CN114840801A - Distributed target tracking method and system suitable for directed switching topology - Google Patents
Distributed target tracking method and system suitable for directed switching topology Download PDFInfo
- Publication number
- CN114840801A CN114840801A CN202210384402.0A CN202210384402A CN114840801A CN 114840801 A CN114840801 A CN 114840801A CN 202210384402 A CN202210384402 A CN 202210384402A CN 114840801 A CN114840801 A CN 114840801A
- Authority
- CN
- China
- Prior art keywords
- sensor node
- target state
- information
- ith
- representing
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 47
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 20
- 238000001914 filtration Methods 0.000 claims abstract description 19
- 239000011159 matrix material Substances 0.000 claims description 150
- 239000013598 vector Substances 0.000 claims description 75
- 238000005259 measurement Methods 0.000 claims description 39
- 239000000126 substance Substances 0.000 claims description 38
- 238000005070 sampling Methods 0.000 claims description 37
- 238000004364 calculation method Methods 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 10
- 238000012546 transfer Methods 0.000 claims description 7
- 238000012937 correction Methods 0.000 claims description 6
- 230000008878 coupling Effects 0.000 claims description 6
- 238000010168 coupling process Methods 0.000 claims description 6
- 238000005859 coupling reaction Methods 0.000 claims description 6
- 238000012986 modification Methods 0.000 claims description 3
- 230000004048 modification Effects 0.000 claims description 3
- 239000002775 capsule Substances 0.000 claims 1
- 239000002245 particle Substances 0.000 claims 1
- 238000004891 communication Methods 0.000 abstract description 22
- 230000007246 mechanism Effects 0.000 abstract description 5
- 230000009466 transformation Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000004088 simulation Methods 0.000 description 2
- 239000000243 solution Substances 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000002347 injection Methods 0.000 description 1
- 239000007924 injection Substances 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- 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
-
- 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
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- 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
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention relates to a distributed target tracking method and a system suitable for directed switching topology, wherein the method comprises the following steps: predicting the target state of each sensor node by adopting a volume information filtering algorithm to obtain target state prediction information; and carrying out consistency encryption on target state prediction information transmitted among the sensor nodes based on a pushing and average consistency algorithm of state decomposition. On the basis of a distributed volume information filtering algorithm, the privacy protection mechanism based on state decomposition is utilized to ensure the privacy of the local information of the nodes, and the algorithm is suitable for directed switching topology by combining a push-sum average consistency method. The algorithm can carry out privacy protection on the state of each node on the premise of ensuring the estimation precision, prevent the locally estimated convergence value from being acquired by an eavesdropper, and meanwhile, can also be robust to the unidirectional communication network of the transformation structure.
Description
Technical Field
The invention relates to the technical field of cooperative target tracking, in particular to a distributed target tracking method and a distributed target tracking system suitable for directed switching topology.
Background
The positioning and tracking of maneuvering targets is an important field in present scientific research, and is widely applied in military reconnaissance, mobile robots, traffic control and the like. With the development of network communication technology, in order to complete a target tracking task with higher precision at lower cost, collaborative state estimation is applied to a wireless sensor network. However, the communication between the nodes of the sensor network, which is completely transparent to the real state information of the target, has the risk of privacy disclosure, and the privacy of the local values of the nodes in the network is virtually violated, so that the sensor network is easily attacked by an eavesdropper, and such information security hidden danger can bring fatal one-click to the collaborative estimation network containing sensitive information or needing to keep the state information secret at a certain moment.
Conventional privacy protection methods are noise injection and coding decoding of information, but are not suitable for low power consumption sensor networks due to the large amount of computation and communication load they require. Some privacy protection consistency algorithms based on the fuzzy mechanism also influence the consistency convergence precision due to differential privacy. Therefore, a privacy protection mechanism based on state decomposition is provided, which can achieve the same and accurate estimation value while protecting the local state information of each node through lightweight computation under a distributed framework, and is more suitable for a sensor network with limited computation and communication resources.
However, most communication networks in practical application are directed networks, and especially for the problem of tracking a maneuvering target under limited communication conditions such as battlefields and complex environments, the situations of one-way communication and asymmetric change of the communication network structure are easy to occur. The existing privacy protection algorithm based on state decomposition only considers the sensor network of undirected communication, is not suitable for a target tracking task requiring communication protection under directed switching topology, and has higher requirement on ideal communication conditions.
Disclosure of Invention
In view of this, the present invention provides a distributed target tracking method and system suitable for a directed switching topology, so as to implement a target tracking task requiring communication protection under the directed switching topology.
In order to achieve the purpose, the invention provides the following scheme:
a distributed target tracking method suitable for use in a directed switching topology, the method comprising the steps of:
predicting the target state of each sensor node by adopting a volume information filtering algorithm to obtain target state prediction information;
and carrying out consistency encryption on target state prediction information transmitted among the sensor nodes based on a pushing and average consistency algorithm of state decomposition.
Optionally, the predicting the target state of each sensor node by using the volume information filtering algorithm to obtain target state prediction information specifically includes:
initializing and acquiring a target state information vector and a predictive value covariance matrix of each sensor node at an initial moment; the target state information vector and the predictive value covariance matrix at the initial moment are used for predicting target state prediction information at the next moment at the initial moment;
predicting and acquiring target state prediction information of each sensor node at the k moment according to the target state prediction information of each sensor node at the k-1 moment;
and correcting the target state prediction information of each sensor node at the k moment by using the target observation value of each sensor node at the k moment.
Optionally, the predicting and obtaining the target state prediction information at the k time of each sensor node according to the target state prediction information at the k-1 time of each sensor node specifically includes:
according to the target state prediction information at the k-1 moment, calculating volume sampling points at the k-1 moment as follows:
wherein the content of the first and second substances,
represents the calculated r-th volume sample point at time k-1 of the ith sensor node,
representing a covariance matrix of the ith sensor node at the time k-1; c. C r A matrix representing dimensions n x 2n,
e s a column vector representing the n dimensions is represented,
s represents c r The number of columns of (a) is,
representing the target state predicted value of the corrected ith sensor node at the moment of k-1;
predicting and obtaining a target state predicted value at the k moment according to the volume sampling points at the k-1 moment:
wherein the content of the first and second substances,
an r volume sampling point representing the ith sensor node at the predicted k time, f (-) represents the nonlinear state transfer function of the target,
indicating the target state prediction value at the moment k obtained by prediction,
representing the covariance matrix, Q, of the predicted i-th sensor node at time k k-1 A covariance matrix representing the noise of the target motion process at the k-1 moment;
acquiring target state prediction information of each sensor node at the k moment according to the target state prediction value of each sensor node at the k moment, wherein the target state prediction information of each sensor node at the k moment is as follows:
wherein the content of the first and second substances,
a target state information matrix representing the ith sensor node at time k,
and representing the target state information vector of the ith sensor node at the time k.
