CN113048984B - Dynamic positioning information fusion method for underwater unmanned robot cluster - Google Patents
Dynamic positioning information fusion method for underwater unmanned robot cluster Download PDFInfo
- Publication number
- CN113048984B CN113048984B CN202110359019.5A CN202110359019A CN113048984B CN 113048984 B CN113048984 B CN 113048984B CN 202110359019 A CN202110359019 A CN 202110359019A CN 113048984 B CN113048984 B CN 113048984B
- Authority
- CN
- China
- Prior art keywords
- value
- local
- posterior
- average
- moment
- 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
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/20—Instruments for performing navigational calculations
- G01C21/203—Specially adapted for sailing ships
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention relates to the technical field of unmanned robot clusters, in particular to a dynamic positioning information fusion method for an underwater unmanned robot cluster. By introducing the concept of a consistency method under dynamic topology and combining a distributed unscented Kalman filtering method, a local unscented filter and a weighted average consistency filter are embedded in each node of a communication network, and the local unscented filter obtains local unscented estimation quantity about the positioning state of a tested node by utilizing local multisource observation information of each AUV node; and simultaneously, the weighted average consistency filter carries out consistent fusion on all local posterior estimation, so that the global posterior estimation result is output in a mode of weighting the posterior estimation mean value by an information matrix.
Description
Technical Field
The invention relates to the technical field of unmanned robot clusters, in particular to a dynamic positioning information fusion method for an underwater unmanned robot cluster.
Background
With the increasing maturity of AUV technology, the task difficulty and complexity of AUV are also increased, and single AUV cannot meet the new demands of increasing accuracy, diversity and complexity of tasks due to the limitations of AUV, and AUV is developing toward miniaturization, structural simplification, intellectualization, mixing and swarmization. In particular, multiple AUVs have advantages not available with some single AUVs, such as higher fault tolerance, robustness, efficient operation, reconfigurability, distributed sensing and coordination, wider application fields, and the like.
However, compared with the traditional land and air radio communication network-based co-location method, the multi-AUV cluster co-location based on the acoustic communication network has completely different technical characteristics and difficulties. The single AUV only has local sensing and communication capability, and the underwater acoustic communication has the problems of narrow channel bandwidth, low data transmission rate, long measurement time, poor real-time performance and the like, and the method of centralized decision and global decision is not suitable for the needs under the submarine environment. The deep sea environment is complex and changeable, so that the communication of AUV cluster formation presents sparse and dynamic change characteristics, and the message transmission method in the distributed co-location method depends on tree-shaped or ring-shaped communication topology, so that the method is hardly applicable to AUV cluster formation of dynamic topology. The message diffusion form represented by the channel filter only needs to carry out single-hop communication between adjacent nodes, has strong universality on communication topology in theory, and gradually becomes the main research direction of the distributed method. The LED light and low-frequency radio waves are used as dual communication media, detection equipment is changed from high-power-consumption sonar to low-power-consumption LED light strips and underwater cameras, power consumption is reduced, real-time monitoring of the experimental platform on shore is realized, experimental phenomena are convenient to observe, but the system energy consumption, calculated amount, communication amount and the like are greatly increased in system control and co-location scheme design in the face of severe deep sea environment, so that under the severe constraint condition, optimization of system energy consumption, calculated amount and communication amount is completed, and the method has very important significance for rapid aggregation of multiple AUV cluster systems and even adjustment of formation structures.
Disclosure of Invention
In order to solve the technical problems, the invention provides a dynamic positioning information fusion method for an underwater unmanned robot cluster, which is characterized in that a local unscented filter and a weighted average consistency filter are embedded in each node of a communication network by introducing the concept of a consistency method under dynamic topology and combining a distributed unscented Kalman filtering method, and the local unscented filter obtains local unscented estimation quantity about the positioning state of a tested node by utilizing local multisource observation information of each AUV node; and simultaneously, the weighted average consistency filter carries out consistent fusion on all local posterior estimation, so that the global posterior estimation result is output in a mode of weighting the posterior estimation mean value by an information matrix.
