CN104090262B - A kind of method for tracking moving target merging estimation based on multi-sampling rate multi-model - Google Patents

A kind of method for tracking moving target merging estimation based on multi-sampling rate multi-model Download PDF

Info

Publication number
CN104090262B
CN104090262B CN201410223531.7A CN201410223531A CN104090262B CN 104090262 B CN104090262 B CN 104090262B CN 201410223531 A CN201410223531 A CN 201410223531A CN 104090262 B CN104090262 B CN 104090262B
Authority
CN
China
Prior art keywords
sampling rate
estimation
cluster node
state
model
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201410223531.7A
Other languages
Chinese (zh)
Other versions
CN104090262A (en
Inventor
张文安
杨旭升
俞立
刘安东
陈博
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.)
Zhejiang University of Technology ZJUT
Original Assignee
Zhejiang University of Technology ZJUT
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 Zhejiang University of Technology ZJUT filed Critical Zhejiang University of Technology ZJUT
Priority to CN201410223531.7A priority Critical patent/CN104090262B/en
Publication of CN104090262A publication Critical patent/CN104090262A/en
Application granted granted Critical
Publication of CN104090262B publication Critical patent/CN104090262B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
    • G01S5/02—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using radio waves
    • G01S5/0294—Trajectory determination or predictive filtering, e.g. target tracking or Kalman filtering
    • 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
    • G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
    • G01S5/16—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using electromagnetic waves other than radio waves
    • 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
    • G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
    • G01S5/18—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using ultrasonic, sonic, or infrasonic waves
    • 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
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Electromagnetism (AREA)
  • Radar Systems Or Details Thereof (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

A kind of method for tracking moving target merging estimation based on multi-sampling rate multi-model, first wireless sensor network is divided into m bunch by the method, sets up the state-space model under different sampling rate.Leader cluster node utilizes EKF method to obtain partial estimation result.Fusion center is synchronized to synchronization point all partial estimation, and utilizes CI fusion method to obtain merging estimated result, is adjusted the sampling rate of network node by the energy information of the estimated value of target velocity and leader cluster node.The present invention provides a kind of under the premise ensureing tracking accuracy, robustness and quick-reaction capability, effectively reduces the energy consumption of sensor network, promotes the method for tracking moving target estimated based on the fusion of multi-sampling rate multi-model of motility.

