CN106156790B - Distributed cooperation algorithm and data fusion mechanism applied to sensor network - Google Patents

Distributed cooperation algorithm and data fusion mechanism applied to sensor network Download PDF

Info

Publication number
CN106156790B
CN106156790B CN201610405473.9A CN201610405473A CN106156790B CN 106156790 B CN106156790 B CN 106156790B CN 201610405473 A CN201610405473 A CN 201610405473A CN 106156790 B CN106156790 B CN 106156790B
Authority
CN
China
Prior art keywords
sensor
data fusion
matrix
measurement
level data
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.)
Expired - Fee Related
Application number
CN201610405473.9A
Other languages
Chinese (zh)
Other versions
CN106156790A (en
Inventor
谌云莉
孟晗
王楠
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing University of Technology
Original Assignee
Beijing University of Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing University of Technology filed Critical Beijing University of Technology
Priority to CN201610405473.9A priority Critical patent/CN106156790B/en
Publication of CN106156790A publication Critical patent/CN106156790A/en
Application granted granted Critical
Publication of CN106156790B publication Critical patent/CN106156790B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques

Landscapes

  • Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Testing Or Calibration Of Command Recording Devices (AREA)
  • Indication And Recording Devices For Special Purposes And Tariff Metering Devices (AREA)

Abstract

The invention relates to the technical field of sensor data processing, in particular to a distributed cooperative algorithm and a data fusion mechanism applied to a sensor network, which comprises sensor-level data fusion and central-level data fusion, and the method comprises the following steps: the method of the invention uses the result of the central-level data fusion to correct the result of the sensor-level data fusion, can solve the problems of low precision, single information and large data fusion calculation amount of a single sensor system, and can be widely applied to multi-source image composition, robots, intelligent instrument systems, battlefields and unmanned planes, image analysis and understanding, target detection and tracking and automatic target identification.