In order to achieve the above purpose, the present invention provides the following technical solutions:
the dynamic positioning information fusion method for the underwater unmanned robot cluster comprises the following steps of:
step one: constructing a state space model of a multi-AUV co-location system and a communication network model based on an average network undirected graph, comprehensively considering the dynamics characteristics of a platform, determining input parameters, adding Gaussian noise, and establishing a motion equation;
step two: threshold weighting eliminates coarse differences in local information: a threshold weighting module is added to the rear end of the local observation value to optimize special node information;
step three: local information filter: approximating probability density of the function by using unscented transformation, solving expected and variance of a target event, converting a nonlinear problem into a Kalman filtering problem, and outputting a local posterior value;
step four: weighted average consistency filter: the weighted average consistency filter carries out consistent fusion on all local posterior estimation, so that a global posterior estimation result is output in a mode of weighting a posterior estimation mean value by an information matrix;
step five: optimizing global posterior results: when the consistency filter outputs the global posterior estimated value and covariance, judging whether the consistency filter corresponds to the latest moment and optimizing the latest moment, and generating a stable estimated value as an input value of the local filter so as to solve the problem of filtering asynchronism.
The invention is further improved, in the first step, the deep sea sparse dynamic wireless communication network is described based on an average network undirected graph model, the sparsity is described by an effective communication connection edge set between any nodes, and a state space model of any nodes is as follows:
the equation of state: x is x k ＝f(x k-1 )+Q k-1
Observation equation:
wherein x is k As a real state variable of the AUV under test,for observing the variable, the state equation and the observation equation are nonlinear functions, Q k-1 To predict noise, R k For observing noise, it is assumed that both are gaussian white noise, satisfying normal distribution;
random dynamic variation of sparse networks by monitoring the change of the correlation probability values in the average adjacency matrix and the average laplace matrix, using undirected graph G k ＝(V，ε k ) Describing a sparse dynamic network communication topology at any time k, wherein V is a set of sensor nodes, and the cardinality of the set of sensor nodes is N, epsilon k For the set of effective communication connection edges between nodes at the current moment, the base number of the set of effective communication connection edges is M k M in sparse dynamic network k The sparse condition is satisfied:
the adjacent matrix at the previous k moment is synthesized to obtain an average adjacent matrix, and the average adjacent matrix is differenced with the average degree matrix to obtain an average Laplace matrix (L=D-A), wherein the change value of elements in the matrix can clearly reflect the random dynamic change condition of a sparse network, the average Laplace matrix is a semi-positive definite matrix, and the minimum characteristic value of the average Laplace matrix is larger than zero and is a filling condition for communicating an average network undirected graph;
the convergence of the consistency method is described through the angle of mean square convergence, and in the average network undirected graph, the communication topology is randomly and dynamically changed with a certain probability, so that the method is more suitable for examining the convergence of the consistency method from the angle of mean square convergence, and if the average network undirected graph is communicated, the average consistency method can enable all node states to be converged to an average consistency value.
The invention is further improved, a threshold weighting module is added at the rear end of the local observed value in the second step to optimize special node information, the condition that the signal to be detected corresponds to the moment k-2 is assumed, and the state quantity such as the position, the speed, the heading angle and the like of the tracked node, the sampling period and the maximum error under the environmental interference such as ocean current and the like are based on the moment k-2Calculating the existence range of the estimated expected value of the tracking node at the moment k-1, and setting a threshold value +.>Calculating the existence range of the estimated expected value of the tracking node at the current moment, setting a threshold value according to the speed of the previous moment and the signal sampling period, and measuring the distance l between the target and the body at the previous moment and the next moment k-1 L k-2 Setting weight +.>Wherein Δl= |l k-1 -l k-2 If the observed signal exceeds the threshold value, a virtual position setpoint value is generated +.>Instead of the observed values, the method participates in local filtering calculation, and avoids the risk of excessive distortion of remote signals.
The invention is further improved, and the specific steps of the local information filter in the step three are as follows:
knowing the AUV node state x detected at the current time k-1 Is estimated (mean value of posterior estimation for the previous moment) and covarianceDecomposing covariance matrix->Wherein the requirement->Must be positive;
1) In order to transfer the state transfer function f (x k-1 ) Is approximately normal distribution, and is subjected to UT conversion to obtain sigma points
L (i) Represents the ith column of matrix L, and the weight is
2)
Then
Obtaining the prior value of the detected node stateIs satisfying the normal distribution +.>
5) In the same way, the first step is toDecomposing, let->
6)
L 1(i ) Representative matrix L 1 Is the (i) th column of the weight value
5)y k (i) ＝f(x k-1 (i) ) Then
Priori expectation of observed value of measured node at k moment
Its a priori variance
6) Observe the observed value y of the AUV node to be tested m
7)
Note k=p xy (P y ) -1
8) State posterior expectation of AUV node under test at k moment
Posterior variance of
So far, the state posterior estimation value and covariance of the k moment are already obtained, and the next cycle can be entered to continue recursion.