Description

A kind of method for tracking moving target merging estimation based on multi-sampling rate multi-model
Technical field
The present invention relates to movable object tracking field, especially a kind of mobile object real-time tracking method.
Background technology
Wireless sensor network is due to its self-organization, robustness and the feature that can cover on a large scale so that it has important using value in the field such as environment measuring, vehicle tracking, military investigation and military target tracking.In target tracking domain, when evaluating tracking, tracking accuracy need to be considered, follow the tracks of robustness and energy expenditure and follow the tracks of the indexs such as response time.In order to improve tracking accuracy, it is proposed that multi-sensor information fusion method of estimation, namely improve tracking accuracy by merging the metrical information of multiple sensor.Especially, adopt distributions to merge method of estimation, can pass through to merge each local state and estimate to obtain the fusion estimated result that precision is higher, make the failure tolerant ability of Target Tracking System and robustness be strengthened and improve.In existing target following technology, have and adopt this distributions based on many group sensors to merge methods of estimation, but be all the mode adopting many group single sample rates of sensor and single mobile object module, tracking strategy can not be adjusted according to the energy state of the state (such as translational speed) of mobile target and sensor node, cause that system flexibility is not enough and be unfavorable for the energy-saving and cost-reducing of wireless sensor network.Adjustment information acquisition rate can be carried out to reduce the energy consumption of sensor network according to mobile target travel situation but without technology at present in wireless sensor network.
Summary of the invention
For the deficiency that the motility overcoming existing method for tracking moving target is poor, energy consumption is bigger, the present invention provides a kind of under the premise ensureing tracking accuracy, robustness and quick-reaction capability, effectively reduces the energy consumption of sensor network, promotes the method for tracking moving target estimated based on the fusion of multi-sampling rate multi-model of motility.
The technical solution adopted for the present invention to solve the technical problems is:
A kind of method for tracking moving target merging estimation based on multi-sampling rate multi-model, said method comprising the steps of:
Step 1) speed of mobile target is divided into L different grade, wireless sensor network be divided into m different bunch, sensor node is divided into n different sampling rate, select the state variable of mobile target, described state variable is position, speed or acceleration, sets up the state-space model of movable object tracking under n sampling rate；
Step 2) leader cluster node collects the measurement information of its bunch of interior nodes, according to state-space model under present sample speed, application extension Kalman's method obtains the partial estimation of mobile target, and partial estimation result and its dump energy information are sent to fusion center；
Step 3) for low sampling rate bunch, in the moment point that its partial estimation lacks, fusion center, by the state estimation in a upper moment is predicted, was synchronized to synchronization point all local estimated result；
Step 4) fusion center error co-variance matrix according to each partial estimation, determine the fusion parameters in CI fusion method online, application CI merges method of estimation, obtains the fusion estimated result of mobile target, and described fusion estimated result includes movement rate estimated value；
Step 5) fusion center energy information according to the movement rate estimated value and each leader cluster node that move target, if the movement velocity estimated value of mobile target is lower than corresponding threshold value, will reduce the sampling rate of minimum energy cluster node；Otherwise, the movement velocity estimated value of mobile target, higher than corresponding threshold value, will accelerate the sampling rate of the highest cluster node of energy；And sampling rate adjustment result is sent to each leader cluster node；
Step 6) if leader cluster node receives sampling rate adjustment information, the sampling rate of bunch interior nodes will be adjusted, and be switched to the state-space model under corresponding sampling rate, otherwise, each cluster node is undertaken sampling and state estimation by former sampling rate.
Further, described step 1) in, require to determine the value of L, m and n according to the power conservation requirement of system and tracking accuracy, wherein, L=m × n+1, correspondence L threshold speed V of L speed classc, c=1 ..., L-1 and Vc-1< Vc。