Description

Distributed cooperation algorithm and data fusion mechanism applied to sensor network
Technical Field
The invention relates to the technical field of sensor data processing, in particular to a distributed cooperation algorithm and a data fusion mechanism applied to a sensor network.
Background
Data fusion integrates the multiband information of a single sensor or the information provided by different sensors, eliminates the redundancy and contradiction possibly existing among the information of multiple sensors, complements the redundancy and contradiction, improves the timeliness and reliability of remote sensing information extraction, and improves the use efficiency of data. Data fusion was first applied in the field of military target identification, and has now been widely applied in multi-source image composition, robots and intelligent instrument systems, battlefields and unmanned planes, image analysis and understanding, target detection and tracking, and automatic target identification.
The existing data fusion technology is generally that a sensor-level data fusion result is used in a central-level data fusion process, and the central-level data fusion result is directly used as a final fusion result.
Disclosure of Invention
The invention aims to solve the problems of low fusion precision and poor reliability caused by the fact that the central-level data fusion can not be used for correcting the result of the sensor-level data fusion in the data fusion process, and provides a distributed cooperation algorithm and a data fusion mechanism applied to a sensor network, which mainly comprise (1) sensor-level data fusion; (2) and (4) performing central-level data fusion.
The sensor-level data fusion specifically comprises the following steps:
(21) acquiring sufficient original measurement data through a sensor, eliminating abnormal data in measurement, linearizing a nonlinear sensor measurement model near a certain nominal value, and solving according to the linear measurement model;
(22) let s ═ H θ, H be the sensor input, H be a full rank nxp observation matrix (N > P), H be composed of observation samples, and let s ═ s [0 [, H is the input to the sensor]，s[1]，...，s[N-1]]TIs the model output of the sensor, s is a matrix of Nx 1, x ═ x [0 [ ]]，x[1]，...，x[N-1]]TIs the measurement quantity output by the sensor, theta is the parameter model of the sensor to be solved, and theta is a matrix of P multiplied by 1;
(23) obtaining a difference J (θ) between a measurement quantity x output from the sensor and s obtained by a sensor linear measurement model by a linear weighting method, where W is a diagonal matrix of N × N, and W ═ C-1And C represents a covariance matrix of zero mean noise;
Figure BSA0000130970070000021
(24) solving the gradient of the J (theta) function to the theta variable;
Figure BSA0000130970070000022
(25) let the gradient of the J (theta) function equal to 0, an estimate of the sensor's linear measurement model can be obtained
Figure BSA0000130970070000023
Figure BSA0000130970070000024
(26) If the noise is uncorrelated during the measurement, the C estimate is a diagonal matrix, let C n]＝diag(σ0，σ1，...，σn)，
Figure BSA0000130970070000025
Wherein H [ n ]]Is a matrix of n × p, hT[n]Is a matrix of 1 xp, X n]＝[x[0]，x[1]，...，x[n]]TThen there is an estimator
Figure BSA0000130970070000026
C[n]Is a covariance matrix of the noise, least squares estimator
Figure BSA0000130970070000027
The covariance matrix of (A) is
Figure BSA0000130970070000028
(27) Calculating an estimate
Figure BSA0000130970070000029
And covariance of least squares estimate ∑ [ n [ [ n ]]Updating of (1);
the update of the estimator is calculated as follows:
Figure BSA00001309700700000210
wherein
Figure BSA00001309700700000211
The covariance of the least squares estimate is updated as follows:
∑[n]＝(I-K[n]hT[n])∑[n-1]。
(28) and performing Dempster-Shafer data fusion on state estimation values obtained by sensors with different properties or the same properties to provide initial conditions for central-level data fusion.
The central level data fusion comprises the following steps:
(31) establishing a state equation and a measurement equation of the observed system.
Setting the system state equation as xk＝Axk-1+Buk-1+wk-1The measurement equation is yk＝Cxk-1+vkWherein x iskIs composed of (x)1，k，x2，k，...，xN，k) State variables, u, constituting Nx 1kIs (u)1，k，u2，k，...uL，k) Constituent input vectors of L x 1, ykIs prepared from (y)1，k，y2，k，...yM，k) Constituent Mx 1 observation vectors, wkIs prepared from (w)1，k，w2，k，...wN，k) Forming a Nx 1 process noise vector, vkIs prepared from (v)1，k，v2，k，...vM，k) Forming an M multiplied by 1 measurement noise vector, wherein A is an N multiplied by N state transition matrix, B is an N multiplied by L input relation matrix, and C is an M multiplied by N measurement relation matrix;
(32) on the basis of the optimal estimated value of the k-1 th state variable, predicting the prior state value at the k moment by using a state equation of the system.
Figure BSA0000130970070000031
Pk -＝APk-1AT+ Q, wherein
Figure BSA0000130970070000032