The invention further improves, in step four, the consistency method of multi-agent clusters is used, and all local posterior estimation is subjected to consistency fusion through a weighted average consistency filter, so that the global posterior estimation result is output in the form of weighting the posterior estimation mean value by an information matrix:
considering a network of n AUVs, the AUV monomers estimate a signal that is disturbed by gaussian noise from the measurements, the signal model is slightly modified in form as follows:
x(k+1)＝x(k)+w(k)
each AUV measures the signal as
z i (k)＝H i (k)x(k)+v i (k)
Wherein the process noise and the measured value are respectively
E[w(k)w(l) T ]＝Q(k)δ kl ，
E[v i (k)v j (l) T ]＝R i (k)δ kl δ ij
Wherein when k=l, δ kl =1; otherwise, delta kl ＝0。
Consider a network consisting of n AUVs, assuming n (a, H) are observable; assume that each AUV applies the following distributed estimation method
And initial condition P i ＝P 0 ，Then, estimate error->Is a stable linear system with Lyapunov function of +.>
The invention is further improved, when the consistency filter outputs the global posterior estimation value and covariance, judging the timeliness of the consistency filter, namely, whether the consistency filter corresponds to the current moment, if so, the consistency filter is used as the input of the local filter; if not, entering an estimation module, generating a reliable estimated value as a posterior value of the local filter, and generating the reliable estimated value as the posterior value of the local filter according to the following steps:
1) Calculating the change delta of the global estimated value of the AUV state to be measured at the moment k-1;
2) Obtaining a state assumption value of the AUV to be measured at the moment k under the variation delta;
3) Setting an estimated value of the AUV state as a weighted average of the assumed value and the local posterior value;
4) The predicted value is entered into the local method as a global posterior estimate and covariance.
The invention has the beneficial effects that:
(1) The invention combines the knowledge of graph theory and topology network and unscented Kalman filtering and consistency method, and is fully applicable to the situations of sparse communication, random dynamic change of communication topology and clutter interference of multi-AUV co-location in a deepwater environment.
(2) The AUV observation value threshold weighting method reduces measurement deviation caused by harsh underwater signal conditions and dynamic and static obstacles, eliminates observation rough differences according to the state of the AUV and environmental factors, and improves the reliability of positioning information.
(3) According to the invention, a global posterior value estimation mechanism is introduced, whether the local filtering at the next moment is carried out is selected by judging the aging of the consistency filtering, and the delayed global posterior result is replaced by the estimated value, so that the aging of the consistency method is ensured, and the filtering asynchronism problem of the local filtering and the consistency filtering is effectively avoided.
Drawings
FIG. 1 is a partial information filtering method of the present invention;
FIG. 2 is a threshold weighting module in a local filter of the present invention;
FIG. 3 is a parallel fusion consistency distributed filtering method of the present invention;
FIG. 4 is a schematic diagram of a predictive module for handling filtering asynchrony in accordance with the present invention.
Detailed Description
The present invention will be described in further detail with reference to the following examples, which are only for the purpose of illustrating the invention and are not to be construed as limiting the scope of the invention.
Examples: as shown in fig. 1 to 4, a dynamic positioning information fusion method for an underwater unmanned robot cluster includes the steps of:
step one: the method comprises the steps of constructing a state space model of a multi-AUV co-location system, comprehensively considering the dynamics characteristics of a platform, determining input parameter values, adding Gaussian noise, and establishing a motion equation. Constructing a communication network model based on an average network undirected graph, describing sparseness through an effective communication connection edge set among any nodes, describing random dynamic changes of the sparse network by monitoring changes of relevant probability values in an average adjacency matrix and an average Laplace matrix, and describing convergence of a consistency method through a mean square convergence angle.
Step two: the rough difference of the local information is eliminated by using threshold weighting: and a threshold weighting module is added at the rear end of the local observation value to perform rough difference elimination, and special node information is optimized.
Step three: improving unscented Kalman filtering to process local information: the probability density of the function is approximated by a Unscented Transform (UT), the expected and variance of the target event are calculated, and the nonlinear problem is converted into a kalman filter problem.
Step four: weighted average consistency filter: and the weighted average consistency filter carries out consistent fusion on all the local posterior estimation, so that the global posterior estimation result is output in the form of weighting the posterior estimation mean value by an information matrix.
Step five: optimizing global posterior results: when the consistency filter outputs the global posterior estimated value and covariance, judging whether the global posterior estimated value and covariance correspond to the latest moment, if the corresponding moment of the value is correct, the value is used as the input of the local filter, and if the value does not correspond to the current moment, the value enters the estimation module to generate a stable estimated value as the input value of the local filter so as to solve the problem of filtering asynchronism.