Described step 1) in, described state variable is two dimension or three-dimensional parameter.
Described step 1) in, the state-space model under each sampling rate is stored in leader cluster node, points to each state-space model by array of pointers M [n], i.e. M [j], j=1 ..., n points to the state-space model under sampling rate j.
Further, described step 2) in, described measurement information is sensor node and the distance of mobile target, the acceleration moving target and athletic posture.
In step 3) in, if this partial estimation result of subsequent time continues disappearance, the predictive value that current time is obtained is predicted again, to obtain the state estimation of subsequent time.
In step 4) in, described fusion parameters ωiDetermined online by the mark of the error co-variance matrix of each bunch of partial estimation, namelyWherein, tr (Pi,k|k)、tr(Pl,k|k) it is the mark of the error co-variance matrix of the partial estimation of bunch i, l, Pi,k|k、Pl,k|kFor the error co-variance matrix of the partial estimation of current time k bunch of i, l,αi、αlFor weight coefficient, and it is inversely proportional to prediction number of times, &Sigma; i = 1 m &alpha; i = 1 , &Sigma; l = 1 m &alpha; l = 1 .
Further, in step 5) in, when mobile target velocity is lower than threshold value VL-1Time, the sampling rate of minimum energy cluster node reduces by 1 times, when mobile target velocity is lower than threshold value VL-2Time, the sampling rate not yet reducing sampling rate and minimum energy cluster node reduces by 1 times, when mobile target velocity is lower than threshold value VL-m-2Time, m cluster node has reduced by 1 sampling speed all, and the sampling rate of minimum energy cluster node reduces by 2 times, by that analogy, when mobile target velocity is lower than threshold value V1Time, n-fold lower drops in the sampling rate of all nodes；Otherwise, increase the sampling rate of the highest cluster node of energy.
In step 5) in, when described measurement information is range information, the location estimation value according to the location estimation value of current time and a upper moment, extrapolate mobile target velocity.
Beneficial effects of the present invention is mainly manifested in: owing to quick sampling rate helps only small for improving the slowly moving target tracking accuracy that moves.For this problem, the invention provides the distributed method for tracking moving target dynamically adjusting sensor sampling rate.When target travel is slow, reduces the sampling rate of each cluster node and generate the speed of partial estimation, but fusion center still keeps original speed to carry out merging the quick-reaction capability estimated to ensure system.Comparing existing method for tracking target, the method is while reducing wireless sensor network energy consumption, it is ensured that the tracking accuracy of system, robustness and quick-reaction capability.
Accompanying drawing explanation
Fig. 1 is the Target Tracking System schematic diagram of wireless sensor network.
Fig. 2 adopts many group sensors to carry out merging the method for tracking moving target schematic diagram estimated.
Fig. 3 is the partial estimation flow chart of each leader cluster node.
The mobile chain objective state fusion that Fig. 4 is fusion center estimates flow chart.
Fig. 5 is multi-speed sample, estimation, fusion schematic diagram.
Detailed description of the invention
Below in conjunction with accompanying drawing, the invention will be further described.
With reference to Fig. 1～Fig. 5, a kind of method for tracking moving target merging estimation based on multi-sampling rate multi-model, said method comprising the steps of:
Step 1) speed of mobile target is divided into L different grade, wireless sensor network is divided into m different bunches, and sensor node is divided into n different sampling rate.Select the state variable (position, speed, acceleration) of mobile target, set up the state-space model of movable object tracking under n sampling rate；
Step 2) leader cluster node collects the measurement information of its bunch of interior nodes, according to state-space model under present sample speed, application extension Kalman (EKF) method obtains the partial estimation of mobile target, and partial estimation result and its dump energy information are sent to fusion center (aggregation node)；
Step 3) for low sampling rate bunch, in the moment point that its partial estimation lacks, fusion center, by the state estimation in a upper moment is predicted, was synchronized to synchronization point all local estimated result；
Step 4) fusion center error co-variance matrix according to each partial estimation, determine the fusion parameters in CI fusion method online, application CI merges method of estimation, obtains the fusion estimated result of mobile target, and described fusion estimated result includes movement rate estimated value；