Represents a prior estimate of the state variable at time k,
Figure BSA0000130970070000033
represents the optimal estimate of the state variable at time k-1,
Figure BSA0000130970070000034
covariance representing prior estimation error at time kThe matrix is a matrix of a plurality of matrices,
Figure BSA0000130970070000035
covariance matrix representing the posterior estimation error at time k, Q ═ E [ w [ ]kwk T]Representing the covariance of the process noise vector.
(33) Correcting the state prior estimated value obtained by the last step of prediction by using the actual measured value
Figure BSA0000130970070000036
Obtaining the optimal estimated value of the system state
Figure BSA0000130970070000037
Kk＝Pk -CT(CPk -CT+R)-1，
Figure BSA0000130970070000038
Pk＝(I-KkC)Pk -Wherein R ═ E [ v ═kvk T]Representing the covariance of the measured noise vector.
(34) In order to avoid matrix inversion operation, scalar sequential processing method is adopted for processing, correlation among measurement noises is assumed, and orthogonal decomposition is carried out on measurement noise covariance R, wherein R is MDMTWhere M is an orthogonal matrix, D ═ diag (σ)1，σ2，...，σM) For diagonal matrix, the measurement equation is subjected to matrix transformation MTyk＝MTCxk-1+MTvkDefining a new observation vector yk′＝MTykThe measurement matrix C ═ MTC, measuring noise vector vk′＝MTvkThe new measurement noise vector has a covariance of R ═ E [ v ═ vk′vk′T]＝E[MTvkvk TM]＝MTRecording the measurement matrix when RM is equal to D
Figure BSA0000130970070000041
yk′＝(y1，k′，y2，k′，...，yM，k′)，
Figure BSA0000130970070000042
P0，k＝Pk -Then the state and mean square error matrix can be corrected using the following scalar algorithm:
Figure BSA0000130970070000043
Pm，k＝(I-km，kcm′)Pm-1，kwherein M is 1, 2.. times.m,
Figure BSA0000130970070000044
Pk＝PM，k。
(35) use the center-level data fusion result to correct the sensor-level data fusion result.
The invention brings the following beneficial results: the invention uses the result of the central-level data fusion to correct the result of the sensor-level data fusion, can solve the problems of low precision, single information and large data fusion calculation amount of a single sensor system, and can be widely applied to multi-source image composition, robots and intelligent instrument systems, battlefields and unmanned planes, image analysis and understanding, target detection and tracking and automatic target identification.
Drawings
The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
FIG. 1 is a block diagram of the present invention.
FIG. 2 is a flow chart of sensor-level data fusion.
Fig. 3 is a flow chart of sensor measurement modeling.
FIG. 4 is a flow chart of rejecting abnormal sensor measurement data.
FIG. 5 is a flow chart of center level data fusion.
Detailed Description
FIG. 1 is a general diagram of data fusion of the present invention, in which sensor-level data fusion obtains partial state information of an observed system and measurement noise variance of sensors by fusing the sensors distributed in different categories of different objects to be measured, and provides an initial condition for central-level data fusion, the central-level data fusion obtains optimal state information of the system to be measured by filtering the result of the sensor-level data fusion through fuzzy Kalman filtering, and the result of the sensor-level data fusion is corrected by using the result of the central-level data fusion.
Fig. 2 is a flow chart of sensor-level data fusion, in which a homogeneous sensor measures the same measured state to improve the reliability of the measured state, a heterogeneous sensor measures different measured states to provide more comprehensive system state information, and system state data measured by the homogeneous sensor and the heterogeneous sensor provides initial conditions for central-level data fusion through Dempster-Shafer data fusion. The sensor-level data fusion specifically comprises the following steps:
(21) firstly, establishing a measurement model of a sensor, specifically referring to a flow chart in fig. 3, for a nonlinear sensor measurement model, linearizing the model near a certain nominal value, then solving according to the linear measurement model, obtaining sufficient measurement original data through the sensor, and removing abnormal data in measurement, for a time-invariant system, specifically referring to fig. 4, a mean value α and a variance σ of a measured state can be firstly solved, then removing data of the sensor measurement data outside an interval (α -3 σ, α +3 σ), and then solving the mean value and the variance;
(22) let s ═ H θ, H be the sensor input, H be a full rank nxp observation matrix (N > P), H be composed of observation samples, and let s ═ s [0 [, H is the input to the sensor]，s[1]，...，s[N-1]]TIs the model output of the sensor, s is a matrix of Nx 1, x ═ x [0 [ ]]，x[1]，...，x[N-1]]TIs the measurement quantity output by the sensor, theta is the parameter model of the sensor to be solved, and theta is a matrix of P multiplied by 1;