In this embodiment, the first step specifically includes: the platform state of the single AUV consists of elements such as the position, the speed, the gesture and the like of the platform. Comprehensively considering the dynamics characteristics of the platform, determining input parameters, adding Gaussian noise, and establishing a motion equation. The state equation of the whole co-location system can be obtained from the state, input and noise of the whole platform. Similarly, when the observation equation between the single platform and the platform is properly expanded, the observation equation of the whole system can be obtained. The state space model is a starting point of the design of the co-location method, and in the sparse dynamic wireless sensor network, the state space model of any node i is as follows:
the equation of state: x is x k ＝f(x k-1 )+Q k-1
Observation equation:
wherein x is k The real state variable of the AUV to be measured contains information such as position, speed, gesture and the like,for observing variables, the state equation and the observation equation are nonlinear functions in consideration of nonlinearity of the track push and the relative distance measurement model, and the actual observation value of the AUV to be measured. Q (Q) k-1 To predict noise, R k To observe noise, it is assumed that both are gaussian white noise, satisfying a normal distribution.
First, using undirected graph G k ＝(V，ε k ) Describing a sparse dynamic network communication topology at any time k, wherein V is a set of sensor nodes, and the cardinality of the set of sensor nodes is N, epsilon k Connecting a set of edges with a base ofNumber of active communication connection sides M k . M in sparse dynamic network k The sparse condition is satisfied:
the average adjacent matrix can be obtained by integrating the adjacent matrix at the previous k time, and the average Laplace matrix (L=D-A) can be obtained by making a difference with the average degree matrix, and the change value of the elements in the matrix can clearly reflect the random dynamic change condition of the sparse network. And the average Laplace matrix is a semi-positive definite matrix, and the minimum eigenvalue of the matrix is larger than zero, so that the average network undirected graph is a connected condition.
In the average network undirected graph, the communication topology changes randomly and dynamically with a certain probability, so that the method is more suitable for investigating the convergence of the consistency method from the aspect of mean square convergence. If the average network undirected graph is connected, the average consistency method can enable all node states to converge to an average consistency value in a mean square.
In this embodiment, the second step specifically includes: assuming that the signal to be detected corresponds to time k-2, and according to the state quantity such as the position, the speed, the heading angle and the like of the tracked node at time k-2, the sampling period and the maximum error under the environmental interference such as ocean current and the likeCalculating the existence range of the estimated expected value of the tracking node at the moment k-1, and setting a threshold value +.>According to the distance l between the measuring object and the body at the front and back time k-1 L k-2 Setting weight +.>Wherein Δl= |l k-1 -l k-2 | a. The invention relates to a method for producing a fibre-reinforced plastic composite. If the observed signal exceeds the threshold value, a virtual position expectation value is generated>Instead of looking atAnd measuring values, participating in local filtering calculation, and avoiding the risk of excessive distortion of remote signals.
In this embodiment, the third step specifically includes: knowing the AUV node state x detected at the current time k-1 Is estimated (mean value of posterior estimation for the previous moment) and covarianceDecomposing covariance matrix->Wherein the requirement->Must be positive;
1) In order to transfer the state transfer function f (x k-1 ) Is approximately normal distribution, and is subjected to UT conversion to obtain sigma points
L (i) Represents the ith column of matrix L, and the weight is
2)
Then
Obtaining the prior value of the detected node stateIs satisfying the normal distribution +.>
3) In the same way, the first step is toDecomposing, let->
4)
L 1(i) Representative matrix L 1 Is the (i) th column of the weight value
5)y k (i) ＝f(x k-1 (i) ) Then
Priori expectation of observed value of measured node at k moment
Its a priori variance
6) Observe the observed value y of the AUV node to be tested m
7)
Note k=p xy (P y ) -1
8) State posterior expectation of AUV node under test at k moment
Posterior variance of
So far, the state posterior estimation value and covariance of the k moment are already obtained, and the next cycle can be entered to continue recursion.