Optionally, the modifying, by using the target observation value at the time k of each sensor node, the target state prediction information at the time k of each sensor node specifically includes:
calculating a target state measurement value of each volume sampling point of the ith sensor node through a nonlinear observation function;
respectively taking the average value of the target state measurement values corresponding to each volume sampling point of the ith sensor node as the target state observation value of the ith sensor node;
calculating a measured value covariance matrix of the ith sensor node according to the target state measured value and the target state observed value of each volume sampling point of the ith sensor node;
calculating a cross covariance matrix of the predicted value and the measured value of the ith sensor node according to the target state predicted value and the target state measured value of each volume sampling point;
according to the measured value covariance matrix of the ith sensor node and the cross covariance matrix of the predicted value and the measured value, determining a pseudo measurement matrix of the ith sensor node as follows:
wherein the content of the first and second substances,
a pseudo measurement matrix representing the ith sensor node,
a cross covariance matrix representing the predicted value and the measured value of the ith sensor node;
according to the pseudo measurement matrix of the ith sensor node, determining the information increment and the associated information matrix of the ith sensor node at the time k as follows:
wherein, the first and the second end of the pipe are connected with each other,
the information increment of the ith sensor node at the moment k,
a covariance matrix representing the measured noise of the ith sensor node at time k,
white gaussian noise representing an average of 0;
correcting target state prediction information of the ith sensor node at the time k by using the following formula according to the information increment of the ith sensor node at the time k and the associated information matrix;
wherein the content of the first and second substances,
a target state information vector representing the corrected k-time of the i-th sensor node,
and representing the corrected target state information matrix of the ith sensor node at the k moment.
Optionally, the performing consistency encryption on the target state prediction information transmitted between the sensor nodes by using the state decomposition-based push and average consistency algorithm specifically includes:
the updating mode of the target state prediction information transmitted between the adjacent sensor nodes is as follows:
wherein l represents the number of consistent iterations of the target state prediction information,
and
respectively representing alpha items of sub information in target state information vectors of the ith sensor node at the k time under the l iteration and the l +1 iteration,
and
respectively representing beta items of sub information in a target state information vector at the k time under the ith iteration and the l +1 iteration of the ith sensor node,
and
respectively representing alpha items of sub information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration,
and
respectively represents beta items of sub information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration,
representing the consistency weight within the ith sensor node,
the consistency weight between the ith sensor node and the jth sensor node is shown, omega is the consistency weight of the sub information exchange in each sensor node,
coupling weight for information exchange between two sub-states in the ith sensor node;
the updating mode of the deduction item corresponding to the target state prediction information transmitted between the adjacent sensor nodes is as follows:
wherein the content of the first and second substances,
and
respectively representing the sum of alpha item sub information in the target state information vector of the ith sensor node at the k time under the l iteration and the l +1 iteration;
and
a sum term of beta item sub information in a target state information vector at the k time under the ith iteration and l +1 iteration of the ith sensor node is represented;
a sum term of alpha item sub information in a target state information vector of a jth sensor node at the k moment under the ith iteration is represented;
and
respectively representing the sum of alpha item sub-information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration;
and
respectively representing the sum of beta item sub-information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration;
a sum item of alpha item sub information in a target state information matrix at the k moment of the jth sensor node under the ith iteration is represented;
a set of all neighboring sensor nodes representing that the ith sensor node is capable of receiving information;
and
is satisfied with the initial value of iteration
And
is satisfied with the initial value of iteration
Wherein 1 is n×1 Representing an n-dimensional all-1 vector, 1 n×n Representing a full 1-dimensional square matrix of n dimensions.
The consistency encryption result of the target state prediction information transmitted between the adjacent sensor nodes is as follows:
wherein the content of the first and second substances,
a consistent encrypted target state information vector representing the ith sensor node,
and representing a consistency encryption target state information matrix of the ith sensor node.
A distributed target tracking system adapted for use in a directed handover topology, the system comprising:
the target state prediction module is used for predicting the target state of each sensor node by adopting a volume information filtering algorithm to obtain target state prediction information;
and the consistency encryption module is used for carrying out consistency encryption on the target state prediction information transmitted among the sensor nodes based on a pushing and average consistency algorithm of state decomposition.
Optionally, the target state prediction module specifically includes:
the initialization submodule is used for initializing and acquiring a target state information vector and a predictive value covariance matrix of each sensor node at the initial moment; the target state information vector and the predictive value covariance matrix at the initial moment are used for predicting target state prediction information at the next moment at the initial moment;
the target state prediction information prediction submodule is used for predicting and acquiring target state prediction information of each sensor node at the k moment according to the target state prediction information of each sensor node at the k-1 moment;
and the target state prediction information correction submodule is used for correcting the target state prediction information of each sensor node at the k moment by using the target observation value of each sensor node at the k moment.
Optionally, the target state prediction information prediction sub-module specifically includes:
and the volume sampling point calculation unit is used for calculating the volume sampling points at the k-1 moment as follows according to the target state prediction information at the k-1 moment:
wherein the content of the first and second substances,
denotes the calculated r-th volume sample point at the time k-1 of the i-th sensor node, 2n denotes the number of volume sample points,
representing a covariance matrix of the ith sensor node at the time k-1; c. C r A matrix representing dimensions n x 2n,
e s a column vector representing the n dimensions is represented,
s represents c r The number of columns of (a) is,
representing the target state predicted value of the corrected ith sensor node at the moment of k-1;
and the target state predicted value prediction unit is used for predicting and acquiring a target state predicted value at the k moment according to the volume sampling points at the k-1 moment as follows:
wherein the content of the first and second substances,
an r volume sampling point representing the ith sensor node at the predicted k time, f (-) represents the nonlinear state transfer function of the target,
indicating the target state prediction value at the moment k obtained by prediction,
representing the covariance matrix, Q, of the predicted i-th sensor node at time k k-1 A covariance matrix representing the noise of the target motion process at the k-1 moment;
a target state prediction information obtaining unit, configured to obtain, according to the target state prediction value at the time k of each sensor node, target state prediction information at the time k of each sensor node:
wherein the content of the first and second substances,
a target state information matrix representing the ith sensor node at time k,
and representing the target state information vector of the ith sensor node at the time k.