Step 5) fusion center energy information according to the movement rate estimated value and each leader cluster node that move target, if the movement velocity estimated value of mobile target is lower than corresponding threshold value, will reduce the sampling rate of minimum energy cluster node；Otherwise, the movement velocity estimated value of mobile target, higher than corresponding threshold value, will accelerate the sampling rate of the highest cluster node of energy；And sampling rate adjustment result is sent to each leader cluster node；
Step 6) if leader cluster node receives sampling rate adjustment information, the sampling rate of bunch interior nodes will be adjusted, and be switched to the system model under corresponding sampling rate.Otherwise, each cluster node rate in tempo carries out sampling and state estimation.
Repeated execution of steps 1)-6), it is achieved the tracking precisely in real time to mobile target.
In step 1) described in mobile dbjective state variable include mobile target x-axis and the position of y-axis, speed and acceleration in monitor area coordinate system.Can require to determine the value of L, m and n according to the power conservation requirement of real system and tracking accuracy.Wherein, L=m × n+1, correspondence L-1 threshold speed V of L speed classc, c=1 ..., L-1 and Vc-1< Vc。
Step 2) described in measurement information be sensor node and the distance of mobile target, the mobile acceleration of target, athletic posture (direction), the range information that sensor can include ultrasound wave, infrared, laser sensor, signal attenuation (RSSI) obtain, it is also possible to include accelerometer, gyro sensor.It addition, each leader cluster node is only when obtaining measurement information, generate the partial estimation result of mobile target.
In step 3) in, if this partial estimation result of subsequent time continues disappearance, the predictive value that current time is obtained is predicted again, to obtain the state estimation of subsequent time.Also include fusion center and predicted that by successive ignition make up partial estimation engraves the disappearance of partial estimation when multiple fusion.
Step 4) in, described fusion parameters ωiDetermined online by the mark of the error co-variance matrix of each bunch of partial estimation, namelyWherein, tr (Pi,k|k)、tr(Pl,k|k) it is the mark of the error co-variance matrix of the partial estimation of bunch i, l, Pi,k|k、Pl,k|kFor the error co-variance matrix of the partial estimation of current time k bunch of i, l,αi、αlFor weight coefficient, and it is inversely proportional to prediction number of times,It addition, fusion center only stores the system state space model under the highest sampling rate.
In step 5) in, when mobile target velocity is lower than threshold value VL-1Time, the sampling rate of minimum energy cluster node reduces by 1 times, when mobile target velocity is lower than threshold value VL-2Time, the sampling rate not yet reducing sampling rate and minimum energy cluster node reduces by 1 times, when mobile target velocity is lower than threshold value VL-m-2Time (m cluster node has reduced by 1 sampling speed all), the sampling rate of minimum energy cluster node reduces by 2 times, by that analogy, when mobile target velocity is lower than threshold value V1Time, n-fold lower drops in the sampling rate of all nodes.Otherwise, increase the sampling rate of the highest cluster node of energy.Only during range information Observable, it is possible to the location estimation value according to the location estimation value of current time and a upper moment, extrapolated mobile target velocity.
As shown in Figure 1, wireless sensor network node is divided into m bunch, leader cluster node is responsible for collecting bunch interior nodes measurement information and generating the partial estimation result of position of mobile robot information, speed, acceleration, and fusion center is responsible for collecting the whole network partial estimation and carrying out State fusion estimation.Fig. 1 describes wireless sensor network environment and moves down the real-time tracking of mobile robot.Target Tracking System state estimation block diagram is as shown in Figure 2.The model of mobile apparatus people's Target Tracking System can describe by the state-space model shown in such as formula (1) and (2):
X (k)=fj(x(k-1))+Gwj(k)(1)
zi(kj)=hi(x(kj))+vi(kj) i=1 ..., mj=1 ..., n (2)
Wherein, k is discrete-time series, and i is each cluster node sequence number, and j is sampling rate grade, is divided into n grade.For the Position And Velocity state of k moment target,For a bunch k of i (partial estimation i)jThe state observation inscribed during=j (k-1)+1.For bunch i state-transition matrix under sampling rate j,Partial estimation observing matrix for bunch i.NxFor the dimension of system mode, ni,zDimension for local observation vector.Wj(k),vi(kj) for zero-mean, covariance matrix respectively QjAnd RijUncorrelated white Gaussian noise.Original state and initial covariance matrix respectively x0|0And P0|0, and x0|0With wj(k), vi(kj) statistical iteration.