In this embodiment, the fourth step specifically includes: considering a network of n AUVs, the AUV monomers estimate a signal that is disturbed by gaussian noise from the measurements, the signal model is slightly modified in form as follows:
x(k+1)＝x(k)+w(k)
each AUV measures the signal as
z i (k)＝H i (k)x(k)+v i (k)
Wherein the process noise and the measured value are respectively
E[w(k)w(l) T ]＝Q(k)δ kl ，
E[v i (k)v j (l) T ]＝R i (k)δ kl δ ij
Wherein when k=l, δ kl =1; otherwise, delta kl ＝0。
Consider a network consisting of n AUVs, assuming n (a, H) are observable. Assume that each AUV applies the following distributed estimation method
And initial condition P i ＝P 0 ，Then, estimate error->Is a stable linear system with Lyapunov function of +.>
In this embodiment, the fifth step specifically includes: in the case of a significant improvement in hardware level, the computation cycle is also greatly shortened, and the latest global posterior result should be fully utilized by the system. Outputting global posterior estimation value by consistency filterCovariance->Judging whether the value corresponds to the k moment or not, and if the value corresponds to the moment correctly, taking the value as the input of the local filter; if the current time is not corresponding, entering an estimation module, and taking a reliable estimated value as a posterior value of the local filter according to the generation of the reliable estimated value, wherein the method specifically comprises the following steps:
1) Calculating the change delta of the global estimated value of the AUV state to be measured at the moment k-1;
2) Obtaining a state assumption value of the AUV to be measured at the moment k under the variation delta;
3) Setting an estimated value of the AUV state as a weighted average of the assumed value and the local posterior value;
4) The predicted value is entered into the local method as a global posterior estimate and covariance.
The foregoing is merely exemplary embodiments of the present invention, and specific structures and features that are well known in the art are not described in detail herein. It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Claims (6)
1. The dynamic positioning information fusion method for the underwater unmanned robot cluster is characterized by comprising the following steps of:
step one: constructing a state space model of a multi-AUV co-location system and a communication network model based on an average network undirected graph, comprehensively considering the dynamics characteristics of a platform, determining input parameters, adding Gaussian noise, and establishing a motion equation;
step two: threshold weighting eliminates coarse differences in local information: a threshold weighting module is added to the rear end of the local observation value to optimize special node information;
step three: local information filter: approximating probability density of the function by using unscented transformation, solving expected and variance of a target event, converting a nonlinear problem into a Kalman filtering problem, and outputting a local posterior value;
step four: weighted average consistency filter: the weighted average consistency filter carries out consistent fusion on all local posterior estimation, so that a global posterior estimation result is output in a mode of weighting a posterior estimation mean value by an information matrix;
step five: optimizing global posterior results: when the consistency filter outputs the global posterior estimated value and covariance, judging whether the consistency filter corresponds to the latest moment and optimizing the latest moment, and generating a stable estimated value as an input value of the local filter so as to solve the problem of filtering asynchronism.
2. The method for dynamic positioning information fusion of an underwater unmanned robot cluster according to claim 1, wherein in the first step, a deep sea sparse dynamic wireless communication network is described based on a communication network model of an average network undirected graph, sparseness is described by an effective communication connection edge set between any nodes, and a state space model of any node is as follows:
the equation of state: x is x k ＝f(x k-1 )+Q k-1
Observation equation:
wherein x is k As a real state variable of the AUV under test,for observing the variable, the state equation and the observation equation are nonlinear functions, Q k-1 To predict noise, R k For observing noise, it is assumed that both are gaussian white noise, satisfying normal distribution;
random dynamic variation of sparse networks by monitoring the change of the correlation probability values in the average adjacency matrix and the average laplace matrix, using undirected graph G k ＝(V,ε k ) Describing a sparse dynamic network communication topology at any time k, wherein V is a set of sensor nodes, and the cardinality of the set of sensor nodes is N, epsilon k For the set of effective communication connection edges between nodes at the current moment, the base number of the set of effective communication connection edges is M k M in sparse dynamic network k Satisfy sparse barPiece (2):
the adjacent matrix at the previous k moment is synthesized to obtain an average adjacent matrix, and the average adjacent matrix is differenced with the average degree matrix to obtain an average Laplace matrix (L=D-A), wherein the change value of elements in the matrix can clearly reflect the random dynamic change condition of a sparse network, the average Laplace matrix is a semi-positive definite matrix, and the minimum characteristic value of the average Laplace matrix is larger than zero and is a filling condition for communicating an average network undirected graph;
the convergence of the consistency method is described through the angle of mean square convergence, and in the average network undirected graph, the communication topology is randomly and dynamically changed with a certain probability, so that the method is more suitable for examining the convergence of the consistency method from the angle of mean square convergence, and if the average network undirected graph is communicated, the average consistency method can enable all node states to be converged to an average consistency value.