Optionally, the target state prediction information modification sub-module specifically includes:
the target state measurement value calculation unit is used for calculating a target state measurement value of each volume sampling point of the ith sensor node through a nonlinear observation function;
the target state observation value calculation unit is used for respectively serving as a target state observation value of the ith sensor node according to the average value of the target state measurement values corresponding to each volume sampling point of the ith sensor node;
the cross covariance matrix calculation unit is used for calculating a cross covariance matrix of the predicted value and the measured value of the ith sensor node according to the target state predicted value and the target state measured value of each volume sampling point;
the pseudo measurement matrix determining unit is used for determining a pseudo measurement matrix of the ith sensor node as follows according to the measured value covariance matrix of the ith sensor node and the cross covariance matrix of the predicted value and the measured value:
wherein, the first and the second end of the pipe are connected with each other,
a pseudo measurement matrix representing the ith sensor node,
a cross covariance matrix representing the predicted value and the measured value of the ith sensor node;
an adjustment quantity determining unit, configured to determine, according to the pseudo measurement matrix of the ith sensor node, that the information increment and the associated information matrix of the ith sensor node at the time k are:
wherein the content of the first and second substances,
the information increment of the ith sensor node at the moment k,
a covariance matrix representing the measured noise of the ith sensor node at time k,
white gaussian noise representing an average of 0;
the information correction unit is used for correcting the target state prediction information of the ith sensor node at the time k by using the following formula according to the information increment of the ith sensor node at the time k and the associated information matrix;
wherein the content of the first and second substances,
a target state information vector representing the corrected k-time of the i-th sensor node,
and representing the corrected target state information matrix of the ith sensor node at the k moment.
Optionally, the consistency encryption module specifically includes:
the updating mode of the target state prediction information transmitted between the adjacent sensor nodes is as follows:
wherein l represents the number of consistent iterations of the target state prediction information,
and
respectively represent the ith sensor nodeThe alpha-term sub-information in the target state information vector at time k at the i-th and l +1 iterations,
and
respectively represents beta items of sub information in a target state information vector at the k time under the l iteration and the l +1 iteration of the ith sensor node,
and
respectively representing alpha items of sub information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration,
and
respectively represents beta items of sub information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration,
representing the consistency weight within the ith sensor node,
the consistency weight between the ith sensor node and the jth sensor node is shown, omega is the consistency weight of the sub information exchange in each sensor node,
coupling weight for information exchange between two sub-states in the ith sensor node;
the updating mode of the deduction item corresponding to the target state prediction information transmitted between the adjacent sensor nodes is as follows:
wherein the content of the first and second substances,
and
respectively representing the sum of alpha item sub information in the target state information vector of the ith sensor node at the k time under the l iteration and the l +1 iteration;
and
a sum term of beta item sub information in a target state information vector at the k time under the ith iteration and l +1 iteration of the ith sensor node is represented;
a sum term of alpha item sub information in a target state information vector of a jth sensor node at the k moment under the ith iteration is represented;
and
respectively representing the sum of alpha item sub-information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration;
and
respectively representing the sum of beta item sub-information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration;
a sum item of alpha item sub information in a target state information matrix at the k moment of the jth sensor node under the ith iteration is represented;
a set of all neighboring sensor nodes representing that the ith sensor node is capable of receiving information;
and
is satisfied with the initial value of iteration
And
is satisfied with the initial value of iteration
Wherein 1 is n×1 All-1-vectors, 1, representing n dimensions n×n A full 1-dimensional matrix representing n dimensions;
the consistency encryption result of the target state prediction information transmitted between the adjacent sensor nodes is as follows:
wherein the content of the first and second substances,
a consistent encrypted target state information vector representing the ith sensor node,
and representing a consistency encryption target state information matrix of the ith sensor node.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the invention discloses a distributed target tracking method and a system suitable for directed switching topology, wherein the method comprises the following steps: predicting the target state of each sensor node by adopting a volume information filtering algorithm to obtain target state prediction information; and carrying out consistency encryption on target state prediction information transmitted among the sensor nodes based on a pushing and average consistency algorithm of state decomposition. On the basis of a distributed volume information filtering algorithm, the privacy protection mechanism based on state decomposition is utilized to ensure the privacy of the local information of the nodes, and the algorithm is suitable for directed switching topology by combining a push-sum average consistency method. The algorithm can carry out privacy protection on the state of each node on the premise of ensuring the estimation precision, prevent the locally estimated convergence value from being acquired by an eavesdropper, and meanwhile, can also be robust to the unidirectional communication network of the transformation structure.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.
Fig. 1 is a flowchart of a distributed target tracking method suitable for a directed switching topology according to embodiment 1 of the present invention;
fig. 2 is a schematic diagram of a distributed target tracking method suitable for a directed switching topology according to embodiment 1 of the present invention;
fig. 3 is a diagram of a directed switching topology communication structure of a multi-sensor network node according to embodiment 1 of the present invention;
FIG. 4 is a comparison graph of the real trajectory of the moving object and the estimated trajectories of CIF, C-CIF, PC-CIF provided in embodiment 2 of the present invention;
FIG. 5 is a comparison graph of the real state of the moving object and the estimated state of the PC-CIF according to embodiment 2 of the present invention;
FIG. 6 is a comparison graph of the position root mean square error of the CIF, C-CIF, PC-CIF algorithms provided in embodiment 2 of the present invention;
FIG. 7 is a comparison graph of root mean square error of velocity for CIF, C-CIF, PC-CIF algorithms provided in embodiment 2 of the present invention;
fig. 8 is a diagram of the convergence effect of lateral position information and eavesdropper estimation values of a single sensor node according to embodiment 2 of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention aims to provide a distributed target tracking method and a distributed target tracking system which are suitable for a directed switching topology, so as to realize a target tracking task requiring communication protection under the directed switching topology.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
The invention relates to a collaborative nonlinear target tracking algorithm for a multi-sensor network, which is characterized in that a collaborative estimation algorithm with a privacy protection function is designed on the basis of a volume information filtering algorithm, a privacy protection mechanism based on state decomposition is adopted, a push and average consistency algorithm is adopted, the robustness of the algorithm under a directed switching topology is increased, and the estimation precision is ensured while the algorithm security is improved.
As shown in fig. 1 and 2, the present invention provides a distributed target tracking method suitable for a directed switching topology, including the following steps:
in fig. 2, maneveringtarget represents a maneuvering target, observer model represents an observation model, sensor represents a sensor, dvnamic model represents a dynamic model, stateprogress represents state observation, StateUpdate represents state update, weightsupate represents weight update, substtates Consensus information, statedecompensation represents state decomposition, Privacy-preserving Push-sum, and LocalCIF represents a local CIF (cubatelnformationfilter).
