor-driven amplitude of
Figure 642056DEST_PATH_IMAGE006
expression,
Figure 586265DEST_PATH_IMAGE007
representes constantly;
Definition likelihood ratio function
Figure 323277DEST_PATH_IMAGE008
Figure 784346DEST_PATH_IMAGE009
(2)
Wherein, is new breath; Have the white Gaussian noise characteristic,
Figure 377187DEST_PATH_IMAGE011
be the joint probability density function of new breath under the motor-driven hypothesis;
The decay memory Sequential Detection Algorithm of recursive form is following:
Figure 917890DEST_PATH_IMAGE012
?，
Figure 561361DEST_PATH_IMAGE013
(3)
Wherein,
Figure 834210DEST_PATH_IMAGE014
is decay factor
Relation according to formula (3) and 0 is made as 0 with detection threshold, the probability density function Gaussian distributed, and motor-driven decision rule is following:
(1) if
Figure 563132DEST_PATH_IMAGE015
; Then accepting
Figure 828897DEST_PATH_IMAGE016
supposes; Promptly exist motor-driven; So motor-driven zero-time is: , carry out state compensation afterwards;
(2)? If
Figure 770625DEST_PATH_IMAGE018
? continue the inspection process .
CN201210190065.8A 2012-06-11 2012-06-11 Maneuvering target tracking method based on fading memory sequential detector Active CN102706345B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201210190065.8A CN102706345B (en) 2012-06-11 2012-06-11 Maneuvering target tracking method based on fading memory sequential detector

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201210190065.8A CN102706345B (en) 2012-06-11 2012-06-11 Maneuvering target tracking method based on fading memory sequential detector

Publications (2)

Publication Number Publication Date
CN102706345A true CN102706345A (en) 2012-10-03
CN102706345B CN102706345B (en) 2015-01-28

Family

ID=46899297

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201210190065.8A Active CN102706345B (en) 2012-06-11 2012-06-11 Maneuvering target tracking method based on fading memory sequential detector

Country Status (1)

Country Link
CN (1) CN102706345B (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103744058A (en) * 2013-12-24 2014-04-23 河海大学 Ballistic trajectory formation method based on exponential weighting attenuated memory filtering
CN103901426A (en) * 2012-12-28 2014-07-02 重庆凯泽科技有限公司 Single-maneuvering target tracking system
CN104180801A (en) * 2014-08-20 2014-12-03 广州海格通信集团股份有限公司 Method and system for predicting track points based on ADS-B system
CN106772351A (en) * 2016-11-16 2017-05-31 西安电子科技大学 Kalman filter method based on the memory of limited step
CN106910211A (en) * 2015-12-21 2017-06-30 中国石油天然气股份有限公司 Multi-maneuvering-target tracking method under complex environment
CN108332939A (en) * 2017-12-31 2018-07-27 中国科学院声学研究所 A method of across medium time-varying fluid parameter On-line Estimation
CN108629935A (en) * 2018-05-17 2018-10-09 山东深图智能科技有限公司 A kind of method and system for climbing building pivot frame larceny based on video monitoring detection
CN113155122A (en) * 2021-04-01 2021-07-23 广州大学 Maneuvering target tracking method based on adaptive filtering

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2128644A1 (en) * 2007-03-22 2009-12-02 Furuno Electric Co., Ltd. Gps composite navigation apparatus
US20090322590A1 (en) * 2007-04-18 2009-12-31 Schoettl Alfred Method with a system for ascertaining and predicting a motion of a target object
CN102043906A (en) * 2010-12-15 2011-05-04 北京航空航天大学 Diagonal-form fusion method for maneuvering target tracking

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2128644A1 (en) * 2007-03-22 2009-12-02 Furuno Electric Co., Ltd. Gps composite navigation apparatus
US20090322590A1 (en) * 2007-04-18 2009-12-31 Schoettl Alfred Method with a system for ascertaining and predicting a motion of a target object
CN102043906A (en) * 2010-12-15 2011-05-04 北京航空航天大学 Diagonal-form fusion method for maneuvering target tracking