3. The method for dynamic positioning information fusion of an underwater unmanned robot cluster according to claim 1, wherein a threshold weighting module is added to the rear end of the local observation value in the second step to optimize special node information, and the state quantities and sampling periods of the tracked nodes at the time k-2 are based on the position, speed, heading angle and the like of the tracked nodes and the maximum error under the environmental interference of ocean currents and the like under the assumption that the signal to be detected corresponds to the time k-2Calculating the existence range of the estimated expected value of the tracking node at the moment k-1, and setting a threshold value +.>Calculating the existence range of the estimated expected value of the tracking node at the current moment, setting a threshold value according to the speed of the previous moment and the signal sampling period, and measuring the distance l between the target and the body at the previous moment and the next moment k-1 L k-2 Setting weight +.>Wherein Δl= |l k-1 -l k-2 If the observed signal exceeds the threshold value, a virtual position setpoint value is generated +.>Instead of the observed values, the method participates in local filtering calculation, and avoids the risk of excessive distortion of remote signals.
4. The method for dynamic positioning information fusion of an underwater unmanned robot cluster according to claim 1, wherein the specific steps of the local information filter in the third step are as follows:
knowing the AUV node state x detected at the current time k-1 Is estimated (mean value of posterior estimation for the previous moment) and covarianceDecomposing covariance matrix->Wherein the requirement->Must be positive;
1) In order to transfer the state transfer function f (x k-1 ) Is approximately normal distribution, and is subjected to UT conversion to obtain sigma points
L (i) Represents the ith column of matrix L, and the weight is
2)
Then
Obtaining the prior value of the detected node stateIs satisfying the normal distribution +.>
3) In the same way, the first step is toDecomposing, let->
4)
L 1(i) Representative matrix L 1 Is the (i) th column of the weight value
5)y k (i) ＝f(x k-1 (i) ) Then
Priori expectation of observed value of measured node at k moment
Its a priori variance
6) Observe the observed value y of the AUV node to be tested m
7)
Note k=p xy (P y ) -1
8) State posterior expectation of AUV node under test at k moment
Posterior variance of
So far, the state posterior estimation value and covariance of the k moment are already obtained, and the next cycle can be entered to continue recursion.
5. The method for dynamic positioning information fusion of underwater unmanned robot clusters according to claim 1, wherein in the fourth step, a multi-agent cluster consistency method is used, and all local posterior estimates are consistently fused through a weighted average consistency filter, so that a global posterior estimation result is output in a form of weighting a posterior estimation mean value by an information matrix:
considering a network of n AUVs, the AUV monomers estimate a signal that is disturbed by gaussian noise from the measurements, the signal model is slightly modified in form as follows:
x(k+1)＝x(k)+w(k)
each AUV measures the signal as
z i (k)＝H i (k)x(k)+v i (k)
Wherein the process noise and the measured value are respectively
E[w(k)w(l) T ]＝Q(k)δ kl ，
E[v i (k)v j (l)T]＝R i (k)δ kl δ ij
Wherein when k=l, δ kl =1; otherwise, delta kl ＝0；
Consider a network consisting of n AUVs, assuming n (a, H) are observable; assume that each AUV applies the following distributed estimation method
And initial condition P i ＝P 0 ，Then, estimate error->Is a stable linear system with Lyapunov function of +.>
6. The method for dynamic positioning information fusion of an underwater unmanned robot cluster according to claim 1, wherein the fifth step is to judge the timeliness of the global posterior estimate and covariance when the consistency filter outputs the global posterior estimate and covariance, namely, whether the global posterior estimate and covariance correspond to the current moment, and if so, the global posterior estimate and covariance are used as the input of the local filter; if not, entering an estimation module, generating a reliable estimated value as a posterior value of the local filter, and generating the reliable estimated value as the posterior value of the local filter according to the following steps:
1) Calculating the change delta of the global estimated value of the AUV state to be measured at the moment k-1;
2) Obtaining a state assumption value of the AUV to be measured at the moment k under the variation delta;