Bunch i obtains partial estimation value by spreading kalman method, shown in concrete grammar such as formula (3)-(7):
x ^ i , k j + 1 | k j = f j ( x ^ i , k j | k j ) - - - ( 3 )
P i , k j + 1 | k j = F k j P i , k j | k j F k j T + Q j - - - ( 4 )
K i , k j + 1 = P k j + 1 | k j H i , k j + 1 T [ H i , k j + 1 P i , k j + 1 | k j H i , k j + 1 T + R ij ] - 1 - - - ( 5 )
P i , k j + 1 | k j + 1 = [ I - K i , k j + 1 H i , k j + 1 ] P i , k j + 1 | k j - - - ( 6 )
x ^ i , k j + 1 | k j + 1 = x ^ i , k j + 1 | k j + K i , k j + 1 [ z i , k j + 1 - h i ( x ^ k j + 1 | k j ) ] - - - ( 7 )
Wherein,It is that bunch i is at k respectivelyjThe partial estimation value of dbjective state and predictive value to subsequent time in=j (k-1)+1 sampling instant.The respectively local estimation error covariance matrix of bunch i dbjective state in j (k-1)+1 sampling instant and predicting covariance matrix to subsequent time.For the system equation Jacobian matrix under present sample speed jAndFor a bunch Jacobian matrix for i partial estimation measurement equationQj, RijThe respectively covariance matrix of the covariance matrix of system noise and bunch i partial estimation measurement noise under present sample speed j.For EKF gain, Ι is unit matrix.
Leader cluster node stores the mobile object module under each sampling rate, leader cluster node is according to different sampling rates, switch corresponding system model, the partial estimation of bunch i is obtained according to formula (3)-(7), and partial estimation result and its energy information are sent to fusion center, concrete partial estimation flow chart can refer to Fig. 3.If a bunch i reduces sampling rate, only at kjIt is carved with new observation during=j (k-1)+1 and produces and produce partial estimation value, at moment point kj=j (k-1)+1+s, s=1,2 ..., j-1 does not have new observation and local estimated value produce.As it is shown in figure 5, fusion center all carries out State fusion estimation on every moment point k, when not receiving the partial estimation of bunch i, utilize kthjThe partial estimation value of moment bunch i carries out one-step prediction and obtains kjThe partial estimation state in+1 moment, namely filtering gain is equal to zero, When the partial estimation state synchronized of all bunches to kjIn+1 moment point, application fast method determines CI fusion parameters ω onlinei, and merged estimated result accordingly.Wherein, &Sigma; i = 1 m &omega; i = 1 .
Concrete CI merges method of estimation such as formula (9)-(11)
&omega; i = &alpha; i / tr ( P i , k j + 1 | k j + 1 ) &Sigma; l = 1 m &alpha; l / tr ( P l , k j + 1 | k j + 1 ) - - - ( 9 )
P CI , k j + 1 | k j + 1 - 1 = &Sigma; i = 1 l &omega; i P i , k j + 1 | k j + 1 - 1 - - - ( 10 )
x ^ CI , k j + 1 | k j + 1 = P CI , k j + 1 | k j + 1 &Sigma; i = 1 l &omega; i P i , k j + 1 | k j + 1 - 1 x ^ i , k j + 1 | k j + 1 - - - ( 11 )
Wherein,Respectively move target at kj+1(kj+ 1=j (k-1)+2) the CI State fusion estimation value in moment and error co-variance matrix.ωiThe weights of each partial estimation adopted when being fusion estimated value and the error co-variance matrix calculating mobile dbjective state.Mark for the local estimation error covariance matrix of bunch i.αiFor weight coefficient, it is inversely proportional to prediction number of times, and requirement &Sigma; i = 1 m &alpha; i = 1 .
Fusion center carries out State fusion estimation according to formula (9)-(11), obtains more accurate estimated result.If the movement velocity of mobile target is lower than corresponding threshold value, the sampling rate of minimum energy cluster node will be reduced；Otherwise, the movement velocity of mobile target, higher than corresponding threshold value, will accelerate the sampling rate of the highest cluster node of energy.Each leader cluster node that fusion center is sent to the adjustment result of sampling rate, leader cluster node adjusts the sampling rate of bunch interior nodes accordingly and is switched to corresponding state-space model.
A kind of system adopting the distributed mobile object real-time tracking organizing sensor multi-sampling rate suitable in wireless transducer network energy saving more, described system includes:
Bunch, in wireless sensor network, the sub-network being made up of several sensor nodes, bunch interior node is referred to as cluster node.
Leader cluster node, is responsible for a bunch node for inner sensor child node, computing capability and energy reserve and will be higher than other common nodes.For generating the partial estimation of mobile dbjective state, and partial estimation result is sent to fusion center (aggregation node).
Aggregation node, is responsible for setting up and managing whole wireless sensor network, collects the partial estimation result of all leader cluster nodes, and generate fusion estimated result.Power typically via fixed power source, there is the strongest disposal ability.