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
JIFENG RU: "Sequential Detection of Target Maneuvers", 《2005 7TH INTERNATIONAL CONFERENCE ON INFORMATION FUSION(FUSION)》 *
盛琥等: "基于新息序列修正的输入估计算法", 《***工程与电子技术》 *

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103901426A (en) * 2012-12-28 2014-07-02 重庆凯泽科技有限公司 Single-maneuvering target tracking system
CN103744058A (en) * 2013-12-24 2014-04-23 河海大学 Ballistic trajectory formation method based on exponential weighting attenuated memory filtering
CN104180801A (en) * 2014-08-20 2014-12-03 广州海格通信集团股份有限公司 Method and system for predicting track points based on ADS-B system
CN104180801B (en) * 2014-08-20 2017-06-06 广州海格通信集团股份有限公司 Forecasting Methodology and system based on ADS B system track points
CN106910211A (en) * 2015-12-21 2017-06-30 中国石油天然气股份有限公司 Multi-maneuvering-target tracking method under complex environment
CN106772351A (en) * 2016-11-16 2017-05-31 西安电子科技大学 Kalman filter method based on the memory of limited step
CN108332939A (en) * 2017-12-31 2018-07-27 中国科学院声学研究所 A method of across medium time-varying fluid parameter On-line Estimation
CN108332939B (en) * 2017-12-31 2020-04-07 中国科学院声学研究所 Cross-medium time-varying fluid parameter online estimation method
CN108629935A (en) * 2018-05-17 2018-10-09 山东深图智能科技有限公司 A kind of method and system for climbing building pivot frame larceny based on video monitoring detection
CN113155122A (en) * 2021-04-01 2021-07-23 广州大学 Maneuvering target tracking method based on adaptive filtering

Also Published As

Publication number Publication date
CN102706345B (en) 2015-01-28

Similar Documents

Publication Publication Date Title
CN102706345A (en) Maneuvering target tracking method based on fading memory sequential detector
CN107045125B (en) Interactive multi-model radar target tracking method based on predicted value measurement conversion
CN107832575B (en) Pseudo-measurement-based maneuvering target asynchronous track fusion algorithm with feedback
CN103383261B (en) A kind of modified can't harm the indoor moving targets location method of Kalman filtering
CN102568004A (en) Tracking algorithm for high maneuvering targets
Deng et al. Adaptive unscented Kalman filter for parameter and state estimation of nonlinear high-speed objects
CN107677997A (en) Extension method for tracking target based on GLMB filtering and Gibbs samplings
CN105549049A (en) Adaptive Kalman filtering algorithm applied to GPS navigation
CN106443622A (en) Distributed target tracking method based on improved joint probability data association
CN106407677A (en) Multi-target tracking method in case of loss of measurement data
Wang et al. Fuzzy-control-based particle filter for maneuvering target tracking
CN104504728B (en) Multiple maneuver target tracking methods, system and broad sense JPDA device thereof
CN101221238A (en) Dynamic deviation estimation method based on gauss average value mobile registration
CN110231620B (en) Noise-related system tracking filtering method
CN111711432B (en) Target tracking algorithm based on UKF and PF hybrid filtering
CN109712173A (en) A kind of picture position method for estimating based on Kalman filter
CN101299271A (en) Polynomial forecast model of maneuvering target state equation and tracking method
CN110174907A (en) A kind of human body target follower method based on adaptive Kalman filter
Liu et al. Application of adaptive UKF algorithm in multi-target tracking and positioning system
CN105549003A (en) Automobile radar target tracking method
Chen et al. Research on adaptive Monte Carlo location method based on fusion posture estimation
CN115859626A (en) Self-adaptive unscented Kalman filter design method for periodic moving target
CN104035328A (en) Multi-moving-body tracking control method by adopting interference estimator
CN104931932A (en) Improved debiased coordinate transform Kalman filtering method
Zhao et al. An adaptive UKF with noise statistic estimator

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20121003

Assignee: HANGZHOU DAZHI ELECTROMECHANICAL Co.,Ltd.

Assignor: HANGZHOU DIANZI University

Contract record no.: X2020330000073

Denomination of invention: A maneuvering target tracking method based on attenuated memory sequential detector

Granted publication date: 20150128

License type: Common License

Record date: 20201010

EE01 Entry into force of recordation of patent licensing contract