(23) obtaining the difference J between the measurement quantity x output from the sensor and s solved by the sensor linear measurement model in a linear weighting manner(θ), W is a diagonal matrix of N × N, and W ═ C-1And C represents a covariance matrix of zero mean noise;
Figure BSA0000130970070000061
(24) solving the gradient of the J (theta) function to the theta variable;
Figure BSA0000130970070000062
(25) let the gradient of the J (theta) function equal to 0, an estimate of the sensor's linear measurement model can be obtained
Figure BSA0000130970070000063
Figure BSA0000130970070000064
(26) If the noise is uncorrelated during the measurement, the C estimate is a diagonal matrix, let C n]＝diag(σ0，σ1，...，σn)，
Figure BSA0000130970070000065
Wherein H [ n ]]Is a matrix of n × p, hT[n]Is a matrix of 1 xp, X n]＝[x[0]，x[1]，...，x[n]]TThen there is an estimator
Figure BSA0000130970070000066
C[n]Is a covariance matrix of the noise, least squares estimator
Figure BSA0000130970070000067
The covariance matrix of (A) is
Figure BSA0000130970070000068
(27) Calculating an estimate
Figure BSA0000130970070000069
And minimumCovariance of the two-times estimator ∑ [ n [ [ n ]]Updating of (1);
the update of the estimator is calculated as follows:
Figure BSA00001309700700000610
wherein
Figure BSA00001309700700000611
The covariance of the least squares estimate is updated as follows:
∑[n]＝(I-K[n]hT[n])∑[n-1]。
fig. 5 is a flowchart of the central-level data fusion, which specifically includes the following steps:
(31) establishing a state equation and a measurement equation of the observed system, and converting the nonlinear state equation and the measurement equation of the observed system into a linear system for solving by using a local linearization method.
Setting the system state equation as xk＝Axk-1+Buk-1+wk-1The measurement equation is yk＝Cxk-1+vkWherein x iskIs composed of (x)1，k，x2，k，...，xN，k) State variables, u, constituting Nx 1kIs (u)1，k，u2，k，...uL，k) Constituent input vectors of L x 1, ykIs prepared from (y)1，k，y2，k，...yM，k) Constituent Mx 1 observation vectors, wkIs prepared from (w)1，k，w2，k，...wN，k) Forming a Nx 1 process noise vector, vkIs prepared from (v)1，k，v2，k，...vM，k) Forming an M multiplied by 1 measurement noise vector, wherein A is an N multiplied by N state transition matrix, B is an N multiplied by L input relation matrix, and C is an M multiplied by N measurement relation matrix;
(32) on the basis of the optimal estimated value of the k-1 th state variable, predicting the prior state value at the k moment by using a state equation of the system.
Figure BSA0000130970070000071
Pk -＝APk-1AT+ Q, wherein
Figure BSA0000130970070000072
Represents a prior estimate of the state variable at time k,
Figure BSA0000130970070000073
represents the optimal estimate of the state variable at time k-1,
Figure BSA0000130970070000074
a covariance matrix representing the prior estimation error at time k,
Figure BSA0000130970070000075
covariance matrix representing the posterior estimation error at time k, Q ═ E [ w [ ]kwk T]Representing the covariance of the process noise vector.
(33) Correcting the state prior estimated value obtained by the last step of prediction by using the actual measured value
Figure BSA0000130970070000076
Obtaining the optimal estimated value of the system state
Figure BSA0000130970070000077
Kk＝Pk -CT(CPk -CT+R)-1，
Figure BSA0000130970070000078
Pk＝(I-KkC)Pk -Wherein R ═ E [ v ═kvk T]Representing the covariance of the measured noise vector.
(34) In order to avoid matrix inversion operation, scalar sequential processing method is adopted for processing, correlation among measurement noises is assumed, and orthogonal decomposition is carried out on measurement noise covariance R, wherein R is MDMTWhere M is an orthogonal matrix, D ═ diag (σ)1，σ2，...，σM) For diagonal matrix, the measurement equation is subjected to matrix transformation MTyk＝MTCxk-1+MTvkDefining a new observation vector yk′＝MTykThe measurement matrix C ═ MTC, measuring noise vector vk′＝MTvkThe new measurement noise vector has a covariance of R ═ E [ v ═ vk′vk′T]＝E[MTvkvk TM]＝MTRecording the measurement matrix when RM is equal to D
Figure BSA0000130970070000079
yk′＝(y1，k′，y2，k′，...，yM，k′)，
Figure BSA0000130970070000081
P0，k＝Pk -Then the state and mean square error matrix can be corrected using the following scalar algorithm:
Figure BSA0000130970070000082
Pm，k＝(I-km，kcm′)Pm-1，kwherein M is 1, 2.. times.m,
Figure BSA0000130970070000083
Pk＝PM，k。
(35) use the center-level data fusion result to correct the sensor-level data fusion result.
It will be appreciated by those of ordinary skill in the art that the examples described herein are intended to assist the reader in understanding the manner in which the invention is practiced, and it is to be understood that the scope of the invention is not limited to such specifically recited statements and examples. Those skilled in the art can make various other specific changes and combinations based on the teachings of the present invention without departing from the spirit of the invention, and these changes and combinations are within the scope of the invention.