3) Setting an estimated value of the AUV state as a weighted average of the assumed value and the local posterior value;
4) The predicted value is entered into the local method as a global posterior estimate and covariance.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110359019.5A CN113048984B (en) | 2021-04-01 | 2021-04-01 | Dynamic positioning information fusion method for underwater unmanned robot cluster |
| PCT/CN2021/088258 WO2022205526A1 (en) | 2021-04-01 | 2021-04-20 | Dynamic positioning information fusion method for unmanned underwater vehicle cluster |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110359019.5A CN113048984B (en) | 2021-04-01 | 2021-04-01 | Dynamic positioning information fusion method for underwater unmanned robot cluster |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113048984A CN113048984A (en) | 2021-06-29 |
| CN113048984B true CN113048984B (en) | 2023-10-03 |
Family
ID=76517514
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110359019.5A Active CN113048984B (en) | 2021-04-01 | 2021-04-01 | Dynamic positioning information fusion method for underwater unmanned robot cluster |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN113048984B (en) |
| WO (1) | WO2022205526A1 (en) |
Families Citing this family (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113673076B (en) * | 2021-07-09 | 2022-09-02 | 华南理工大学 | Dynamic response solving method suitable for ocean floating structure |
| CN114111796B (en) * | 2021-11-26 | 2023-10-24 | 江苏科技大学 | Parallel fusion positioning method and system of underwater unmanned robot based on information gain |
| CN115797749B (en) * | 2023-02-02 | 2023-05-23 | 吉林大学 | Method and system for detecting state of autonomous underwater vehicle |
| CN115792796B (en) * | 2023-02-13 | 2023-06-06 | 鹏城实验室 | Co-location method, device and terminal based on relative observation equivalent model |
| CN116359642B (en) * | 2023-03-10 | 2024-01-23 | 湖南金烽信息科技有限公司 | Transformer running state 5G intelligent monitoring system and method |
| CN116740922B (en) * | 2023-05-08 | 2024-04-05 | 海南大学 | Control method of intelligent traffic system based on fuzzy observation protocol |
| CN116358564B (en) * | 2023-06-01 | 2023-07-28 | 中国人民解放军战略支援部队航天工程大学 | Unmanned aerial vehicle bee colony centroid motion state tracking method, system, equipment and medium |
| CN117278941B (en) * | 2023-09-15 | 2024-06-25 | 广东省机场管理集团有限公司工程建设指挥部 | Vehicle driving auxiliary positioning method and device based on 5G network and data fusion |
| CN117269885B (en) * | 2023-11-23 | 2024-02-20 | 中国飞行试验研究院 | Aircraft positioning method and device based on opportunistic signal fusion |
| CN117290115B (en) * | 2023-11-27 | 2024-07-16 | 北京航空航天大学 | Calculation method and system of unmanned aerial vehicle cluster distributed target distribution model |
| CN117540342B (en) * | 2024-01-05 | 2024-03-26 | 山东省科学院海洋仪器仪表研究所 | Short-term prediction method and system for fusion of evaporation waveguide |
| CN117686920B (en) * | 2024-02-02 | 2024-04-12 | 河南科技学院 | Terminal voltage prediction method for single battery in battery system |
| CN117824625B (en) * | 2024-03-05 | 2024-05-14 | 河海大学 | High dam large warehouse underwater environment sensing and composition method based on improved visual odometer |
| CN117990112B (en) * | 2024-04-03 | 2024-07-02 | 中国人民解放军海军工程大学 | Unmanned aerial vehicle photoelectric platform target positioning method based on robust unscented Kalman filtering |
| CN118030040B (en) * | 2024-04-11 | 2024-07-30 | 克拉玛依市富城油气研究院有限公司 | Production dynamic monitoring system and method for oil extraction engineering |
| CN118070556B (en) * | 2024-04-17 | 2024-08-02 | 南京理工大学 | Design method of probability-preserving set member estimator for complex network |
| CN118170034B (en) * | 2024-05-16 | 2024-07-30 | 山东科技大学 | Method for controlling mean square consistency of heterogeneous multi-agent system with communication packet loss |
| CN118331033B (en) * | 2024-06-14 | 2024-09-06 | 东北大学 | Virtual power plant system consistency control method and device based on multiple intelligent agents |
| CN118427716B (en) * | 2024-07-04 | 2024-08-30 | 山东环球软件股份有限公司 | Intelligent park monitoring method based on digital twin technology |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103335646A (en) * | 2013-06-20 | 2013-10-02 | 哈尔滨工程大学 | Multi-boat cooperated navigation method based on distributed augmented information filtering |
| CN108871324A (en) * | 2018-04-28 | 2018-11-23 | 南京信息工程大学 | A kind of underwater passive integrated navigation system decaying adaptive information fusion method |
| CN111397607A (en) * | 2020-03-19 | 2020-07-10 | 哈尔滨工程大学 | Information filtering method adopting parallel fusion mechanism |
| CN112527017A (en) * | 2020-12-11 | 2021-03-19 | 中国科学院沈阳自动化研究所 | Ocean observation method based on multiple AUVs |