(I) constructing a target motion model and a sensor observation model
The nonlinear motion model of a single maneuvering target is:
x k ＝f(x k-1 )+w k-1 ；
wherein x is k Representing the n-dimensional state vector of the target at time k, f (-) representing the nonlinear state transfer function, w k-1 Is white Gaussian noise with mean 0, representing a covariance matrix of Q at the previous moment k-1 Is measured.
The sensor network consists of N sensor nodes, and the observation model of the ith sensor is as follows:
wherein the content of the first and second substances,
representing m dimensions of the target for the ith sensor pair at time kState observation vector, h (·) i Represents the non-linear measurement function of the ith sensor,
is white Gaussian noise with a mean value of 0, representing the covariance matrix of the ith sensor at time k
M-dimensional measurement noise.
(II) Single-node local volume information filtering estimation
The volume information filtering algorithm of the single node mainly executes two links of state prediction and updating according to the flow.
Firstly, initializing the estimation value of each sensor node when k is 0, and taking the state estimation and the covariance matrix as follows:
then, the state of the target at time k is predicted. In this link, 2n volume samples are first calculated based on the estimate at time k-1:
wherein, c r Satisfies the following conditions:
and e s Is an n-dimensional column vector:
substituting the state sampling point at the time of k-1 into a nonlinear state transfer function to calculate a predicted value of the state sampling point at the next time, taking the average value of the sampling points as the predicted value of the sensor node to the target state at the time of k, and then:
thus, the predicted information matrix and information vector can be calculated as:
and finally, updating and correcting the target prediction state through the observation value of the sensor node at the moment k. Based on a prediction link, calculating the corresponding measurement values of 2n state sampling points through a nonlinear observation function, and taking the average value as the prediction measurement value of k time:
and then calculating the covariance matrix of the predicted measurement values and the cross covariance matrix between the states and the measurements:
the pseudo measurement matrix is found by:
based on the matrix, obtaining an information increment and a correlation information matrix of the ith sensor at the time k:
finally, updating the information vector and the information matrix as follows:
102, performing consistency encryption on target state prediction information transmitted among the sensor nodes based on a state decomposition push and average consistency algorithm, specifically comprising the following steps:
at the k moment in the filtering process, the information vector and the information matrix are used as consistency information pairs, and the information vector of the sensor node i is initially processed in consistency iteration
Decomposition into satisfying equations
Two sub-states of
And
the corresponding sub-information matrix of the same satisfies
The consistency process at time k can be expressed as:
where l represents the number of consistency iterations. In addition, the iterative initial values of the sum-pushing items corresponding to the sub information vectors and the sub information matrixes respectively satisfy
In consistencyIn the iterative process, each sensor node broadcasts sub-information to adjacent points
But only pass the substate to itself
The information of (1). Similarly, each sensor node receives the alpha sub-information sent by the adjacent point and only receives the beta sub-information transmitted by the sensor node. Thus, in the (l + 1) th iteration, the information pair is updated as:
wherein, w k Is a consistency weight that varies over time and satisfies:
omega is the sensor node internal consistency weight,
coupling weights for information exchange between two sub-states inside the sensor node are obtained, and the sensor node i is kept secret.
Representing the set of all the adjacency points that the sensor node i can receive the information. Similarly, the updates of the push-sum term corresponding to the information pair in the (l + 1) th iteration are respectively:
finally, take the proportion
And
as a final result of the consistency of the information vector and the information matrix:
at the k moment in the filtering process, the information vector and the information matrix are used as consistency information pairs, and the information vector of the sensor node i is initially processed in consistency iteration
Decomposition into satisfying equations
Two sub-states of
And
the corresponding sub information matrix also satisfies
The consistency process at time k can be expressed as:
wherein l representsThe number of consistent iterations. In addition, the iterative initial values of the sum-pushing items corresponding to the sub information vectors and the sub information matrixes respectively satisfy
In the consistency iteration process, each sensor node broadcasts sub-information to adjacent points thereof
But only pass the substate to itself
The information of (1). Similarly, each sensor node receives the alpha sub-information sent by the adjacent point and only receives the beta sub-information transmitted by the sensor node. Thus, in the l +1 th iteration, the information pair is updated as:
wherein, w k Is a consistency weight that varies over time and satisfies:
omega is the sensor node internal consistency weight,
coupling weights for information exchange between two sub-states inside the sensor node are obtained, and the sensor node i is kept secret.
Indicating that sensor node i can receive the messageAnd all adjacency points of information. Similarly, the updates of the push-sum term corresponding to the information pair in the (l + 1) th iteration are respectively:
finally, taking the proportion
And
as a final result of the consistency of the information vector and the information matrix:
example 2
Representing the state vector of the tracking target as
Wherein (x) k ，y k ，z k ) Representing the three-dimensional coordinates of the object at time k,
the velocity of the target in the X, Y, Z directions at time k is shown. Representing the measurement vector of the ith sensor node as
Wherein
Object representation with sensorThe relative distance of (a) to (b),
the pitch angle is expressed in terms of,
indicating the yaw angle. Suppose the three-dimensional coordinates of sensor i are
The following relationship is satisfied between the measured value and the state value:
considering a multi-sensor network composed of 4 sensor nodes, the communication topology of the system is shown in fig. 3, eavesdropper represents an eavesdropper, and host-but-curous node 2 represents a relevant sensor node 2. Communication topologies are respectively at T 1 ＝40，T 2 ＝100，T 3 Switching is performed at three times of 150. A consistency weight matrix is given based on the Metropolis weight rule:
wherein, delta represents the sum of the maximum out-degree and the in-degree of all the sensor nodes in the sensor network, and epsilon represents an edge set in the directed graph of the sensor network. To evaluate the privacy-preserving effect of the algorithm, it is assumed that there is an external eavesdropper interested in the state information of the sensor node 1 and that all the consistency weights of the sensor node communicating with its neighbors are known. The eavesdropper has the following observer:
in order to verify the effectiveness of the invention, 200s simulation verification is carried out on the proposed distributed volume information filtering algorithm based on the state decomposition push and average consistency algorithm. Firstly, the tracking accuracy is shown, and fig. 4 shows the comparison between the real track of the moving object and the estimated track of CIF (volume Information Filter), C-CIF (sense-based volume Information Filter, consistency-based volume Information Filter), PC-CIF (Privacy-preserving Push-sum sense-based volume Information Filter, Privacy-preserving Push-based volume Information Filter, consistency-based volume Information Filter), so that it can be seen intuitively that the estimated accuracy of C-CIF and PC-CIF is maintained at the same level, and is better than that of a single sensor node CIF. Fig. 5 further shows the comparison between the Actual Track state of the moving target and the estimated state of the PC-CIF, and it can be seen that the estimated state of the proposed algorithm and the Actual state of the target increase with time tend to be consistent, confirming the validity of the state estimation result of the distributed target tracking method of the present invention. The nodes in fig. 6-8 represent sensor nodes, and according to evaluation indexes, fig. 6 and 7 respectively show position root mean square errors and speed root mean square errors of three methods of CIF, C-CIF and PC-CIF, so that it can be more accurately seen that the estimation accuracy of the original consistency filtering algorithm based on state decomposition and the average consistency algorithm cannot be influenced, and the superiority of the original algorithm relative to single-sensor node state estimation is retained. Fig. 8 verifies the effectiveness of the algorithm in privacy protection of the directed switching topology network, average in fig. 8 represents average, and fig. 8 shows the convergence condition of the sub-state of the sensor node 1 and the observed value of the eavesdropper (eavesdropp) at 30 iteration times: it can be found that as the number of consistent iterations increases, the sub-state of the sensor node 1 gradually converges to the average value of the initial local and adjacent point states of the iteration, and the observed value of the eavesdropper gradually deviates from the true value of the sensor node 1, thereby protecting the local state of the sensor node 1. In conclusion, the distributed nonlinear target tracking estimation algorithm which has the privacy protection function, guarantees certain estimation accuracy and can be used for the directed switching topology sensor network under the limited communication condition is realized.
