CN106650785A - Weighted evidence fusion method based on evidence classification and conflict measurement - Google Patents

Weighted evidence fusion method based on evidence classification and conflict measurement Download PDF

Info

Publication number
CN106650785A
CN106650785A CN201610984406.7A CN201610984406A CN106650785A CN 106650785 A CN106650785 A CN 106650785A CN 201610984406 A CN201610984406 A CN 201610984406A CN 106650785 A CN106650785 A CN 106650785A
Authority
CN
China
Prior art keywords
evidence
conflict
coefficient
evidences
unit
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201610984406.7A
Other languages
Chinese (zh)
Other versions
CN106650785B (en
Inventor
李军伟
刘先省
胡振涛
金勇�
周林
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.)
Henan University
Original Assignee
Henan University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Henan University filed Critical Henan University
Priority to CN201610984406.7A priority Critical patent/CN106650785B/en
Publication of CN106650785A publication Critical patent/CN106650785A/en
Application granted granted Critical
Publication of CN106650785B publication Critical patent/CN106650785B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques
    • G06F18/251—Fusion techniques of input or preprocessed data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques
    • G06F18/257—Belief theory, e.g. Dempster-Shafer

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)
  • Complex Calculations (AREA)

Abstract

The invention discloses a weighted evidence fusion method based on evidence classification and conflict measurement. The method comprises the following steps: firstly, measurement information of multiple sensors is acquired and converted to evidence information; then, whether to have consistency is calculated on each two evidences; if yes, a consistency conflict coefficient is calculated, and if not, an inconsistency conflict coefficient is calculated; then, the weight coefficient of an evidence is solved jointly, and a fused evidence is corrected according to the weight coefficient; and finally, a Dempster combination rule is adopted to carry out one-by-one fusion on evidences after correction, and a decision result for final target recognition is outputted. Compared with the traditional algorithm, the scheme of the invention comprehensively considers the difference of basic probability assignment of focal elements in a single subset between evidences and the support degree of the part in which intersections of different focal elements between evidences are not empty, the conflict degree between evidences is measured jointly, the weight coefficient of the fused evidence is determined on the basis, the fused evidence is corrected, and important theoretical significance and practical value are facilitated.