Claims (9)

1. one kind merges the method for tracking moving target estimated based on multi-sampling rate multi-model, it is characterised in that: said method comprising the steps of:
Step 1) speed of mobile target is divided into L different grade, wireless sensor network be divided into m different bunch, sensor node is divided into n different sampling rate, select the state variable of mobile target, described state variable is position, speed or acceleration, sets up the state-space model of movable object tracking under n sampling rate；
Step 2) leader cluster node collects the measurement information of its bunch of interior nodes, according to state-space model under present sample speed, application extension kalman filter method obtains the partial estimation of mobile target, and partial estimation result and its dump energy information are sent to fusion center；
Step 3) for low sampling rate bunch, in the moment point that its partial estimation lacks, fusion center, by the state estimation in a upper moment is predicted, was synchronized to synchronization point all local estimated result；
Step 4) fusion center error co-variance matrix according to each partial estimation, determine the fusion parameters in CI fusion method online, application CI merges method of estimation, obtains the fusion estimated result of mobile target, and described fusion estimated result includes movement velocity estimated value；
Step 5) fusion center energy information according to the movement velocity estimated value and each leader cluster node that move target, if the movement velocity estimated value of mobile target is lower than corresponding threshold value, will reduce the sampling rate of minimum energy cluster node；Otherwise, if the movement velocity estimated value of mobile target is higher than corresponding threshold value, the sampling rate of the highest cluster node of energy will be accelerated；And sampling rate adjustment result is sent to each leader cluster node；
Step 6) if leader cluster node receives sampling rate adjustment information, the sampling rate of bunch interior nodes will be adjusted, and be switched to the state-space model under corresponding sampling rate, otherwise, each cluster node is undertaken sampling and state estimation by former sampling rate.
2. the method for tracking moving target merging estimation based on multi-sampling rate multi-model as claimed in claim 1, it is characterized in that: described step 1) in, power conservation requirement and tracking accuracy according to system require to determine the value of L, m and n, wherein, correspondence L-1 threshold speed V of L=m × n+1, L speed classc, c=1 ..., L-1 and Vc-1< Vc。
3. the method for tracking moving target merging estimation based on multi-sampling rate multi-model as claimed in claim 2, it is characterised in that: described step 1) in, described state variable is two dimension or three-dimensional parameter.
4. the method for tracking moving target merging estimation based on multi-sampling rate multi-model as claimed in claim 2, it is characterized in that: described step 1) in, state-space model under each sampling rate is stored in leader cluster node, each state-space model is pointed to by array of pointers M [n], i.e. M [j], j=1 ..., n points to the state-space model under sampling rate j.
5. the method for tracking moving target merging estimation based on multi-sampling rate multi-model as described in one of Claims 1 to 4, it is characterized in that: described step 2) in, described measurement information is sensor node and the distance of mobile target, the acceleration moving target and athletic posture.
6. the method for tracking moving target merging estimation based on multi-sampling rate multi-model as described in one of Claims 1 to 4, it is characterized in that: in step 3) in, if this partial estimation result of subsequent time continues disappearance, the predictive value that current time is obtained is predicted again, to obtain the state estimation of subsequent time.
7. the method for tracking moving target merging estimation based on multi-sampling rate multi-model as described in one of Claims 1 to 4, it is characterised in that: in step 4) in, described fusion parameters ωiDetermined online by the mark of the error co-variance matrix of each bunch of partial estimation, namelyWherein, tr (Pi,k|k)、tr(Pl,k|k) it is the mark of the error co-variance matrix of the partial estimation of bunch i, l, Pi,k|k、Pl,k|kFor the error co-variance matrix of the partial estimation of current time k bunch of i, l,αi、αlFor weight coefficient, and it is inversely proportional to prediction number of times,
8. the method for tracking moving target merging estimation based on multi-sampling rate multi-model as described in one of Claims 1 to 4, it is characterised in that: in step 5) in, when mobile target velocity is lower than threshold value VL-1Time, the sampling rate of minimum energy cluster node reduces by 1 times, when mobile target velocity is lower than threshold value VL-2Time, the sampling rate not yet reducing sampling rate and minimum energy cluster node reduces by 1 times, when mobile target velocity is lower than threshold value VL-m-2Time, m cluster node has reduced by 1 sampling speed all, and the sampling rate of minimum energy cluster node reduces by 2 times, by that analogy, when mobile target velocity is lower than threshold value V1Time, n-fold lower drops in the sampling rate of all nodes；Otherwise, increase the sampling rate of the highest cluster node of energy.
9. the method for tracking moving target merging estimation based on multi-sampling rate multi-model as claimed in claim 8, it is characterized in that: in step 5) in, when described measurement information is range information, location estimation value according to current time and the location estimation value in a upper moment, extrapolate mobile target velocity.
CN201410223531.7A 2014-05-23 2014-05-23 A kind of method for tracking moving target merging estimation based on multi-sampling rate multi-model Active CN104090262B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410223531.7A CN104090262B (en) 2014-05-23 2014-05-23 A kind of method for tracking moving target merging estimation based on multi-sampling rate multi-model