Example 3
the target state prediction module is used for predicting the target state of each sensor node by adopting a volume information filtering algorithm to obtain target state prediction information;
the target state prediction module specifically includes: the initialization submodule is used for initializing and acquiring a target state information vector and a predictive value covariance matrix of each sensor node at the initial moment; the target state prediction information prediction submodule is used for predicting and acquiring target state prediction information of each sensor node at the k moment according to the target state prediction information of each sensor node at the k-1 moment; and the target state prediction information correction submodule is used for correcting the target state prediction information of each sensor node at the k moment by using the target observation value of each sensor node at the k moment.
The target state prediction information prediction sub-module specifically includes:
and the volume sampling point calculation unit is used for calculating the volume sampling points at the k-1 moment as follows according to the target state prediction information at the k-1 moment:
wherein, the first and the second end of the pipe are connected with each other,
denotes the calculated r-th volume sample point at the time k-1 of the i-th sensor node, 2n denotes the number of volume sample points,
representing a covariance matrix of the ith sensor node at time k-1; c. C r A matrix representing dimensions n x 2n,
e s a column vector representing the n dimensions is represented,
s represents c r The number of columns of (a) is,
representing the target state predicted value of the corrected ith sensor node at the moment of k-1;
the target state predicted value prediction unit is used for predicting and obtaining the target state predicted value at the k moment according to the volume sampling point at the k-1 moment as follows:
wherein the content of the first and second substances,
an r volume sampling point representing the ith sensor node at the predicted k time, f (-) represents the nonlinear state transfer function of the target,
indicating the target state prediction value at the moment k obtained by prediction,
covariance matrix, Q, representing the predicted k-th sensor node time k-1 A covariance matrix representing the noise of the target motion process at the k-1 moment;
a target state prediction information obtaining unit, configured to obtain, according to the target state prediction value at the time k of each sensor node, target state prediction information at the time k of each sensor node:
wherein the content of the first and second substances,
a target state information matrix representing the ith sensor node at time k,
and representing the target state information vector of the ith sensor node at the time k.
The target state prediction information modification sub-module specifically includes:
the target state measurement value calculation unit is used for calculating a target state measurement value of each volume sampling point of the ith sensor node through a nonlinear observation function;
the target state observation value calculation unit is used for respectively serving as a target state observation value of the ith sensor node according to the average value of the target state measurement values corresponding to each volume sampling point of the ith sensor node;
the cross covariance matrix calculation unit is used for calculating a cross covariance matrix of the predicted value and the measured value of the ith sensor node according to the target state predicted value and the target state measured value of each volume sampling point;
the pseudo measurement matrix determining unit is used for determining a pseudo measurement matrix of the ith sensor node as follows according to the measured value covariance matrix of the ith sensor node and the cross covariance matrix of the predicted value and the measured value:
wherein the content of the first and second substances,
a pseudo measurement matrix representing the ith sensor node,
a cross covariance matrix representing the predicted value and the measured value of the ith sensor node;
an adjustment quantity determining unit, configured to determine, according to the pseudo measurement matrix of the ith sensor node, that the information increment and the associated information matrix of the ith sensor node at the time k are:
wherein the content of the first and second substances,
the information increment of the ith sensor node at the moment k,
a covariance matrix representing the measured noise of the ith sensor node at time k,
white gaussian noise representing an average of 0;
the information correction unit is used for correcting the target state prediction information of the ith sensor node at the time k by using the following formula according to the information increment of the ith sensor node at the time k and the associated information matrix;
wherein the content of the first and second substances,
a target state information vector representing the corrected k-time of the i-th sensor node,
and representing the corrected target state information matrix of the ith sensor node at the k moment.
And the consistency encryption module is used for carrying out consistency encryption on the target state prediction information transmitted among the sensor nodes based on a pushing and average consistency algorithm of state decomposition.
The updating mode of the target state prediction information transmitted between the adjacent sensor nodes in the consistency encryption module is as follows:
wherein l represents the number of consistent iterations of the target state prediction information,
and
respectively representing alpha items of sub information in target state information vectors of the ith sensor node at the k time under the l iteration and the l +1 iteration,
and
respectively representing beta items of sub information in a target state information vector at the k time under the ith iteration and the l +1 iteration of the ith sensor node,
and
respectively representing alpha items of sub information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration,
and
respectively represents beta items of sub information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration,
representing the consistency weight within the ith sensor node,
the consistency weight between the ith sensor node and the jth sensor node is shown, omega is the consistency weight of the sub information exchange in each sensor node,
coupling weight for information exchange between two sub-states in the ith sensor node;
the updating mode of the deduction item corresponding to the target state prediction information transmitted between the adjacent sensor nodes is as follows:
wherein the content of the first and second substances,
and
respectively representing the sum of alpha item sub information in the target state information vector of the ith sensor node at the k time under the l iteration and the l +1 iteration;
and
a sum term of beta item sub information in a target state information vector at the k time under the ith iteration and l +1 iteration of the ith sensor node is represented;
a sum term of alpha item sub information in a target state information vector of a jth sensor node at the k moment under the ith iteration is represented;
and
respectively representing the sum of alpha item sub-information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration;
and
respectively representing the sum of beta item sub-information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration;
a sum item of alpha item sub information in a target state information matrix at the k moment of the jth sensor node under the ith iteration is represented;
a set of all neighboring sensor nodes representing that the ith sensor node is capable of receiving information;
and
is satisfied with the initial value of iteration
And
is satisfied with the initial value of iteration
Wherein 1 is n×1 Representing an n-dimensional all-1 vector, 1 n×n A full 1-dimensional matrix representing n dimensions;
the consistency encryption result of the target state prediction information transmitted between the adjacent sensor nodes is as follows:
wherein the content of the first and second substances,
a consistent encrypted target state information vector representing the ith sensor node,
and representing a consistency encryption target state information matrix of the ith sensor node.