Description

Based on the classification of evidence and the weighted evidence fusion method of measure method for conflict
Technical field
The present invention relates to technical field of multisource information fusion, more particularly to it is a kind of based on the classification of evidence and measure method for conflict plus Warrant is according to fusion method.
Background technology
At present, developing rapidly with computer technology and information technology, various many sensings towards complicated applications background Device information system is continued to bring out, but sensor has various uncertain factors in perception, therefore, the data that system is obtained With all multiple features such as uncertain and unreliable.In order to extract real-time, effective and accurate information from multi-sensor information, use To judge the attribute and feature of identification target, it is necessary to carry out effective information fusion to multi-sensor data.Multi-source information melts Multiple sensors are provided conjunction technology complementary and redundancy in the time or spatially, improve target identification system performance and Reliability, obtains to more complete, more accurate, the more reliable inference of things.Dempster-Shafer evidence theories are in uncertain letter The unique advantage of the aspect such as breath expression and fusion method, the sign and fusion for decision level uncertain information provides strong Means, obtain a wide range of applications in fields such as target identification, image procossings.But in actual applications, due to sensor itself Limitation and monitoring of environmental in interference or the factor such as artificial disturbance, the identification target information for causing its output there may be The conflict even situation of contradiction.Evidence theory generally weighs the conflict spectrum between evidence using conflict coefficient, but research shows Conflict coefficient has some shortcomings, and such as two on all four evidences, the conflict coefficient calculated between evidence is but not zero. In addition, in real application systems, during conflicting evidence fusion problem high using the process of Dempster rules of combination, often obtaining Run counter to the fusion results of intuition, it is impossible to carry out effective decision-making, greatly have impact on the decision-making performance of emerging system.
The content of the invention
It is an object of the invention to provide a kind of based on the classification of evidence and the weighted evidence fusion method of measure method for conflict, Neng Gouyou Correct decisions are made to recognizing target in effect ground.
The technical solution used in the present invention is：
Based on the classification of evidence and the weighted evidence fusion method of measure method for conflict, including following step：
A, the Basic Probability As-signment by obtaining the burnt unit of the corresponding evidence of multiple sensor measurement informations, by each evidence Regard a vector, the vector m of i-th evidence asi=(mi(θ1),…,mi(θr),…,mi(θk))TRepresent, wherein i= 1,2 ..., n, n are the sum of evidence vector, and k is Jiao unit number in framework of identification Θ, r=1,2 ..., k；
B, to above-mentioned i-th evidence miWith j-th evidence mjWhether it is that consistent evidence is judged：According to miMiddle maximum The corresponding burnt unit of Basic Probability As-signment and evidence mjWhether the corresponding burnt unit of Basic Probability As-signment of middle maximum is that same burnt unit is next Judge, if same burnt unit, then claim evidence miAnd mjFor consistent evidence, otherwise claim evidence miAnd mjFor non-uniform evidence, wherein i, J=1,2 ..., n；i≠j；
C, by following formulaDifference property coefficient is calculated, any i-th evidence m is obtainediWith J evidence mjBetween otherness coefficient d (mi,mj), M in formulaiRepresent a row vector, DiRepresent a column vector；
D, by any evidence miAnd mjBetween otherness coefficient d (mi,mj) public affairs are passed through respectively according to classification of evidence result Formula：WithCalculate and appoint Anticipate consistent evidence miAnd mjWith non-uniform evidence miAnd mjBetween conflict coefficient conf (mi,mj)；
E, by any i-th evidence m for obtainingiWith j-th evidence mjBetween conflict coefficient conf (mi,mj) by public affairs Formula： Try to achieve i-th evidence with Total conflict spectrum factor conf (m of other n-1 evidencei) and i-th evidence and other n-1 evidences relative degree of support Factor t ruf (mi), and using relative degree of support factor t ruf maximum in n evidencemaxWith i evidence and other n-1 The relative degree of support factor t ruf (m of evidencei) obtain weight coefficient ω by following formulai,
Jiao unit θ in F, i-th evidence of noterBasic Probability As-signment mi(θr) represent, wherein r=1,2 ..., k, after amendment I-th evidence in Jiao unit θrBasic Probability As-signment mi d(θr) represent, according to the weight coefficient ω obtained in step EiPass through Formula：
Evidence to merging is modified；
It is G, last, revised evidence is merged one by one using Dempster rules of combination, the base of Jiao unit A after fusion The corresponding burnt first corresponding identification target of the result of decision for target identification of maximum of this probability assignment m (A), as decision-making are most Termination fruit.
Row vector M described in described step Cr=[- mi(θr)mj(θ1),…,|mi(θr)-mj(θr)|,
…,-mi(θr)mj(θk)], column vector DrExpression formula is Wherein, r=1,2 ..., k.
Described Dempster rules of combination are：
Wherein, m (A) represents the Basic Probability As-signment of Jiao unit A, and K is conflict coefficient, r, l=1,2 ..., k,For empty set.
As application background, the information that sensor is provided is converted target identification of the present invention based on multisensor measurement For evidence, evidence is classified from the angle of evidence burnt unit's Basic Probability As-signment, using the otherness system between evidence Number and exponential function, the conflict coefficient characterized between evidence is constructed respectively according to classification of evidence result, so as to the evidence for merging It is modified.The present invention program considers the difference of the burnt unit's Basic Probability As-signment of list collection between evidence compared with traditional algorithm Different burnt unit's common factors different and evidence between are not the degree of support of empty set part, the common conflict spectrum weighed between evidence, The weight coefficient of fusion evidence is determined on the basis of this, and is modified to merging evidence, finally using Dempster rules of combination Revised evidence is merged one by one and is made the decision-making last to target identification, with important theory significance and applied valency Value.
Description of the drawings
Fig. 1 is the flow chart of the present invention.
Specific embodiment
As shown in figure 1, the present invention includes following step：
A, the Basic Probability As-signment by obtaining the burnt unit of the corresponding evidence of multiple sensor measurement informations, each is demonstrate,proved According to regarding a vector, the vector m of i-th evidence asiRepresent, wherein i=1,2 ..., n, n is the sum of evidence vector, K is Jiao unit number in framework of identification Θ；The different sensor of multiple properties will be obtained first to convert the identification information of target For multiple evidences, and regard each evidence for merging as a vector.Assume that obtaining n evidence is respectively m1,m2,…,mn, Assume that Jiao unit in framework of identification Θ is θ1,θ2,…,θk, the corresponding burnt unit's Basic Probability As-signment of i-th evidence is respectively mi (θ1),mi(θ2),…,mi(θk), regard evidence as vector, then the corresponding element of i-th evidence vector is followed successively by：mi(θ1),mi (θ2),…,mi(θk)。
During acquisition multisensor observation information of the present invention, also had according to the different modes for obtaining of actual conditions Institute is different, for example：In the data fusion of multisensor syste target identification, the species of target regards proposition as, each sensor The judged result to targeted species be given by measurement, process is seen and is testified.
B, to above-mentioned each evidence vector miAnd mjCarry out difference property coefficient calculating：And record any i-th evidence vector mi With j-th evidence vector mjBetween difference property coefficient be d (mi,mj), wherein i, j=1,2 ..., n；i≠j；
Difference property coefficient between evidence calculates the otherness system that evidence is obtained by relation between row vector and column vector Number.Difference property coefficient between evidence embodies the otherness between two evidences, and the difference property coefficient between evidence is bigger, card Conflict spectrum according between is bigger.Difference property coefficient is calculated and passes through following formula in specific step C of the present inventionObtain any i-th evidence vector miWith j-th evidence vector mjBetween otherness coefficient d (mi, mj), M in formulaiRepresent a row vector, DiRepresent a column vector.Described row vector MrExpression formula is Mr=[- mi (ωr)mj(ω1),…,|mi(ωr)-mj(ωr)|,…,-mi(ωr)mj(ωk)], column vector DrExpression formula isWherein, r=1,2 ..., k.
C, by any evidence miAnd mjBetween otherness coefficient d (mi,mj) index of the combination with natural constant e as the truth of a matter Function, according to the classification of evidence result formula is passed through respectively：With Calculate any evidence miAnd mjBetween conflict coefficient conf (mi,mj)；Wherein, e is natural constant, is one and is approximately equal to 2.71828182845904523536 ... irrational number.
D, by any i-th evidence vector miWith j-th evidence vector mjBetween conflict coefficient conf (mi,mj) pass through 3. 2. 1. following three formula try to achieve total conflict spectrum factor conf (m of i-th evidence and other n-1 evidencei) (if two Conflict coefficient between evidence is bigger, illustrates that the conflict spectrum between two evidences is bigger) and i-th evidence and other n-1 The relative degree of support factor t ruf (m of evidencei), and using relative degree of support factor t ruf maximum in n evidencemaxWith I-th evidence degree of support factor t ruf (m relative with other n-1 evidencesi) 4. obtain weight coefficient ω by following formulai； Assume evidence mi,mj(i, j=1,2 ..., n) between conflict coefficient conf (mi,mj), i-th evidence and other n-1 evidence Total conflict spectrum factor conf (mi) represent, i-th evidence degree of support factor relative with other n-1 evidence is used truf(mi) represent, maximum relative degree of support factor truf in n evidencemaxRepresent, the weight coefficient of i-th evidence Use ωiRepresent.Specifically formula is：
trufmax=max (truf (m1),…,truf(mi),…,truf(mn))③
Jiao unit θ in E, i-th evidencerBasic Probability As-signment mi(θr) represent, burnt unit in revised i-th evidence θrBasic Probability As-signment useRepresent, according to weight coefficient ωiBy formula：
Evidence to merging is modified：Then Dempster groups are adopted Normally merged, described Dempster rules of combination are：
Wherein, m (A) represents the Basic Probability As-signment of Jiao unit A, and K is conflict coefficient, r, l=1,2 ..., k,For empty set. The corresponding burnt unit of the maximum of the Basic Probability As-signment m (A) of Jiao unit A is the corresponding identification of the result of decision of target identification after fusion Target, as decision-making final result.The present invention program considers the burnt unit of list collection in evidence substantially general compared with traditional algorithm The difference of rate assignment and non-list integrate burnt unit and intersect not as the conflict spectrum between the characterized evidence of the degree of support of empty set, lead to The weight coefficient that evidences conflict coefficient determines fusion evidence is crossed, and Dempster rules of combination pair are adopted after being modified to evidence Revised evidence is merged one by one, obtains the rational result of decision, can be very good to be applied in field of target recognition, is had There are important theory significance and using value.
Illustrating conflict coefficient in evidence theory with concrete instance below can not weigh conflict spectrum between evidence.Example 1 assumes that framework of identification is Θ={ θ1,θ2,θ3,θ4, there are two evidence m that property is different1And m2, its basic probability assignment function Respectively：
m1:m1(θ1)=0.25, m1(θ2)=0.25, m1(θ3)=0.25, m1(θ4)=0.25；
m2:m2(θ1)=0.25, m2(θ2)=0.25, m2(θ3)=0.25, m2(θ4)=0.25.
By calculating conflict coefficient K=0.75, evidence m may determine that according to K values1And m2Between there is conflict, this is with two The intuition that there is no conflict between individual on all four evidence judges to contradict.Can according to the measure method for conflict method of this paper Conf=0 is obtained, it is consistent with theoretical analysis result.
Carried out using specific example the measure method for conflict method in description of test invention can effectively weigh evidence it Between conflict spectrum：
Example 2 assumes that framework of identification is Θ={ 1,2 ..., 20 }, has two basic probability assignment functions to be respectively:
m1:m1(a)=0.8, m1(2,3,4)=0.05, m1(7)=0.05, m1(Θ)=0.1；
m2:m2(1,2,3,4,5)=1.
According to { 1 }, { 1,2 } ..., { 1,2 ..., 20 } change wherein subset a.
The conflict spectrum tested between evidence by various measure method for conflict methods, evidence m1And m2Between conflict coefficient with The results contrast the change of subset a and change is as shown in table 1.Wherein, K represents rushing in Dempster-Shafer evidence theories Prominent coefficient, difBetP be Pignistic probability metricses, dJJousselme evidence distances are represented, conf is in patent of the present invention Evidences conflict balancing method.
Found out by table 1, with the change of subset a, conflict coefficient K is always 0.05, it is impossible to reflect evidence m1And m2Between Conflict situation about changing with subset, is not inconsistent with intuition analysis.Evidences conflict balancing method and its other party in patent of the present invention Method changes with the change of subset a, and when the conflict between evidence becomes hour, measure method for conflict coefficient also diminishes, and meets theoretical point Analysis, can effectively weigh the conflict spectrum between evidence.
The different evidence measure method for conflict coefficients comparison results of table 1
Illustrate that the fusion method in this patent can be overcome with the fusion of the high conflicting evidence of effectively solving with specific experiment below The result of running counter to intuition that Dempster rules of combination occur when high conflicting evidence is solved：
Example 3 assumes that framework of identification is Θ={ θ1,θ2,θ3, wherein θ1Represent bomber, θ2Represent airliner, θ3Represent war Bucket machine, obtains 4 different evidences of observation information composition property and is respectively using 4 different sensors：
m1:m1(θ1)=0.5, m1(θ2)=0.2, m1(θ3)=0.3；
m2:m2(θ1)=0, m2(θ2)=0.9, m2(θ3)=0.1；
m3:m3(θ1)=0.6, m3(θ2)=0.1, m3(θ3)=0.3；
m4:m4(θ1)=0.8, m4(θ2)=0.1, m4(θ3)=0.1.
The evidences conflict balancing method that the evident information that comprehensive 4 sensors are given is combined in the inventive method can obtain evidence Relative reliability be respectively 0.3033,0.1141,0.2955,0.2871.It can thus be appreciated that evidence m2It is former because of sensor Cause or external environmental interference cause and other sensors information collision, and fusion results should be the maximum bomber of possibility, And the possibility of airliner and fighter plane is very little.Carry out the result of target identification and compare such as table 2 using various combination rule It is shown.As seen from the results in Table 2：Dempster rules of combination can not be effectively treated to conflicting evidence, if there is evidence to certain The Basic Probability As-signment of one proposition is 0, and no matter follow-up fusion evidence supports much to the proposition, then fusion results are always 0. The method of document [1] is overly conservative, and with the increase of evidence number, the uncertainty of fusion results is also increasing, and does not meet not The purpose of certainty reasoning.The method convergence rate of document [2] is slow, when 4 evidence fusions are collected, still can not correctly do Go out decision-making.Document [3,4] although method can correctly make a policy when 3 evidences are collected, in 3 evidence fusions Result m (θ afterwards1)<0.5, m (θ1) and m (θ2) numerical value is closer to.The method of document [5] can be very in 3 evidence fusions Make a policy well, and the last fusion results proposition θ of the inventive method1Basic Probability As-signment m (θ1) maximum, reach 0.9231, and can preferably make a policy compared with the method for document [5] when 3 evidence fusions are collected, m (θ1) Numerical value reach 0.6503.The inventive method reduces the impact of interference evidence, and fast convergence rate reduces risk of policy making, carries The reliability of result when high conflicting evidence merges.
Table 2 is carried out the result of target identification and is compared using various combination rule
Each document for using in table 2 is as follows：
[1]Yager R R.On the Dempster-Shafer framework and new combination rules[J].Information Sciences,1987,41(2):93-137.
[2] Sun Quan, Ye Xiuqing, Gu Weikang. a kind of new composite formula [J] based on evidence theory. electronic letters, vol, 2000,28(8):117-119.
[3] Li Bicheng, Wang Bo, Wei Jun, etc. a kind of effective evidence theory composite formula [J]. data acquisition and procession, 2002,17(1):33-36.
[4] Quan Wen, Wang Xiaodan, Wang Jian, etc. a kind of DST rules of combination [J] distributed based on local conflicts. electronic letters, vol, 2012,40(9):1880-1884.
[5] Hu Changhua, department wins by a narrow margin, Zhou Zhijie, etc. the D-S innovatory algorithms [J] under new evidences conflict criterion. electricity Sub- journal, 2009,37 (7):1578-1583.

Claims (3)

1. based on the classification of evidence and the weighted evidence fusion method of measure method for conflict, it is characterised in that：Including following step：
A, the Basic Probability As-signment by obtaining the burnt unit of the corresponding evidence of multiple sensor measurement informations, each evidence is regarded as One vector, the vector m of i-th evidencei=(mi(θ1),…,mi(θr),…,mi(θk))TRepresent, wherein i=1, 2 ..., n, n are the sum of evidence vector, and k is Jiao unit number in framework of identification Θ, r=1,2 ..., k；
B, to above-mentioned i-th evidence miWith j-th evidence mjWhether it is that consistent evidence is judged：According to miIt is middle maximum basic The corresponding burnt unit of probability assignment and evidence mjWhether the corresponding burnt unit of middle maximum Basic Probability As-signment is same burnt unit judging, If same burnt unit, then claim evidence miAnd mjFor consistent evidence, otherwise claim evidence miAnd mjFor non-uniform evidence, wherein i, j=1, 2,…,n；i≠j；
C, by following formulaDifference property coefficient is calculated, any i-th evidence m is obtainediWith j-th card According to mjBetween otherness coefficient d (mi,mj), M in formulaiRepresent a row vector, DiRepresent a column vector；
D, by any evidence miAnd mjBetween otherness coefficient d (mi,mj) formula is passed through respectively according to classification of evidence result：WithCalculate any one Cause evidence miAnd mjWith non-uniform evidence miAnd mjBetween conflict coefficient conf (mi,mj)；
E, by any i-th evidence m for obtainingiWith j-th evidence mjBetween conflict coefficient conf (mi,mj) pass through formula：Try to achieve i-th evidence and other n-1 Total conflict spectrum factor conf (m of individual evidencei) and i-th evidence degree of support factor relative with other n-1 evidences truf(mi), and using relative degree of support factor t ruf maximum in n evidencemaxWith i evidence and other n-1 evidence Relative degree of support factor t ruf (mi) obtain weight coefficient ω by following formulai,
Jiao unit θ in F, i-th evidence of noterBasic Probability As-signment mi(θr) represent, wherein r=1,2 ..., k, revised the Jiao unit θ in i evidencerBasic Probability As-signment mi d(θr) represent, according to the weight coefficient ω obtained in step EiBy public affairs Formula：
Evidence to merging is modified；
It is G, last, revised evidence to be merged one by one using Dempster rules of combination, Jiao unit A's is substantially general after fusion The corresponding burnt unit of maximum of rate assignment m (A) is the corresponding identification target of the result of decision of target identification, and as decision-making most terminates Really.
2. according to claim 1 based on the classification of evidence and the weighted evidence fusion method of measure method for conflict, it is characterised in that： Row vector M described in described step Cr=[- mi(θr)mj(θ1),…,|mi(θr)-mj(θr)|,
…,-mi(θr)mj(θk)], column vector DrExpression formula isIts In, r=1,2 ..., k.
3. according to claim 1-2 based on the classification of evidence and the weighted evidence fusion method of measure method for conflict, its feature exists In：Described Dempster rules of combination are：
Wherein, m (A) represents the Basic Probability As-signment of Jiao unit A, and K is conflict coefficient, r, l=1,2 ..., k,For empty set.
CN201610984406.7A 2016-11-09 2016-11-09 Weighted evidence fusion method based on the classification of evidence and measure method for conflict Active CN106650785B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610984406.7A CN106650785B (en) 2016-11-09 2016-11-09 Weighted evidence fusion method based on the classification of evidence and measure method for conflict

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610984406.7A CN106650785B (en) 2016-11-09 2016-11-09 Weighted evidence fusion method based on the classification of evidence and measure method for conflict

Publications (2)

Publication Number Publication Date
CN106650785A true CN106650785A (en) 2017-05-10
CN106650785B CN106650785B (en) 2019-05-03

Family

ID=58806116

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610984406.7A Active CN106650785B (en) 2016-11-09 2016-11-09 Weighted evidence fusion method based on the classification of evidence and measure method for conflict

Country Status (1)

Country Link
CN (1) CN106650785B (en)

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107368854A (en) * 2017-07-20 2017-11-21 华北电力大学（保定） A kind of circuit breaker failure diagnostic method based on improvement evidence theory
CN108764312A (en) * 2018-05-17 2018-11-06 河海大学 Optimize multi objective dam defect image detecting method based on DS
CN108875823A (en) * 2018-06-12 2018-11-23 吉林大学 Combined method based on evidence under new evidence measurement standard
CN109061569A (en) * 2018-08-03 2018-12-21 中国人民解放军战略支援部队信息工程大学 A kind of object detection method and system of Spatial-temporal Information Fusion
CN109766933A (en) * 2018-12-26 2019-05-17 中国电子科技集团公司第二十研究所 A kind of multisource data fusion recognition methods based on evidence fuzzy factor
CN110033028A (en) * 2019-03-19 2019-07-19 河南大学 Conflicting evidence fusion method based on arithmetic average approach degree
CN110717504A (en) * 2019-06-28 2020-01-21 武汉大学 Pedestrian positioning method based on evidence theory fusion and multi-source sensor feature matching
CN110889227A (en) * 2019-11-28 2020-03-17 四川泛华航空仪表电器有限公司 Aircraft fuel oil measuring method based on multi-sensor information fusion
CN111079829A (en) * 2019-12-14 2020-04-28 长沙芯连心智慧系统有限责任公司 Multi-sensor self-adaptive fusion target identification method and system
CN111325277A (en) * 2020-02-26 2020-06-23 中国人民解放军军事科学院评估论证研究中心 Information fusion method based on negotiation strategy in target identification
CN111340118A (en) * 2020-02-27 2020-06-26 河南大学 Conflict evidence fusion method based on reliability entropy and BJS divergence
CN111368871A (en) * 2019-12-18 2020-07-03 中国人民解放军战略支援部队信息工程大学 Multi-feature information fusion method, device and system
CN111625775A (en) * 2020-05-28 2020-09-04 河南大学 Hellinger distance and reliability entropy based weighted conflict evidence fusion method
CN113487223A (en) * 2021-07-23 2021-10-08 深圳市城市公共安全技术研究院有限公司 Risk assessment method and risk assessment system based on information fusion
CN113989553A (en) * 2021-10-25 2022-01-28 西安交通大学 Evidence modeling method and system based on fuzzy sample density estimation and oriented to image mode classification

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101383090A (en) * 2008-10-24 2009-03-11 北京航空航天大学 Floating vehicle information processing method under parallel road network structure
CN101441718A (en) * 2008-12-19 2009-05-27 福建三元达通讯股份有限公司 Sensor information fuse device and method
CN102665294A (en) * 2012-04-25 2012-09-12 武汉大学 Vehicular sensor networks (VSN) event region detection method based on Dempster-Shafer (D-S) evidence theory
CN102721941A (en) * 2012-06-20 2012-10-10 北京航空航天大学 Method for fusing and diagnosing fault information of circuit of electric meter on basis of SOM (self-organized mapping) and D-S (Dempster-Shafer) theories
CN103810266A (en) * 2014-01-27 2014-05-21 中国电子科技集团公司第十研究所 Semantic network object identification and judgment method
CN104036257A (en) * 2014-06-25 2014-09-10 华南农业大学 Citrus vein phloem degeneration detection and classification method based on D-S theory through multi-source data fusion
US9141660B2 (en) * 2012-12-17 2015-09-22 International Business Machines Corporation Intelligent evidence classification and notification in a deep question answering system
CN105551031A (en) * 2015-12-10 2016-05-04 河海大学 Multi-temporal remote sensing image change detection method based on FCM and evidence theory

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101383090A (en) * 2008-10-24 2009-03-11 北京航空航天大学 Floating vehicle information processing method under parallel road network structure
CN101441718A (en) * 2008-12-19 2009-05-27 福建三元达通讯股份有限公司 Sensor information fuse device and method
CN102665294A (en) * 2012-04-25 2012-09-12 武汉大学 Vehicular sensor networks (VSN) event region detection method based on Dempster-Shafer (D-S) evidence theory
CN102721941A (en) * 2012-06-20 2012-10-10 北京航空航天大学 Method for fusing and diagnosing fault information of circuit of electric meter on basis of SOM (self-organized mapping) and D-S (Dempster-Shafer) theories
US9141660B2 (en) * 2012-12-17 2015-09-22 International Business Machines Corporation Intelligent evidence classification and notification in a deep question answering system
CN103810266A (en) * 2014-01-27 2014-05-21 中国电子科技集团公司第十研究所 Semantic network object identification and judgment method
CN104036257A (en) * 2014-06-25 2014-09-10 华南农业大学 Citrus vein phloem degeneration detection and classification method based on D-S theory through multi-source data fusion
CN105551031A (en) * 2015-12-10 2016-05-04 河海大学 Multi-temporal remote sensing image change detection method based on FCM and evidence theory

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
毕文豪等: "基于新的证据冲突衡量的加权证据融合方法", 《控制与决策》 *
蔚晓丹: "基于改进 D － S 信息融合方法的发动机故障诊断", 《公路交通科技》 *

Cited By (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107368854B (en) * 2017-07-20 2020-06-09 华北电力大学（保定） Breaker fault diagnosis method based on improved evidence theory
CN107368854A (en) * 2017-07-20 2017-11-21 华北电力大学（保定） A kind of circuit breaker failure diagnostic method based on improvement evidence theory
CN108764312A (en) * 2018-05-17 2018-11-06 河海大学 Optimize multi objective dam defect image detecting method based on DS
CN108875823A (en) * 2018-06-12 2018-11-23 吉林大学 Combined method based on evidence under new evidence measurement standard
CN108875823B (en) * 2018-06-12 2021-06-25 吉林大学 Evidence combination method based on evidence measurement standard
CN109061569A (en) * 2018-08-03 2018-12-21 中国人民解放军战略支援部队信息工程大学 A kind of object detection method and system of Spatial-temporal Information Fusion
CN109766933A (en) * 2018-12-26 2019-05-17 中国电子科技集团公司第二十研究所 A kind of multisource data fusion recognition methods based on evidence fuzzy factor
CN110033028A (en) * 2019-03-19 2019-07-19 河南大学 Conflicting evidence fusion method based on arithmetic average approach degree
CN110033028B (en) * 2019-03-19 2022-09-16 河南大学 Conflict evidence fusion method based on arithmetic mean closeness
CN110717504A (en) * 2019-06-28 2020-01-21 武汉大学 Pedestrian positioning method based on evidence theory fusion and multi-source sensor feature matching
CN110889227A (en) * 2019-11-28 2020-03-17 四川泛华航空仪表电器有限公司 Aircraft fuel oil measuring method based on multi-sensor information fusion
CN111079829A (en) * 2019-12-14 2020-04-28 长沙芯连心智慧系统有限责任公司 Multi-sensor self-adaptive fusion target identification method and system
CN111079829B (en) * 2019-12-14 2023-04-18 长沙芯连心智慧系统有限责任公司 Multi-sensor self-adaptive fusion target identification method and system
CN111368871A (en) * 2019-12-18 2020-07-03 中国人民解放军战略支援部队信息工程大学 Multi-feature information fusion method, device and system
CN111368871B (en) * 2019-12-18 2023-04-07 中国人民解放军战略支援部队信息工程大学 Multi-feature information fusion method, device and system
CN111325277A (en) * 2020-02-26 2020-06-23 中国人民解放军军事科学院评估论证研究中心 Information fusion method based on negotiation strategy in target identification
CN111340118A (en) * 2020-02-27 2020-06-26 河南大学 Conflict evidence fusion method based on reliability entropy and BJS divergence
CN111340118B (en) * 2020-02-27 2021-07-23 河南大学 Conflict evidence fusion method based on reliability entropy and BJS divergence
CN111625775A (en) * 2020-05-28 2020-09-04 河南大学 Hellinger distance and reliability entropy based weighted conflict evidence fusion method
CN111625775B (en) * 2020-05-28 2023-04-07 河南大学 Hellinger distance and reliability entropy based weighted conflict evidence fusion method
CN113487223A (en) * 2021-07-23 2021-10-08 深圳市城市公共安全技术研究院有限公司 Risk assessment method and risk assessment system based on information fusion
CN113487223B (en) * 2021-07-23 2024-04-30 深圳市城市公共安全技术研究院有限公司 Risk assessment method and system based on information fusion
CN113989553A (en) * 2021-10-25 2022-01-28 西安交通大学 Evidence modeling method and system based on fuzzy sample density estimation and oriented to image mode classification
CN113989553B (en) * 2021-10-25 2024-04-05 西安交通大学 Evidence modeling method and system based on mixed sample density estimation for image mode classification

Also Published As

Publication number Publication date
CN106650785B (en) 2019-05-03

Similar Documents

Publication Publication Date Title
CN106650785A (en) Weighted evidence fusion method based on evidence classification and conflict measurement
Li et al. An approach to fuzzy soft sets in decision making based on grey relational analysis and Dempster–Shafer theory of evidence: An application in medical diagnosis
CN106778847B (en) The fusion method of evidences conflict is measured based on logarithmic function
Dong et al. An analytic hierarchy process model of group consensus
CN104021392B (en) A kind of conflicting evidence fusion method based on vector metric
CN111950627B (en) Multi-source information fusion method and application thereof
CN108399340A (en) Based on the onboard networks safety risk estimating method for improving FAHP and cloud model
CN110033028A (en) Conflicting evidence fusion method based on arithmetic average approach degree
CN102393881B (en) A kind of high-precision detecting method of real-time many sensing temperatures data fusion
CN104376400A (en) Risk assessment method based on fuzzy matrix and analytic hierarchy process
CN103929330A (en) Domain name service quality evaluation method and system
CN107895311A (en) A kind of method and apparatus of order matching distribution
CN103473459A (en) Method of processing and fusing multisystem big data
CN105843829A (en) Big data credibility measurement method based on layering model
CN108573339A (en) A kind of consumer&#39;s net purchase methods of risk assessment of multi objective Project Decision Method
Luo et al. Diagnosis related group grouping study of senile cataract patients based on E-CHAID algorithm
CN108491719A (en) A kind of Android malware detection methods improving NB Algorithm
Thwing et al. A robust estimator of malaria incidence from routine health facility data
CN113295421B (en) Engine fault diagnosis method based on improved conflict coefficient and reliability entropy
CN107704869A (en) A kind of corpus data methods of sampling and model training method
Koc Role of national conditions in occupational fatal accidents in the construction industry using interpretable machine learning approach
CN110543417B (en) Method and device for evaluating effectiveness of software test
CN101527016A (en) Screening method of compliant conditions of entry and exit inspection and quarantine
AU2021101948A4 (en) A concrete durability detection method based on cloud model and D-S evidence theory
CN104462826A (en) Method for detecting and measuring multi-sensor evidence conflicts on basis of matrix singular value decomposition

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