Claims (2)

1. A distributed cooperative algorithm and a data fusion method applied to a sensor network are characterized by comprising sensor-level data fusion and central-level data fusion, wherein the sensor-level data fusion obtains partial state information of an observed system and measurement noise variance of sensors by fusing the sensors distributed in different types of different measured objects to provide initial conditions for the central-level data fusion, the central-level data fusion carries out filtering processing on the result of the sensor-level data fusion through fuzzy Kalman filtering to obtain the optimal state information of the measured system, and the result of the sensor-level data fusion is corrected by using the result of the central-level data fusion;
the sensor-level data fusion comprises the following steps:
(21) acquiring sufficient original measurement data through a sensor, eliminating abnormal data in measurement, linearizing a nonlinear sensor measurement model near a certain nominal value, and solving according to the linear measurement model;
(22) let s ═ H θ, H be the sensor input, H be a full rank nxp observation matrix (N > P), H be composed of observation samples, and let s ═ s [0 [, H is the input to the sensor]，s[1]，...，s[N-1]]TIs the model output of the sensor, s is a matrix of Nx 1, x ═ x [0 [ ]]，x[1]，...，x[N-1]]TIs the measurement quantity output by the sensor, theta is the parameter model of the sensor to be solved, and theta is a matrix of P multiplied by 1;
(23) obtaining a difference J (θ) between a measurement quantity x output from the sensor and s obtained by a sensor linear measurement model by a linear weighting method, where W is a diagonal matrix of N × N, and W ═ C-1And C represents a covariance matrix of zero mean noise;
Figure FDA0002313001700000011
(24) solving the gradient of the J (theta) function to the theta variable;
Figure FDA0002313001700000012
(25) let the gradient of the J (theta) function equal to 0, an estimate of the sensor's linear measurement model can be obtained
Figure FDA0002313001700000013
Figure FDA0002313001700000014
(26) If the noise is uncorrelated during the measurement, the C estimate is a diagonal matrix, let C n]＝diag(σ0，σ1，...，σn)，
Figure FDA0002313001700000015
Wherein H [ n ]]Is a matrix of n × p, hT[n]Is a matrix of 1 xp, X n]＝[x[0]，x[1]，...，x[n]]TThen there is an estimator
Figure FDA0002313001700000021
C[n]Is a covariance matrix of the noise, least squares estimator
Figure FDA0002313001700000022
The covariance matrix of (A) is
Figure FDA0002313001700000023
(27) Calculating an update of the covariance ∑ [ n ] of the estimator and the least squares estimator;
the update of the estimator is calculated as follows:
Figure FDA0002313001700000024
wherein
Figure FDA0002313001700000025
The covariance of the least squares estimate is updated as follows:
∑[n]＝(I-K[n]hT[n])∑[n-1]；
(28) and performing Dempster-Shafer data fusion on state estimation values obtained by sensors with different properties or the same properties to provide initial conditions for central-level data fusion.
2. The distributed collaboration algorithm and data fusion method applied to the sensor network as claimed in claim 1, wherein the central level data fusion comprises the following steps:
(31) establishing a state equation and a measurement equation of the observed system;
setting the system state equation as xk＝Axk-1+Buk-1+wk-1The measurement equation is yk＝Cxk-1+vkWherein x iskIs composed of (x)1，k，x2，k，...xNK) State variables, u, constituting Nx 1kIs (u)1，k，u2，k，...uL，k) Constituent input vectors of L x 1, ykIs prepared from (y)1，k，y2，k，...yM，k) Constituent Mx 1 observation vectors, wkIs prepared from (w)1，k，w2，k，...wN，k) Forming a Nx 1 process noise vector, vkIs prepared from (v)1，k，v2，k，...vM，k) Forming an M multiplied by 1 measurement noise vector, wherein A is an N multiplied by N state transition matrix, B is an N multiplied by L input relation matrix, and C is an M multiplied by N measurement relation matrix;
(32) on the basis of the optimal estimated value of the k-1 th state variable, predicting a prior state value at the k moment by using a state equation of the system;
Figure FDA0002313001700000026
Pk-＝APk-1AT+ Q, wherein
Figure FDA0002313001700000027
Represents a prior estimate of the state variable at time k,
Figure FDA0002313001700000028
represents the optimal estimate of the state variable at time k-1,
Figure FDA0002313001700000029
a covariance matrix representing the prior estimation error at time k,
Figure FDA00023130017000000210
covariance matrix representing the posterior estimation error at time k, Q ═ E [ w [ ]kwk T]A covariance representing the process noise vector;
(33) correcting the state prior estimated value obtained by the last step of prediction by using the actual measured value
Figure FDA00023130017000000211
Obtaining the optimal estimated value of the system state
Figure FDA0002313001700000031
Kk＝Pk -CT(CPk -CT+R)-1，
Figure FDA0002313001700000032
Pk＝(I-KkC)Pk -Wherein R ═ E [ v ═kvk T]A covariance representing the measurement noise vector;
(34) in order to avoid matrix inversion operation, scalar sequential processing method is adopted for processing, correlation among measurement noises is assumed, and orthogonal decomposition is carried out on measurement noise covariance R, wherein R is MDMTWhere M is an orthogonal matrix and D is a diag (σ 1, σ 2.., σ M), a matrix transformation M is performed on the measurement equationTyk＝MTCxk-1+MTvkDefining a new observation vector yk′＝MTykThe measurement matrix C ═ MTC, measuring noise vector vk′＝MTvkThe new measurement noise vector has a covariance of R ═ E [ v ═ vk′vk′T]＝E[MTvkvk TM]＝MTRecording the measurement matrix when RM is equal to D
Figure FDA0002313001700000033
yk′＝(y1，k′，y2，k′，...，yM，k′)，
Figure FDA0002313001700000034
P0，k＝PkThe state and mean square error matrix can then be corrected using the following scalar algorithm:
Figure FDA0002313001700000035
Pm，k＝(I-km，kcm′)Pm-1，kwherein M is 1, 2k＝xM，k，Pk＝PM，k；
(35) Use the center-level data fusion result to correct the sensor-level data fusion result.
CN201610405473.9A 2016-06-08 2016-06-08 Distributed cooperation algorithm and data fusion mechanism applied to sensor network Expired - Fee Related CN106156790B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610405473.9A CN106156790B (en) 2016-06-08 2016-06-08 Distributed cooperation algorithm and data fusion mechanism applied to sensor network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610405473.9A CN106156790B (en) 2016-06-08 2016-06-08 Distributed cooperation algorithm and data fusion mechanism applied to sensor network

Publications (2)

Publication Number Publication Date
CN106156790A CN106156790A (en) 2016-11-23
CN106156790B true CN106156790B (en) 2020-04-14

Family

ID=57353197

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610405473.9A Expired - Fee Related CN106156790B (en) 2016-06-08 2016-06-08 Distributed cooperation algorithm and data fusion mechanism applied to sensor network

Country Status (1)

Country Link
CN (1) CN106156790B (en)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108364014A (en) * 2018-01-08 2018-08-03 东南大学 A kind of multi-sources Information Fusion Method based on factor graph
CN108896047B (en) * 2018-05-09 2020-11-06 上海交通大学 Distributed sensor network collaborative fusion and sensor position correction method
US11227486B2 (en) * 2018-07-16 2022-01-18 Here Global B.V. Method, apparatus, and system for estimating vulnerable road users
CN111667073B (en) * 2020-03-30 2021-12-28 西北工业大学 Fuzzy Kalman filtering target tracking method improved by evidence theory
CN111860613B (en) * 2020-06-30 2023-10-27 江苏科技大学 Multi-target tracking and state predicting method based on multi-source heterogeneous signals
CN114260890B (en) * 2020-08-26 2023-11-03 北京市商汤科技开发有限公司 Method and device for determining state of robot, robot and storage medium

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102629418A (en) * 2012-04-09 2012-08-08 浙江工业大学 Fuzzy kalman filtering-based traffic flow parameter prediction method
CN104408324A (en) * 2014-12-11 2015-03-11 云南师范大学 D-S evidence theory based multi-sensor information fusion method
CN104484410A (en) * 2014-12-16 2015-04-01 芜湖乐锐思信息咨询有限公司 Data fusion method and system applied to big data system
CN105352535A (en) * 2015-09-29 2016-02-24 河海大学 Measurement method on the basis of multi-sensor date fusion
CN105427200A (en) * 2015-11-05 2016-03-23 上海电力学院 Power plant fault diagnosis method based on fuzzy theory and data fusion
CN105467382A (en) * 2015-12-31 2016-04-06 南京信息工程大学 SVM (Support Vector Machine)-based multi-sensor target tracking data fusion algorithm and system thereof