The invention expands the application range of the original privacy protection tracking filter algorithm to the communication network under the directed switching topological structure, can complete the protection of the sensor network communication while ensuring the nonlinear target tracking precision, thereby realizing the safety guarantee of the target tracking result under the limitation of certain communication conditions.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.
Claims (10)
1. A distributed target tracking method suitable for a directed switching topology is characterized by comprising the following steps:
predicting the target state of each sensor node by adopting a volume information filtering algorithm to obtain target state prediction information; the target state prediction information comprises a target state information matrix and a target state information vector which are obtained through prediction;
and carrying out consistency encryption on target state prediction information transmitted among the sensor nodes based on a pushing and average consistency algorithm of state decomposition.
2. The distributed target tracking method applicable to the directed switching topology according to claim 1, wherein the predicting the target state of each sensor node by using a volume information filtering algorithm to obtain target state prediction information specifically comprises:
initializing and acquiring a target state information vector and a predictive value covariance matrix of each sensor node at an initial moment; the target state information vector and the predictive value covariance matrix at the initial moment are used for predicting target state prediction information at the next moment at the initial moment;
predicting and acquiring target state prediction information of each sensor node at the k moment according to the target state prediction information of each sensor node at the k-1 moment;
and correcting the target state prediction information of each sensor node at the k moment by using the target observation value of each sensor node at the k moment.
3. The distributed target tracking method applicable to the directed switching topology according to claim 2, wherein the step of obtaining target state prediction information at the time k of each sensor node according to the target state prediction information at the time k-1 of each sensor node specifically includes:
according to the target state prediction information at the k-1 moment, calculating the volume sampling point at the k-1 moment as follows:
wherein the content of the first and second substances,
denotes the calculated r-th volume sample point at the time k-1 of the i-th sensor node, 2n denotes the number of volume sample points,
representing a covariance matrix of the ith sensor node at the time k-1; c. C r A matrix representing dimensions n x 2n,
e s a column vector representing the n dimensions is represented,
s represents c r The number of columns of (a) is,
representing the target state predicted value of the corrected ith sensor node at the moment of k-1;
predicting and obtaining a target state predicted value at the k moment according to the volume sampling points at the k-1 moment:
wherein the content of the first and second substances,
an r-th volume sample point representing the i-th sensor node at the predicted k-time, f (-) represents the nonlinear state transfer function of the target,
indicating the target state prediction value at the moment k obtained by prediction,
representing the covariance matrix, Q, of the predicted i-th sensor node at time k k-1 A covariance matrix representing the noise of the target motion process at the k-1 moment;
acquiring target state prediction information of each sensor node at the k moment according to the target state prediction value of each sensor node at the k moment, wherein the target state prediction information of each sensor node at the k moment is as follows:
wherein the content of the first and second substances,
a target state information matrix representing the ith sensor node at time k,
and representing the target state information vector of the ith sensor node at the time k.
4. The distributed target tracking method applicable to the directed switching topology according to claim 3, wherein the correcting the target state prediction information at the time k of each sensor node by using the target observation value at the time k of each sensor node specifically includes:
calculating a target state measurement value of each volume sampling point of the ith sensor node through a nonlinear observation function;
respectively taking the average value of the target state measurement values corresponding to each volume sampling point of the ith sensor node as the target state observation value of the ith sensor node;
calculating a measured value covariance matrix of the ith sensor node according to the target state measured value and the target state observed value of each volume sampling point of the ith sensor node;
calculating a cross covariance matrix of the predicted value and the measured value of the ith sensor node according to the target state predicted value and the target state measured value of each volume sampling point;
according to the measured value covariance matrix of the ith sensor node and the cross covariance matrix of the predicted value and the measured value, determining a pseudo measurement matrix of the ith sensor node as follows:
wherein the content of the first and second substances,
a pseudo measurement matrix representing the ith sensor node,
a cross covariance matrix representing the predicted value and the measured value of the ith sensor node;
according to the pseudo measurement matrix of the ith sensor node, determining the information increment and the associated information matrix of the ith sensor node at the time k as follows:
wherein the content of the first and second substances,
the information increment of the ith sensor node at the moment k,
a covariance matrix representing the measured noise of the ith sensor node at time k,
white gaussian noise representing an average of 0;
according to the information increment and the associated information matrix of the ith sensor node at the time k, correcting the target state prediction information of the ith sensor node at the time k by using the following formula;
wherein the content of the first and second substances,
a target state information vector representing the corrected k-time of the i-th sensor node,
and representing the corrected target state information matrix of the ith sensor node at the k moment.
5. The distributed target tracking method applicable to the directed switching topology according to claim 4, wherein the push and average consistency algorithm based on state decomposition performs consistency encryption on target state prediction information transmitted between sensor nodes, and specifically includes:
the updating mode of the target state prediction information transmitted between the adjacent sensor nodes is as follows:
wherein l represents the number of consistent iterations of the target state prediction information,
and
respectively representing alpha items of sub information in target state information vectors of the ith sensor node at the k time under the l iteration and the l +1 iteration,
and
respectively representing beta items of sub information in a target state information vector at the k time under the ith iteration and the l +1 iteration of the ith sensor node,
and
respectively representing alpha items of sub information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration,
and
respectively represents beta items of sub information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration,
representing the consistency weight within the ith sensor node,
the consistency weight between the ith sensor node and the jth sensor node is shown, omega is the consistency weight of the sub information exchange in each sensor node,
for the ith sensor sectionCoupling weight of information exchange between two sub-states inside the point;
the updating mode of the deduction item corresponding to the target state prediction information transmitted between the adjacent sensor nodes is as follows:
wherein the content of the first and second substances,
and
respectively representing the sum of alpha item sub information in the target state information vector of the ith sensor node at the k time under the l iteration and the l +1 iteration;
and
a sum term of beta item sub information in a target state information vector at the k time under the ith iteration and l +1 iteration of the ith sensor node is represented;
a sum term of alpha item sub information in a target state information vector of a jth sensor node at the k moment under the ith iteration is represented;
and
respectively representing the sum of alpha item sub-information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration;
and
respectively representing the sum of beta item sub-information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration;
a sum item of alpha item sub information in a target state information matrix at the k moment of the jth sensor node under the ith iteration is represented;
a set of all neighboring sensor nodes representing that the ith sensor node is capable of receiving information;
and
is satisfied with the initial value of iteration
And
is satisfied with the initial value of iteration
Wherein 1 is n×1 A full 1-vector representing the n-dimensions,1 n×n a full 1-dimensional matrix representing n dimensions;
the consistency encryption result of the target state prediction information transmitted between the adjacent sensor nodes is as follows:
wherein the content of the first and second substances,
a consistent encrypted target state information vector representing the ith sensor node,
and representing a consistency encryption target state information matrix of the ith sensor node.