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410223531.7A CN104090262B (en) 2014-05-23 2014-05-23 A kind of method for tracking moving target merging estimation based on multi-sampling rate multi-model

Publications (2)

Publication Number Publication Date
CN104090262A CN104090262A (en) 2014-10-08
CN104090262B true CN104090262B (en) 2016-06-29

Family

ID=51637993

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410223531.7A Active CN104090262B (en) 2014-05-23 2014-05-23 A kind of method for tracking moving target merging estimation based on multi-sampling rate multi-model

Country Status (1)

Country Link
CN (1) CN104090262B (en)

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104833949A (en) * 2015-05-11 2015-08-12 西北工业大学 Multiple-unmanned aerial vehicle cooperative passive location method based on improved distance parameterization
CN104994579B (en) * 2015-06-30 2018-08-17 浙江工业大学 A kind of distributed method for tracking moving target suitable for event triggering
CN105933958A (en) * 2016-04-18 2016-09-07 北京科技大学 Wireless sensor network moving object positioning device and method based on clustering
CN108594652B (en) * 2018-03-19 2021-05-25 江苏大学 Observer information iteration-based vehicle state fusion estimation method
CN110163270B (en) * 2019-05-10 2021-11-09 北京易控智驾科技有限公司 Multi-sensor data fusion method and system
CN111220145A (en) * 2019-11-28 2020-06-02 湖南率为控制科技有限公司 Fusion method of navigation data under multiple sampling rates for automatic driving automobile
CN111829521B (en) * 2020-06-23 2022-05-03 浙江工业大学 Consistent target tracking method based on data driving
CN111883265A (en) * 2020-06-30 2020-11-03 南京理工大学 Target state estimation method applied to fire control system
CN112068080A (en) * 2020-09-09 2020-12-11 中国航空工业集团公司雷华电子技术研究所 Radar sampling method, computer device, and medium

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101511099A (en) * 2009-04-01 2009-08-19 南京邮电大学 Collection method for wireless sensor network data based on time series prediction model
CN103052128A (en) * 2012-12-20 2013-04-17 华南理工大学 Wireless sensor network-based energy-efficient collaborative scheduling method

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101511099A (en) * 2009-04-01 2009-08-19 南京邮电大学 Collection method for wireless sensor network data based on time series prediction model
CN103052128A (en) * 2012-12-20 2013-04-17 华南理工大学 Wireless sensor network-based energy-efficient collaborative scheduling method

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
"Energy Management in Wireless Sensor Networks with Energy-hungry Sensors";Cesare Alippi et al.;《IEEE Instrumentation and Measurement Magazine》;20090430;第12卷(第2期);第16-23页 *
"Energy-Efficient Fusion Estimation for Wireless Sensor Networks with Packet Losses";ZHANG Wen-An et al.;《Proceedings of the 30th Chinese Control Conference》;20110724;第6408-6412页 *
"基于无线传感网的目标跟踪算法及实验系统设计";俞立 等;《浙江工业大学学报》;20121231;第40卷(第6期);第649-652页 *

Also Published As

Publication number Publication date
CN104090262A (en) 2014-10-08

Similar Documents

Publication Publication Date Title
CN104090262B (en) A kind of method for tracking moving target merging estimation based on multi-sampling rate multi-model
CN104076355B (en) Tracking before Dim targets detection in strong clutter environment based on dynamic programming
CN109459040B (en) Multi-AUV (autonomous Underwater vehicle) cooperative positioning method based on RBF (radial basis function) neural network assisted volume Kalman filtering
Chen et al. A hybrid prediction method for bridging GPS outages in high-precision POS application
CN102830402B (en) Target tracking system and method for underwater sensor network
CN107677997B (en) Extended target tracking method based on GLMB filtering and Gibbs sampling
CN110989352B (en) Group robot collaborative search method based on Monte Carlo tree search algorithm
CN110244715B (en) Multi-mobile-robot high-precision cooperative tracking method based on ultra wide band technology
CN104898148B (en) A kind of inexpensive INS/GPS seamless navigations method based on data compression and neutral net
CN104766320B (en) Many Bernoulli Jacob under thresholding is measured filter Faint target detection and tracking
Kim et al. UWB positioning system based on LSTM classification with mitigated NLOS effects
CN103648108B (en) Sensor network distributed consistency object state estimation method
CN104778358B (en) The partly overlapping extension method for tracking target in monitored area be present in multisensor
CN108008099B (en) Pollution source positioning method
Song et al. Neural-network-based AUV navigation for fast-changing environments
CN107728138A (en) A kind of maneuvering target tracking method based on current statistical model
CN108645413A (en) The dynamic correcting method of positioning and map building while a kind of mobile robot
CN107589748A (en) AUV autonomous navigation methods based on UnscentedFastSLAM algorithms
CN105717505A (en) Data association method for utilizing sensing network to carry out multi-target tracking
CN101543400A (en) Aanimal behavior detection and automatic analysis system and animal behavior analysis methods
CN103047982B (en) Adaptive target tracking method based on angle information
CN111862165A (en) Target tracking method for updating Kalman filter based on deep reinforcement learning
CN105444763A (en) IMU indoor positioning method
CN105022055A (en) IMU indoor positioning method
CN110749891A (en) Self-adaptive underwater single beacon positioning method capable of estimating unknown effective sound velocity

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