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102629418A (en) * 2012-04-09 2012-08-08 浙江工业大学 Fuzzy kalman filtering-based traffic flow parameter prediction method
CN104408324A (en) * 2014-12-11 2015-03-11 云南师范大学 D-S evidence theory based multi-sensor information fusion method
CN104484410A (en) * 2014-12-16 2015-04-01 芜湖乐锐思信息咨询有限公司 Data fusion method and system applied to big data system
CN105352535A (en) * 2015-09-29 2016-02-24 河海大学 Measurement method on the basis of multi-sensor date fusion
CN105427200A (en) * 2015-11-05 2016-03-23 上海电力学院 Power plant fault diagnosis method based on fuzzy theory and data fusion
CN105467382A (en) * 2015-12-31 2016-04-06 南京信息工程大学 SVM (Support Vector Machine)-based multi-sensor target tracking data fusion algorithm and system thereof

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
"A Wireless Micro Inertial Measurement Unit(IMU)";Fabian Höflinger等;《IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT》;20130120;第62卷(第9期);参见摘要 *
"一氧化碳气体质量浓度智能检测仪的研究与设计";冯冬青等;《仪表技术与传感器》;20091231(第3期);参见第2.1节 *
"一种改进的自适应指数加权衰减记忆滤波算法";田宠等;《武汉理工大学学报》;20111130;第33卷(第11期);参见第1节 *
"基于卡尔曼滤波的交通信息融合方法";杨兆升等;《吉林大学学报（工学版）》;20070930;第37卷(第5期);参见第1.1节，图2 *

Also Published As

Publication number Publication date
CN106156790A (en) 2016-11-23

Similar Documents

Publication Publication Date Title
CN106156790B (en) Distributed cooperation algorithm and data fusion mechanism applied to sensor network
CN107765347B (en) Short-term wind speed prediction method based on Gaussian process regression and particle filtering
CN111178385B (en) Target tracking method for robust online multi-sensor fusion
CN108960334B (en) Multi-sensor data weighting fusion method
CN108304685A (en) A kind of non-linear degradation equipment method for predicting residual useful life and system
CN104462015B (en) Process the fractional order linear discrete system state updating method of non-gaussian L é vy noises
CN110289989A (en) A kind of distributed state estimation method based on volume Kalman filtering algorithm
CN109785247B (en) Method and device for correcting abnormal point cloud data of laser radar and storage medium
CN110889862B (en) Combined measurement method for multi-target tracking in network transmission attack environment
CN103809439B (en) A kind of be applicable to remote control system controling parameters pass rank discrimination method
CN110133524A (en) Battery state of charge calculation method, device, server and medium
US20240038324A1 (en) Knowledge reuse-based method and system for predicting cell concentration in fermentation process
CN108460481A (en) Unmanned plane spot development law prediction technique based on Recognition with Recurrent Neural Network
CN109523129B (en) Method for fusing information of multiple sensors of unmanned vehicle in real time
CN110057353B (en) Method for interrupting track association based on communication signal assistance
CN104331087B (en) Robust underwater sensor network target tracking method
CN117828285A (en) Production data management system and method
Dubois et al. Performance evaluation of a moving horizon estimator for multi-rate sensor fusion with time-delayed measurements
CN111929585B (en) Battery charge state calculating device, method, server and medium
CN114511088A (en) Bayesian model updating method and system for structure damage recognition
CN110673132B (en) Real-time filtering method for trace point sequence for multi-frame joint detection and tracking
CN107145474B (en) Kalman filtering algorithm based on autoregressive model
CN115391725A (en) Method, system, equipment and medium for identifying parameters of fractional order chaotic system
CN114611068A (en) High maneuvering target tracking method
CN107590509B (en) Cherenov fusion method based on maximum expectation approximation

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20200414