6. A distributed target tracking system adapted for use in a directed switching topology, the system comprising:
the target state prediction module is used for predicting the target state of each sensor node by adopting a volume information filtering algorithm to obtain target state prediction information;
and the consistency encryption module is used for carrying out consistency encryption on the target state prediction information transmitted among the sensor nodes based on a pushing and average consistency algorithm of state decomposition.
7. The distributed target tracking system applicable to the directed switching topology according to claim 6, wherein the target state prediction module specifically includes:
the initialization submodule is used for initializing and acquiring a target state information vector and a predictive value covariance matrix of each sensor node at the initial moment; the target state information vector and the predictive value covariance matrix at the initial moment are used for predicting target state prediction information at the next moment at the initial moment;
the target state prediction information prediction submodule is used for predicting and acquiring target state prediction information of each sensor node at the k moment according to the target state prediction information of each sensor node at the k-1 moment;
and the target state prediction information correction submodule is used for correcting the target state prediction information of each sensor node at the k moment by using the target observation value of each sensor node at the k moment.
8. The distributed target tracking system applicable to the directed handover topology according to claim 7, wherein the target state prediction information prediction sub-module specifically includes:
and the volume sampling point calculation unit is used for calculating the volume sampling points at the k-1 moment as follows according to the target state prediction information at the k-1 moment:
wherein the content of the first and second substances,
denotes the calculated r-th volume sample point at the time k-1 of the i-th sensor node, 2n denotes the number of volume sample points,
representing a covariance matrix of the ith sensor node at the time k-1; c. C r A matrix representing dimensions n x 2n,
e s a column vector representing the n dimensions is represented,
s represents c r The number of columns of (a) is,
representing the target state predicted value of the corrected ith sensor node at the moment of k-1;
and the target state predicted value prediction unit is used for predicting and acquiring a target state predicted value at the k moment according to the volume sampling points at the k-1 moment as follows:
wherein the content of the first and second substances,
an r-th volume sample point representing the i-th sensor node at the predicted k-time, f (-) represents the nonlinear state transfer function of the target,
indicating the target state prediction value at the moment k obtained by prediction,
representing the covariance matrix, Q, of the predicted i-th sensor node at time k k-1 A covariance matrix representing the noise of the target motion process at the k-1 moment;
a target state prediction information obtaining unit, configured to obtain, according to the target state prediction value at the time k of each sensor node, target state prediction information at the time k of each sensor node:
wherein the content of the first and second substances,
a target state information matrix representing the ith sensor node at time k,
and representing the target state information vector of the ith sensor node at the time k.
9. The distributed target tracking system applicable to the directed switching topology according to claim 8, wherein the target state prediction information modification sub-module specifically includes:
the target state measurement value calculation unit is used for calculating a target state measurement value of each volume sampling point of the ith sensor node through a nonlinear observation function;
the target state observation value calculation unit is used for respectively serving as a target state observation value of the ith sensor node according to the average value of the target state measurement values corresponding to each volume sampling point of the ith sensor node;
the cross covariance matrix calculation unit is used for calculating a cross covariance matrix of the predicted value and the measured value of the ith sensor node according to the target state predicted value and the target state measured value of each volume sampling point;
the pseudo measurement matrix determining unit is used for determining a pseudo measurement matrix of the ith sensor node as follows according to the measured value covariance matrix of the ith sensor node and the cross covariance matrix of the predicted value and the measured value:
wherein the content of the first and second substances,
a pseudo measurement matrix representing the ith sensor node,
a cross covariance matrix representing the predicted value and the measured value of the ith sensor node;
an adjustment quantity determining unit, configured to determine, according to the pseudo measurement matrix of the ith sensor node, that the information increment and the associated information matrix of the ith sensor node at the time k are:
wherein the content of the first and second substances,
the information increment of the ith sensor node at the moment k,
a covariance matrix representing the measured noise of the ith sensor node at time k,
white gaussian noise representing an average of 0;
the information correction unit is used for correcting the target state prediction information of the ith sensor node at the time k by using the following formula according to the information increment of the ith sensor node at the time k and the associated information matrix;
wherein the content of the first and second substances,
a target state information vector representing the corrected k-time of the i-th sensor node,
and representing the corrected target state information matrix of the ith sensor node at the k moment.
