CN110097100A - A kind of Bridge State Assessment method based on discrete dynamic Bayesian network - Google Patents
A kind of Bridge State Assessment method based on discrete dynamic Bayesian network Download PDFInfo
- Publication number
- CN110097100A CN110097100A CN201910319693.3A CN201910319693A CN110097100A CN 110097100 A CN110097100 A CN 110097100A CN 201910319693 A CN201910319693 A CN 201910319693A CN 110097100 A CN110097100 A CN 110097100A
- Authority
- CN
- China
- Prior art keywords
- discrete
- state
- bridge
- bayesian network
- dynamic bayesian
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/29—Graphical models, e.g. Bayesian networks
Abstract
The present invention relates to bridge structural state assessment technology fields, disclose a kind of Bridge State Assessment method based on discrete dynamic Bayesian network, solve the defect that existing highway bridge structure condition assessment method cannot be updated by detection information;Bridge State Assessment System Framework is established including the use of analytic hierarchy process (AHP), each component element in evaluation system is set as the node in discrete dynamic Bayesian network, and the discrete codomain of node is assigned according to the divided rank of configuration state, the dynamic Bayesian network model for being suitable for bridge structural state assessment is established in temporal sequence;The discrete state sample generated using analytic hierarchy process (AHP) is learnt, model parameter is obtained;The detection information that the present invention comprehensively utilizes specific structure actual conditions carrys out more new model, and the structure problem of bridge state complex relationship is handled by the dynamic bayesian network of bayesian theory, is effectively assessed bridge structural state.
Description
Technical field
The present invention relates to bridge structural state assessment technology fields, and in particular to one kind is based on discrete dynamic Bayesian network
Bridge State Assessment method more particularly to a kind of Discrete Dynamic Bayesian net using span concrete beam bridge as status assessment object
The Bridge State Assessment method of network.
Background technique
In recent years, researchers are from theoretical research, experimental study, field data, in bridge structural state assessment side
Face achieves many achievements, proposes many methods, such as: expert's Evaluation Method, gray system Evaluation Method, fuzzy theory Evaluation Method,
Analytic hierarchy process (AHP), Reliability assessment method etc..The above appraisal procedure common problem be no matter which kind of method or model all
Place hope on some set formula, cannot effectively combine it is practical, can not be updated by detection information existing method or
Model, therefore also can not just comprehensively utilize the inspection of the priori understanding and reflection specific structure actual conditions of structure degradation universal law
Measurement information.Dynamic bayesian network based on bayesian theory can handle time series data and the structure with complex relationship is asked
Topic, and can use detection information and timely update to model, it is applied in terms of bridge structural state assessment, is expected into
For a main trend of research and development in future.
Summary of the invention
The present invention overcomes the shortcomings of the prior art, proposes a kind of bridge state based on discrete dynamic Bayesian network
Appraisal procedure.
The present invention is achieved through the following technical solutions.
A kind of Bridge State Assessment method based on discrete dynamic Bayesian network, comprising the following steps:
1) it establishes bridge state level evaluation system: bridge is divided into a series of targets from up to down by layer of structure
Sublayer structure form, then the target sublayer structure is divided into several component elements.
2) node state divides: the component element being carried out state grade division, obtains series of discrete value composition
X∈{X1, X2..., Xi..., Xm, wherein discrete value XiThe state grade Classification Index of counterpart member element;According to each component
The state grade of element divides the discrete state for determining the node.
3) discrete dynamic Bayesian network model foundation: each component element in bridge state level evaluation system is set
The node being set in monolithic discrete dynamic Bayesian network, while assigning node discrete codomain, and according to each component element
Interlayer relation and upper and lower layer relationship, establish the directed acyclic graph of monolithic discrete dynamic Bayesian network.
4) model parameter determine: determine the parameter of discrete nodes in model, including in single timeslice discrete nodes it is initial
The transfer ginseng of the observation condition distribution function of discrete nodes and adjacent time piece discrete nodes in state distribution parameter, single timeslice
Number.
5) the initial state distribution parameter of discrete nodes in single timeslice parameter learning: is obtained by statistical method;Pass through
Analytic hierarchy process (AHP) obtains observation condition distribution parameter and adjacent time piece transfer parameters in single timeslice, then determines each parameter
Maximal possibility estimation, i.e., these parameters will make practice according to reach likelihood maximum.
Preferably, the bridge is large span concrete beam bridge.
Preferably, the directed acyclic graph for establishing monolithic discrete dynamic Bayesian network is by setting Bayes
Observer nodes group and concealed nodes group in network, then temporally axis extends by monolithic Bayesian model, according to time interval Δ t
It is divided, obtains complete discrete dynamic Bayesian network structural model.
Preferably, described that observation condition distribution parameter in single timeslice, including following step are obtained by analytic hierarchy process (AHP)
It is rapid:
A) according to the importance scale that variable compares two-by-two, Judgement Matricies.
B) algebraic mean value method is used, canonical normalized weight value is calculated.
C) score value and corresponding State Viewpoint measured value of observation node layer are obtained.
D) score value of destination layer and sub-goal layer is calculated, and one-to-one state value is obtained according to score value.
E) the discrete state sample of all variables of single timeslice is obtained.
F) the discrete state sample of multiple times is obtained according to the time interval point on time shaft.
G) sample is trained using the Parameter Learning Algorithm in dynamic bayesian network, obtains parameter.
The present invention is generated compared with the existing technology to be had the beneficial effect that.
The present invention is directed to the defect that existing highway bridge structure condition assessment method cannot be updated by detection information, proposes
Bridge State Assessment method based on discrete dynamic Bayesian network.Bridge State Assessment system frame is established using analytic hierarchy process (AHP)
Each component element in evaluation system is set as the node in discrete dynamic Bayesian network by frame, and according to configuration state
Divided rank assign the discrete codomain of node, in temporal sequence establish be suitable for bridge structural state assessment dynamic Bayesian networks
Network model.The discrete state sample generated using analytic hierarchy process (AHP) is learnt, model parameter is obtained.The present invention effectively combines
Practical, the detection information of comprehensive utilization specific structure actual conditions carrys out more new model, passes through the Dynamic Bayesian of bayesian theory
The structure problem of network processes bridge state complex relationship, and can use detection information and timely update to model, effectively
Bridge structural state is assessed.
Detailed description of the invention
Fig. 1 is the large span concrete beam bridge state level evaluation system schematic diagram established in embodiment.
Fig. 2 is the dynamic bayesian network structural model figure for the large span concrete beam bridge status assessment established in embodiment.
Specific embodiment
In order to which technical problems, technical solutions and advantages to be solved are more clearly understood, in conjunction with reality
Example and attached drawing are applied, the present invention will be described in further detail.It should be appreciated that specific embodiment described herein is only used to
It explains the present invention, is not intended to limit the present invention.Below with reference to the examples and drawings technical solution that the present invention will be described in detail, but
The scope of protection is not limited by this.
It is using certain large span concrete beam bridge is below research object using the bridge state based on discrete dynamic Bayesian network
Appraisal procedure.
Step 1: bridge state level evaluation system
Bridge State Assessment system is tentatively established using analytic hierarchy process (AHP), by bridge system by distinguishing hierarchy be it is a series of from
It is upper that downward destination layer structure --- sublayer structure form establishes entire evaluation system on this basis.As shown in Figure 1, mixed
Solidifying soil cable-stayed bridge status assessment system W, can be divided into girder S by primary structure level1, bridge tower S2, box beam S3, substructure S4, it is attached
Belong to facility S5, these structures are divided into several component elements i.e. observation layer M again1——M21, entire large span concrete beam bridge state level
Evaluation system.
Step 2: node state divides
Component element in hierarchy system corresponds to the node in discrete dynamic Bayesian network, therefore these nodes
Discrete state depend on each component element state grade divide.The codomain of each variable node, will by it is a series of from
Dissipate value composition X ∈ { X1, X2..., Xi..., Xm, wherein discrete value XiThe state grade Classification Index of counterpart member element.According to
The scoring of " highway maintenance technical specification " (JTGH11-2004), and comprehensively consider the computational efficiency of Bayesian network, by shape
State grade classification is 7, as shown in table 1 below.
The bridge structural state grade classification distribution map that table 1 is established
Step 3: discrete dynamic Bayesian network model foundation
Each component element in large span concrete beam bridge state level evaluation system is set as monolithic Discrete Dynamic pattra leaves
Node in this network, while assigning node discrete codomain, and closed according to the interlayer relation of each component element and upper and lower layer
System, establishes the directed acyclic graph of monolithic discrete dynamic Bayesian network, wherein M label layer element in Fig. 1 is set as Bayes
Observer nodes group in network, W and S label layer element are set as concealed nodes group.Then temporally by monolithic Bayesian model
Axis extends, and is divided according to certain time interval Δ t, as shown in Fig. 2, obtaining complete discrete dynamic Bayesian network knot
Structure model, every lateral arrows indicate time interval in Fig. 2.
Step 4: model parameter determines
A) discrete dynamic Bayesian network parameter
After establishing the discrete dynamic Bayesian network structural model of large span concrete beam bridge status assessment, it is thus necessary to determine that discrete in model
The parameter of node.Parameter in Fig. 2 is as follows:
The initial state distribution parameter of discrete nodes in single timeslice
Wherein,It respectively indicates target node layer and is in stateProbability, sub-goal layer section
Point is in stateProbability, observation node layer be inProbability.I indicates the specific element of present node, and t is indicated
Current time, the observation condition distribution function of discrete nodes in single timeslice
Wherein,Known to expressionMoment nodeIn stateWhen,Moment nodePlace
InConditional probability;Meaning is similar.
The transfer parameters of adjacent time piece discrete nodes
Wherein,Known to expressionMoment nodeIn stateWhen, known to expression
Moment nodeIn stateWhen conditional probability;Meaning is similar.
B) parameter learning
The initial state distribution parameter of discrete nodes can be obtained by statistical method in single timeslice, it is critical that in single timeslice
Observation condition distribution parameter and adjacent time piece transfer parameters determination.It is seen so we utilize analytic hierarchy process (AHP) to determine again
Survey condition distribution parameter, steps are as follows:
1) expert opinion is utilized, obtains the importance scale that variable two-by-two compares, then Judgement Matricies;
2) algebraic mean value method is used, canonical normalized weight value is calculated:;
3) score value and corresponding State Viewpoint measured value of observation node layer are obtained;
4) score value of destination layer and sub-goal layer is calculated:
And one-to-one state value is obtained according to score value
5) the discrete state sample of all variables of single timeslice is obtained;
6) the discrete state sample of multiple times is obtained according to the time interval point on time shaft;
7) sample is trained using the Parameter Learning Algorithm in dynamic bayesian network, obtains parameter.
The purpose of parameter learning is to find the maximal possibility estimation of each parameter, i.e., these parameters will make white silk evidence reach likelihood most
Greatly.
The above content is combine specific preferred embodiment to the further description done of the present invention, and it cannot be said that
A specific embodiment of the invention is only limitted to this, for those of ordinary skill in the art to which the present invention belongs, is not taking off
Under the premise of from the present invention, several simple deduction or replace can also be made, all shall be regarded as belonging to the present invention by being submitted
Claims determine scope of patent protection.
Claims (4)
1. a kind of Bridge State Assessment method based on discrete dynamic Bayesian network, which comprises the following steps:
1) it establishes bridge state level evaluation system: bridge is divided into a series of target sublayers from up to down by layer of structure
Structure type, then the target sublayer structure is divided into several component elements;
2) node state divides: the component element being carried out state grade division, obtains series of discrete value composition, wherein discrete valueThe state grade Classification Index of counterpart member element;According to each component member
The state grade of element divides the discrete state for determining the node;
3) discrete dynamic Bayesian network model foundation: each component element in bridge state level evaluation system is set as
Node in monolithic discrete dynamic Bayesian network, while assigning node discrete codomain, and according to the interlayer of each component element
Relationship and upper and lower layer relationship, establish the directed acyclic graph of monolithic discrete dynamic Bayesian network;
4) model parameter determines: determining the parameter of discrete nodes in model, the original state including discrete nodes in single timeslice
The observation condition distribution function of discrete nodes and the transfer parameters of adjacent time piece discrete nodes in distribution parameter, single timeslice;
5) the initial state distribution parameter of discrete nodes in single timeslice parameter learning: is obtained by statistical method;Pass through level
Analytic approach obtains observation condition distribution parameter and adjacent time piece transfer parameters in single timeslice, then determines each parameter most
Maximum-likelihood estimation, i.e. these parameters will make to practice maximum according to likelihood is reached.
2. a kind of Bridge State Assessment method based on discrete dynamic Bayesian network according to claim 1, feature
It is, the bridge is large span concrete beam bridge.
3. a kind of Bridge State Assessment method based on discrete dynamic Bayesian network according to claim 1, feature
It is, the directed acyclic graph for establishing monolithic discrete dynamic Bayesian network, is by the sight in setting Bayesian network
Node group and concealed nodes group are surveyed, then temporally axis extends by monolithic Bayesian model, according to time intervalIt is divided, is obtained
To complete discrete dynamic Bayesian network structural model.
4. a kind of Bridge State Assessment method based on discrete dynamic Bayesian network according to claim 1, feature
It is, the observation condition distribution parameter obtained by analytic hierarchy process (AHP) in single timeslice, comprising the following steps:
A) according to the importance scale that variable compares two-by-two, Judgement Matricies;
B) algebraic mean value method is used, canonical normalized weight value is calculated;
C) score value and corresponding State Viewpoint measured value of observation node layer are obtained;
D) score value of destination layer and sub-goal layer is calculated, and one-to-one state value is obtained according to score value;
E) the discrete state sample of all variables of single timeslice is obtained;
F) the discrete state sample of multiple times is obtained according to the time interval point on time shaft;
G) sample is trained using the Parameter Learning Algorithm in dynamic bayesian network, obtains parameter.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910319693.3A CN110097100A (en) | 2019-04-19 | 2019-04-19 | A kind of Bridge State Assessment method based on discrete dynamic Bayesian network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910319693.3A CN110097100A (en) | 2019-04-19 | 2019-04-19 | A kind of Bridge State Assessment method based on discrete dynamic Bayesian network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110097100A true CN110097100A (en) | 2019-08-06 |
Family
ID=67445341
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910319693.3A Pending CN110097100A (en) | 2019-04-19 | 2019-04-19 | A kind of Bridge State Assessment method based on discrete dynamic Bayesian network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110097100A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112802021A (en) * | 2021-04-09 | 2021-05-14 | 泰瑞数创科技(北京)有限公司 | Urban bridge road diagnosis method and system based on digital twin technology |
CN112818455A (en) * | 2021-02-22 | 2021-05-18 | 深圳市市政设计研究院有限公司 | Bridge structure response monitoring method and system |
CN113094843A (en) * | 2021-04-30 | 2021-07-09 | 哈尔滨工业大学 | Solving method for conditional probability of beam bridge evaluation based on Bayesian network |
CN113742195A (en) * | 2021-09-18 | 2021-12-03 | 北京航空航天大学 | Bayesian neural network-based system health state prediction method |
CN114021314A (en) * | 2021-10-21 | 2022-02-08 | 北京航空航天大学杭州创新研究院 | System electromagnetic vulnerability evaluation method based on dynamic Bayesian network |
CN114329730A (en) * | 2022-01-06 | 2022-04-12 | 福州大学 | Cable-stayed bridge safety evaluation method based on variable-structure dynamic mixed Bayesian network |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20130027717A (en) * | 2011-09-08 | 2013-03-18 | 한양대학교 산학협력단 | Computerized adaptive testing system and method using bayesian networks |
CN103793853A (en) * | 2014-01-21 | 2014-05-14 | 中国南方电网有限责任公司超高压输电公司检修试验中心 | Overhead power transmission line running state assessment method based on bidirectional Bayesian network |
CN105426970A (en) * | 2015-11-17 | 2016-03-23 | 武汉理工大学 | Meteorological threat assessment method based on discrete dynamic Bayesian network |
CN107016464A (en) * | 2017-04-10 | 2017-08-04 | 中国电子科技集团公司第五十四研究所 | Threat estimating method based on dynamic bayesian network |
-
2019
- 2019-04-19 CN CN201910319693.3A patent/CN110097100A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20130027717A (en) * | 2011-09-08 | 2013-03-18 | 한양대학교 산학협력단 | Computerized adaptive testing system and method using bayesian networks |
CN103793853A (en) * | 2014-01-21 | 2014-05-14 | 中国南方电网有限责任公司超高压输电公司检修试验中心 | Overhead power transmission line running state assessment method based on bidirectional Bayesian network |
CN105426970A (en) * | 2015-11-17 | 2016-03-23 | 武汉理工大学 | Meteorological threat assessment method based on discrete dynamic Bayesian network |
CN107016464A (en) * | 2017-04-10 | 2017-08-04 | 中国电子科技集团公司第五十四研究所 | Threat estimating method based on dynamic bayesian network |
Non-Patent Citations (1)
Title |
---|
贾布裕 等: "基于离散动态贝叶斯网络的桥梁状态评估方法", 《桥梁建设》 * |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112818455A (en) * | 2021-02-22 | 2021-05-18 | 深圳市市政设计研究院有限公司 | Bridge structure response monitoring method and system |
CN112802021A (en) * | 2021-04-09 | 2021-05-14 | 泰瑞数创科技(北京)有限公司 | Urban bridge road diagnosis method and system based on digital twin technology |
CN112802021B (en) * | 2021-04-09 | 2021-07-30 | 泰瑞数创科技(北京)有限公司 | Urban bridge road diagnosis method and system based on digital twin technology |
CN113094843A (en) * | 2021-04-30 | 2021-07-09 | 哈尔滨工业大学 | Solving method for conditional probability of beam bridge evaluation based on Bayesian network |
CN113742195A (en) * | 2021-09-18 | 2021-12-03 | 北京航空航天大学 | Bayesian neural network-based system health state prediction method |
CN114021314A (en) * | 2021-10-21 | 2022-02-08 | 北京航空航天大学杭州创新研究院 | System electromagnetic vulnerability evaluation method based on dynamic Bayesian network |
CN114021314B (en) * | 2021-10-21 | 2024-05-03 | 北京航空航天大学杭州创新研究院 | System electromagnetic vulnerability assessment method based on dynamic Bayesian network |
CN114329730A (en) * | 2022-01-06 | 2022-04-12 | 福州大学 | Cable-stayed bridge safety evaluation method based on variable-structure dynamic mixed Bayesian network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110097100A (en) | A kind of Bridge State Assessment method based on discrete dynamic Bayesian network | |
CN105631483B (en) | A kind of short-term electro-load forecast method and device | |
Conde Rocha Murca et al. | Trajectory clustering and classification for characterization of air traffic flows | |
CN105117602A (en) | Metering apparatus operation state early warning method | |
CN110533331B (en) | Fault early warning method and system based on transmission line data mining | |
CN110210648B (en) | Gray long-short term memory network-based control airspace strategic flow prediction method | |
CN104200288A (en) | Equipment fault prediction method based on factor-event correlation recognition | |
CN110111576A (en) | A kind of urban transportation elastic index and its implementation based on space-time congestion group | |
CN107273273A (en) | A kind of distributed type assemblies hardware fault method for early warning and system | |
CN110378549B (en) | Transmission tower bird damage grade assessment method based on FAHP-entropy weight method | |
JP6185206B1 (en) | Method and apparatus for detecting abnormalities or signs of abnormalities in a photovoltaic system | |
CN104601604A (en) | Network security situation analyzing method | |
CN105912857B (en) | Matching method of power distribution equipment state monitoring sensors | |
CN109246495A (en) | A kind of optical network service method for evaluating quality of oriented multilayer, multi objective | |
Arif et al. | Distribution network outage data analysis and repair time prediction using deep learning | |
CN110889587B (en) | Power distribution network line risk assessment method | |
CN114912635A (en) | Highway maintenance decision-making method and system, storage medium and electronic equipment | |
CN115766501B (en) | Tunnel construction data management system and method based on big data | |
CN107767047A (en) | A kind of power distribution network regulates and controls integral system operation conditions evaluation method | |
CN110059913A (en) | A kind of quantitative estimation method counted and the power failure of future-state is planned | |
CN116090821A (en) | Power distribution network line security risk assessment method considering extreme weather | |
Angalakudati et al. | Improving emergency storm planning using machine learning | |
CN109902344A (en) | Short/Medium Span Bridge group structure performance prediction apparatus and system | |
CN109101659A (en) | A kind of method of power data exception in small power station's data collection system | |
CN107292431A (en) | Power telecom network service reliability Forecasting Methodology based on dynamic bayesian network |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190806 |