| CN112583380A (en) * | 2020-12-15 | 2021-03-30 | 哈尔滨工程大学 | Distributed multi-rate particle filtering algorithm based on convergence optimization |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9963215B2 (en) * | 2014-12-15 | 2018-05-08 | Leidos, Inc. | System and method for fusion of sensor data to support autonomous maritime vessels |
| CN110702095B (en) * | 2019-09-30 | 2022-09-16 | 江苏大学 | Data-driven high-precision integrated navigation data fusion method |
| CN112378396A (en) * | 2020-10-29 | 2021-02-19 | 江苏集萃未来城市应用技术研究所有限公司 | Hybrid high-precision indoor positioning method based on robust LM visual inertial odometer and UWB |
-
2021
- 2021-04-01 CN CN202110359019.5A patent/CN113048984B/en active Active
- 2021-04-20 WO PCT/CN2021/088258 patent/WO2022205526A1/en active Application Filing
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103335646A (en) * | 2013-06-20 | 2013-10-02 | 哈尔滨工程大学 | Multi-boat cooperated navigation method based on distributed augmented information filtering |
| CN108871324A (en) * | 2018-04-28 | 2018-11-23 | 南京信息工程大学 | A kind of underwater passive integrated navigation system decaying adaptive information fusion method |
| CN111397607A (en) * | 2020-03-19 | 2020-07-10 | 哈尔滨工程大学 | Information filtering method adopting parallel fusion mechanism |
| CN112527017A (en) * | 2020-12-11 | 2021-03-19 | 中国科学院沈阳自动化研究所 | Ocean observation method based on multiple AUVs |
| CN112583380A (en) * | 2020-12-15 | 2021-03-30 | 哈尔滨工程大学 | Distributed multi-rate particle filtering algorithm based on convergence optimization |
Non-Patent Citations (2)
| Title |
|---|
| ZHANG L C et al..Cooperative localization for multi-AUVs based on GMPHD filters and information entropy theory.Sensors.2017,全文. * |
| 闫敬 ; 张立 ; 罗小元 ; 濮彬 ; 关新平 ; .异步时钟下基于信息物理融合的水下潜器协同定位算法.自动化学报.2018,(04),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113048984A (en) | 2021-06-29 |
| WO2022205526A1 (en) | 2022-10-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113048984B (en) | Dynamic positioning information fusion method for underwater unmanned robot cluster | |
| Zou et al. | Moving horizon estimation meets multi-sensor information fusion: Development, opportunities and challenges | |
| WO2023045357A1 (en) | Integrated navigation robust filtering method based on statistical similarity measurement | |
| CN112284384B (en) | Co-positioning method of clustered multi-deep sea submarine considering measurement abnormality | |
| CN111780755A (en) | Multisource fusion navigation method based on factor graph and observability degree analysis | |
| Wan et al. | Autonomous vehicle source enumeration exploiting non-cooperative UAV in software defined internet of vehicles | |
| Li et al. | Distributed iterative FIR consensus filter for multiagent systems over sensor networks | |
| Xu et al. | Cooperative localization in harsh underwater environment based on the MC-ANFIS | |
| CN116599602B (en) | Low-energy-consumption long-distance sonar communication system for ocean monitoring | |
| Wen et al. | Intelligent multi-AUG ocean data collection scheme in maritime wireless communication network | |
| Zhang et al. | Cooperative positioning of underwater unmanned vehicle clusters based on factor graphs | |
| Zhang et al. | Gaussian parameterized information aided distributed cooperative underwater positioning algorithm | |
| CN111928851B (en) | TMA technology-based multi-autonomous underwater robot cluster collaborative navigation method | |
| CN110208808B (en) | Passive sonar non-cooperative target line spectrum information fusion method | |
| Chen et al. | Underwater terrain positioning method using maximum a posteriori estimation and PCNN model | |
| El-Banna et al. | Machine learning modeling for IoUT networks: Internet of underwater things | |
| CN114111796B (en) | Parallel fusion positioning method and system of underwater unmanned robot based on information gain | |
| Tang et al. | Factor graph weight particles aided distributed underwater cooperative positioning algorithm | |
| Jie et al. | A fusion method for AUV collaborative location based on information gain weighting | |
| Li et al. | Auv 3d trajectory prediction based on cnn-lstm | |
| CN114415157A (en) | Underwater target multi-model tracking method based on underwater acoustic sensor network | |
| CN112749481A (en) | Factor graph co-location algorithm based on maximum correlation entropy | |
| Shi | Distributed estimation of time-varying biases in relative state measurements | |
| CN112445244B (en) | Target searching method for multiple autonomous underwater robots | |
| CN115061483B (en) | Underwater target state cooperative estimation method based on detection configuration |
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 |