10. The distributed target tracking system applicable to the directed switching topology according to claim 9, wherein the consistency encryption module specifically includes:
the updating mode of the target state prediction information transmitted between the adjacent sensor nodes is as follows:
wherein l represents the number of consistent iterations of the target state prediction information,
and
alpha item subintents in target state information vector respectively representing k time of ith sensor node under l iteration and l +1 iterationIn the form of a capsule, the particles,
and
respectively representing beta items of sub information in a target state information vector at the k time under the ith iteration and the l +1 iteration of the ith sensor node,
and
respectively representing alpha items of sub information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration,
and
respectively represents beta items of sub information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration,
representing the consistency weight within the ith sensor node,
the consistency weight between the ith sensor node and the jth sensor node is shown, omega is the consistency weight of the sub information exchange in each sensor node,
coupling weight for information exchange between two sub-states in the ith sensor node;
the updating mode of the deduction item corresponding to the target state prediction information transmitted between the adjacent sensor nodes is as follows:
wherein the content of the first and second substances,
and
respectively representing the sum of alpha item sub information in the target state information vector of the ith sensor node at the k time under the l iteration and the l +1 iteration;
and
a sum term of beta item sub information in a target state information vector at the k time under the ith iteration and l +1 iteration of the ith sensor node is represented;
a summation item of alpha item sub information in a target state information vector of a jth sensor node at the k moment under the ith iteration is represented;
and
the deduction of alpha item sub information in a target state information matrix at the k time of the ith sensor node under the l iteration and the l +1 iteration respectivelyAnd an item;
and
respectively representing the sum of beta item sub-information in a target state information matrix of the ith sensor node at the k time under the l iteration and the l +1 iteration;
a summation item of alpha item sub information in a target state information matrix of a jth sensor node at the k moment under the ith iteration is represented;
a set of all neighboring sensor nodes representing that the ith sensor node is capable of receiving information;
and
is satisfied with the initial value of iteration
And
is satisfied with the initial value of iteration
Wherein 1 is n×1 Representing an n-dimensional all-1 vector, 1 n×n A full 1-dimensional matrix representing n dimensions;
the consistency encryption result of the target state prediction information transmitted between the adjacent sensor nodes is as follows:
wherein the content of the first and second substances,
a consistent encrypted target state information vector representing the ith sensor node,
and representing a consistency encryption target state information matrix of the ith sensor node.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210384402.0A CN114840801B (en) | 2022-04-13 | 2022-04-13 | Distributed target tracking method and system suitable for directional switching topology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210384402.0A CN114840801B (en) | 2022-04-13 | 2022-04-13 | Distributed target tracking method and system suitable for directional switching topology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114840801A true CN114840801A (en) | 2022-08-02 |
| CN114840801B CN114840801B (en) | 2024-05-31 |
Family
ID=82564682
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210384402.0A Active CN114840801B (en) | 2022-04-13 | 2022-04-13 | Distributed target tracking method and system suitable for directional switching topology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114840801B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107743299A (en) * | 2017-09-08 | 2018-02-27 | 天津大学 | Towards the consensus information filtering algorithm of unmanned aerial vehicle onboard mobile sensor network |
| CN110865334A (en) * | 2019-11-26 | 2020-03-06 | 北京航空航天大学 | Multi-sensor target tracking method and system based on noise statistical characteristics |
| CN110958639A (en) * | 2019-01-31 | 2020-04-03 | 北京航空航天大学 | Target state estimation method and system |
| CN111426322A (en) * | 2020-04-09 | 2020-07-17 | 北京航空航天大学 | Adaptive target tracking filtering method and system for simultaneously estimating state and input |
| CN111829505A (en) * | 2020-01-08 | 2020-10-27 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | Multi-sensor track quality extrapolation track fusion method |
-
2022
- 2022-04-13 CN CN202210384402.0A patent/CN114840801B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107743299A (en) * | 2017-09-08 | 2018-02-27 | 天津大学 | Towards the consensus information filtering algorithm of unmanned aerial vehicle onboard mobile sensor network |
| CN110958639A (en) * | 2019-01-31 | 2020-04-03 | 北京航空航天大学 | Target state estimation method and system |
| CN110865334A (en) * | 2019-11-26 | 2020-03-06 | 北京航空航天大学 | Multi-sensor target tracking method and system based on noise statistical characteristics |
| CN111829505A (en) * | 2020-01-08 | 2020-10-27 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | Multi-sensor track quality extrapolation track fusion method |
| CN111426322A (en) * | 2020-04-09 | 2020-07-17 | 北京航空航天大学 | Adaptive target tracking filtering method and system for simultaneously estimating state and input |
Non-Patent Citations (4)
| Title |
|---|
| ZHA JIRONG 等: "Privacy-Preserving Adaptive Consensus based Cubature Kalman Filter for Distributed Sensor Networks", 《2021 CHINA AUTOMATION CONGRESS》, 24 October 2021 (2021-10-24) * |
| 张露 等: "两阶段高维容积信息滤波及其在目标跟踪中的应用", 《电子学报》, 15 February 2019 (2019-02-15) * |
| 王成 等: "容积信息粒子多传感器融合算法", 《 太赫兹科学与电子信息学报》, 25 December 2021 (2021-12-25) * |
| 董晓飞 等: "有向拓扑条件下针对机动目标的分布式协同制导律设计", 《航空学报》, 10 October 2020 (2020-10-10) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114840801B (en) | 2024-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wu et al. | A distributed Kalman filtering algorithm with fast finite-time convergence for sensor networks | |
| Ullah et al. | Extended Kalman filter-based localization algorithm by edge computing in wireless sensor networks | |
| CN108896047B (en) | Distributed sensor network collaborative fusion and sensor position correction method | |
| Li et al. | Recurrent-neural-network-based unscented Kalman filter for estimating and compensating the random drift of MEMS gyroscopes in real time | |
| CN106991691A (en) | A kind of distributed object tracking being applied under camera network | |
| Keshavarz-Mohammadiyan et al. | Consensus-based distributed unscented target tracking in wireless sensor networks with state-dependent noise | |
| CN103065037B (en) | Nonlinear system is based on the method for tracking target of distributing volume information filtering | |
| Xu et al. | A multi-sensor information fusion method based on factor graph for integrated navigation system | |
| Xia et al. | Distributed cubature H-infinity information filtering for target tracking against uncertain noise statistics | |
| CN109151759B (en) | Sensor network distributed information weighted consistency state filtering method | |
| CN116186643B (en) | Multi-sensor collaborative target tracking method, system, equipment and medium | |
| CN112798021A (en) | Inertial navigation system inter-travelling initial alignment method based on laser Doppler velocimeter | |
| Gao et al. | Distributed state fusion using sparse-grid quadrature filter with application to INS/CNS/GNSS integration | |
| CN111008364A (en) | Method and system for cooperative passive positioning of double observers | |
| Yang et al. | Distributed observer‐based coordination for multiple Lagrangian systems using only position measurements | |
| Zhang et al. | Consensus-based distributed two-target tracking over wireless sensor networks | |
| Wanasinghe et al. | Decentralized cooperative localization approach for autonomous multirobot systems | |
| CN109857127B (en) | Method and device for calculating training neural network model and aircraft attitude | |
| Tang et al. | Information weighted consensus‐based distributed particle filter for large‐scale sparse wireless sensor networks | |
| CN114840801A (en) | Distributed target tracking method and system suitable for directed switching topology | |
| Li et al. | Distributed extended object tracking information filter over sensor networks | |
| Liu et al. | Distributed state estimation for dynamic positioning systems with uncertain disturbances and transmission time delays | |
| John et al. | Fixed‐time orientation estimation and network localisation of multi‐agent systems | |
| Xia et al. | An Exploratory Distributed Localization Algorithm Based on 3D Barycentric Coordinates | |
| Ding et al. | Distributed input and state estimation for non‐linear discrete‐time systems with direct feedthrough |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |