CN107871161A - A kind of bridge entirety damage alarming method based on neutral net - Google Patents
A kind of bridge entirety damage alarming method based on neutral net Download PDFInfo
- Publication number
- CN107871161A CN107871161A CN201711085667.6A CN201711085667A CN107871161A CN 107871161 A CN107871161 A CN 107871161A CN 201711085667 A CN201711085667 A CN 201711085667A CN 107871161 A CN107871161 A CN 107871161A
- Authority
- CN
- China
- Prior art keywords
- bridge
- vibration
- neural network
- frequency
- natural frequency
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Abstract
The invention discloses a kind of bridge entirety damage alarming method based on neutral net, including:Gather the Monitoring Data under health status;Analyze the bridge structure natural frequency of vibration;Analyze several influence factors being had an impact to the natural frequency of vibration;Training sample database is established, using BP neural network algorithm, the BP neural network model established between all influence factors and the natural frequency of vibration, and BP neural network model is tested;The Monitoring Data of bridge under the present situation is gathered, tries to achieve the natural frequency of vibration, and as test samples;The self excited vibrational frequency of bridge span under current influence factor is simulated, draws fitting frequency values;Healthy sample and test samples are established, and healthy sample and test samples are subjected to hypothesis testing.Beneficial effects of the present invention:According to the health monitoring data of bridge, self excited vibrational frequency of bridge span is analyzed, and as damage alarming index, establish perfect damage alarming flow, and combine neutral net, improve the applicability and stability of this method.
Description
Technical field
The present invention relates to railroad bridge technical field, is integrally damaged in particular to a kind of bridge based on neutral net
Hinder method for early warning.
Background technology
Bridge plays vital effect in railway operation, and the growth of the active time with railroad bridge, by
Necessarily damaging more or less can all occur in the influence of external environment and structure in itself, bridge, and when these damages reach certain journey
When spending, it will jeopardize traffic safety.All it is to be rung by the certain sensor of the proper site installation on bridge, collection bridge at present
Should, and Treatment Analysis is carried out to the data that these are collected, extraction can characterize the characteristic parameter of bridge structure, these parameters are entered
Row analysis, observation Bridge performance change, so as to realize damage alarming.
It is existing to be come with some shortcomings for Damage Alarming of Bridge Structures and theory, it is embodied in following a few sides
Face:
1st, the mode of generally use is to carry out threshold value early warning, when bridge monitoring value thinks bridge structure more than a threshold value
There is damage, fail fully to excavate the value of long term monitoring data, the bridge structure that analyzes that can not be quantified is damaged
Degree, cause the waste of a large amount of Monitoring Datas;
2nd, different bridge structure types may differ to the sensitivity of damage criterion, and this, which is required, targetedly selects
Warning index is taken, but existing damage alarming index can not reflect the actual damage situation of bridge completely;
3rd, precision is low, stability is poor, applicability is not strong.
The content of the invention
To solve the above problems, it is an object of the invention to provide a kind of bridge entirety damage alarming based on neutral net
Method, according to the health monitoring data of bridge, self excited vibrational frequency of bridge span is analyzed, and as damage alarming index, establish perfect
Damage alarming flow, and combine neutral net, improve the applicability and stability of this method.
The invention provides a kind of bridge entirety damage alarming method based on neutral net, including:
Step 1, Monitoring Data of the bridge within a period of time under health status is gathered;
Step 2, by the natural frequency of vibration of the Analysis on monitoring data bridge structure collected;
Step 3, according to bridge site environment, analyze several that had an impact on the natural frequency of vibration in step 2 influence because
Element;
Step 4, using the influence factor in the natural frequency of vibration and step 3 in step 2 as training sample, training sample is established
Database, using BP neural network algorithm, the BP neural network model established between these influence factors and the natural frequency of vibration, and it is right
BP neural network model is trained, tested;
Step 5, the Monitoring Data of bridge under the present situation is gathered, tries to achieve the natural frequency of vibration, and using the natural frequency of vibration as inspection
Test sample;
Step 6, according to the BP neural network model established in step 4, the bridge self-vibration under current influence factor is simulated
Frequency, draw fitting frequency values fs, as healthy sample;
Step 7, healthy sample and test samples are subjected to both hypothesis testings, observation and whether there is significant difference:
If there is significant difference, then it is assumed that current bridge structure is damaged, and carries out damage alarming;
If there is no significant difference, then it is assumed that current bridge structure is in a safe condition, and by current inspection
Sample is extended in training sample database, and the BP neural network model established in step 4 is trained again.
As further improvement of the invention, step 4 specifically includes:
Step 401, the topological structure of BP neural network model is determined:
Determine the neuron that input layer, output layer, the hidden layer number of plies and each hidden layer of BP neural network model are included
Number;
Step 402, from input quantity of the influence factor as BP neural network model, using the natural frequency of vibration as output quantity,
Establish training sample database;
Step 403, using most of data in training sample database as training sample, using remaining sub-fraction as
Test sample, BP neural network is trained using conjugate gradient method and BP neural network model carried out using test sample
Test.
As further improvement of the invention, the hidden layer number of plies, the neuron that hidden layer is included are determined using empirical method
Number using object function determine, wherein, object function is defined as the error sum of squares between output valve and desired value, when the mesh
When scalar functions reach minimum, the neuron number that hidden layer is included is optimal.
As further improvement of the invention, step 7 specifically includes:
Step 701, the fitting frequency values f drawn up with the BP neural network pattern die trainedsAs healthy sample, health
The average of sample is μs;
Step 702, with the practical frequency value f of self excited vibrational frequency of bridge span under current influence factormAs test samples, sample is examined
This average is μm;
Step 703, hypothesis testing is carried out to healthy sample and test samples, null hypothesis is:
H0:μs=μm
H1:μs≠μm
If receive it is assumed that i.e. H0Set up, now μs=μm, then healthy sample and test samples do not have notable difference, i.e. bridge
Girder construction is without obvious damage, without carrying out the overall damage alarming of bridge structure;
If refusal is it is assumed that i.e. H1Set up, now μs≠μm, then healthy sample and test samples have notable difference, i.e. bridge
Structure, which has, substantially to be damaged, it is necessary to carry out the overall damage alarming of bridge structure;
Step 704, if bridge damages without obvious, test samples are extended in training sample database, and again
Neural network model is trained.
As further improvement of the invention, frequency values f will be fittedsWith practical frequency value fmCarry out hypothesis testing, observation two
Person whether there is significant difference, judge whether bridge structure occurs substantially damaging with this.
Beneficial effects of the present invention are:
1st, limited bridge monitoring data are make use of, solves and asking for non-destructive tests and early warning is carried out according to measured data
Topic, based on measured data, is analyzed bridge integral status, and when bridge damages, damage alarming can be achieved;
2nd, based on using the natural frequency of vibration of bridge structure, and neutral net structure warning index is combined, finally using phase
The damage alarming method answered, realizes the early warning integrally damaged to bridge;
3rd, from bridge structure, the parameter of structure own characteristic is characterized as non-destructive tests and early warning using bridge
Index can obtain more preferable effect, can be widely used in the non-destructive tests of bridge structure, have larger development prospect;
4th, can not tested person environment influence, employ in statistical analysis hypothesis testing and carry out damage alarming, to bridge
Frequency change has stronger sensitiveness and applicability, and the precision height of identification, stability are good, strong applicability.
Brief description of the drawings
Fig. 1 is that a kind of flow of bridge entirety damage alarming method based on neutral net described in the embodiment of the present invention is shown
It is intended to;
Fig. 2 is influence schematic diagram of the temperature to the single order natural frequency of vibration;
Fig. 3 is influence schematic diagram of the temperature to the five rank natural frequencies of vibration;
Fig. 4 is influence schematic diagram of the vehicular load to the single order natural frequency of vibration;
Fig. 5 is influence schematic diagram of the vehicular load to the five rank natural frequencies of vibration;
Fig. 6 is the object function schematic diagram of the single order natural frequency of vibration;
Fig. 7 is the object function schematic diagram of the five rank natural frequencies of vibration;
Fig. 8 is the trained values and match value schematic diagram of the single order natural frequency of vibration;
Fig. 9 is the trained values and match value schematic diagram of the five rank natural frequencies of vibration;
Figure 10 is the test value and match value schematic diagram of the single order natural frequency of vibration;
Figure 11 is the test value and match value schematic diagram of the five rank natural frequencies of vibration;
Figure 12 is the measured value of the single order natural frequency of vibration and the match value schematic diagram of neutral net under current influence factor;
Figure 13 is the measured value of the five rank natural frequencies of vibration and the match value schematic diagram of neutral net under current influence factor;
Figure 14 is the P value changes schematic diagrames of the single order natural frequency of vibration;
Figure 15 is the P value changes schematic diagrames of the five rank natural frequencies of vibration.
Embodiment
The present invention is described in further detail below by specific embodiment and with reference to accompanying drawing.
A kind of bridge entirety damage alarming method based on neutral net described in the embodiment of the present invention, this method is with bridge
Based on the natural frequency of vibration of structure, using neural net method, other disturbing factors such as environmental factor and the natural frequency of vibration are established
Between correlation, damage alarming is finally realized according to hypothesis testing.As shown in figure 1, this method includes:
Step 1, Monitoring Data of the bridge within a period of time under health status is gathered.
The present embodiment runs the Monitoring Data in 1 year initial stage as healthy number by taking certain bridge as an example after the bridge is built up
According to thinking that the bridge does not damage in this period, bridge structure is in health status.
Step 2, by the natural frequency of vibration of the Analysis on monitoring data bridge structure collected.
By analyzing above-mentioned Monitoring Data, the natural frequency of vibration of bridge structure can be obtained, due to from measured data very
Difficulty analyzes the high order of frequency of bridge, carries out analytic explanation by taking fundamental frequency and five order frequencies as an example herein.The natural frequency of vibration is main
Tried to achieve by carrying out analysis to the acceleration information collected, the frequency tried to achieve mainly has carrier frequency rate.Wherein, use is limited
The self excited vibrational frequency of bridge span and the natural frequency of vibration of actual measurement that member calculates are as shown in table 1 below, during analysis mainly based on practical frequency.Table 1
The middle natural frequency of vibration value by theoretical calculation and the natural frequency of vibration value of actual measurement are abbreviated as calculated value and measured value respectively.
The natural frequency of vibration theoretical value of table 1 and measured value
Step 3, according to bridge site environment, analyze several that had an impact on the natural frequency of vibration in step 2 influence because
Element.
The bridge of the present embodiment is because across great river, present position year temperature Change is larger, and through domestic and international result of study table
It is bright:Temperature has a great influence to the bridge structure natural frequency of vibration, or even can cover influence of the damage to self excited vibrational frequency of bridge span.Therefore, will
Temperature is as the first influence factor caused by the natural frequency of vibration.And the natural frequency of vibration of bridge actual measurement adding when mainly being passed a bridge by vehicle
Speed responsive is tried to achieve, and gained frequency is has carrier frequency rate, therefore, using vehicular load be used as the second influence caused by the natural frequency of vibration because
Element.Wherein, as shown in Figures 2 and 3, vehicular load is to one for influence difference of the temperature to the single order natural frequency of vibration and the five rank natural frequencies of vibration
The influences of the rank natural frequency of vibration and the five rank natural frequencies of vibration as shown in Figure 4 and Figure 5, wherein, train load can use acceleration information
Virtual value represents.
Step 4, using the influence factor in the natural frequency of vibration and step 3 in step 2 as training sample, training sample is established
Database, using BP neural network algorithm, the BP neural network model established between these influence factors and the natural frequency of vibration, and it is right
BP neural network model is trained test.
Specifically include:
Step 401, the topological structure of BP neural network model is determined:Determine input layer, the output of BP neural network model
The neuron number that layer, the hidden layer number of plies and each hidden layer are included.
The hidden layer number of plies is determined using empirical method, the neuron number that hidden layer is included is determined using object function, its
In, object function is defined as output valve fmWith desired value fsBetween error sum of squares, it is hidden when the object function reaches minimum
The neuron number included containing layer is optimal.Object function mse functional expression is as follows:
In above formula:N is sample length, fmFor practical frequency value, fsFor the fitting frequency values of BP neural network.
In the present embodiment, neutral net uses three-layer network, and hidden layer is one layer, using sigmoid as transmission function, is adopted
It is trained with conjugate gradient method, it is 1 to 10 to take hidden layer number of unit respectively, the obtained single order natural frequency of vibration and five ranks
Respectively as shown in Figure 6 and Figure 7, when neuron number is respectively 6 and 5, object function reaches for the object function change of the natural frequency of vibration
To minimum, then the neuron number that hidden layer is included is according to respectively 6 and 5.
Step 402, from input quantity of the influence factor as BP neural network model, using the natural frequency of vibration as output quantity,
Establish training sample database.
Step 403, using most of data in training sample database as training sample, using remaining sub-fraction as
Test sample, BP neural network is trained using conjugate gradient method and BP neural network model carried out using test sample
Test.
Conjugate gradient method its to solve performance be a method between steepest descent method and Newton method, it is only necessary to using leading
Number information, the characteristics of steepest descent method convergence is slow is overcome, it is most useful that conjugate gradient method is not only solution large linear systems
One of method, and solve one of maximally effective algorithm of nonlinear optimization, there is the characteristics of convergence is fast, and stability is high.
Have in the present embodiment and choose 660 groups of natural frequencies of vibration and corresponding influence factor at intervals, build sample data
Storehouse, wherein choosing 500 groups of data as training sample, remainder data is as test sample.Using above-mentioned 500 groups of number of training
According to neural metwork training is carried out, its generalization ability, the i.e. fitting to initial data and the extrapolability to test data are examined.
Wherein shown in capability of fitting below figure 8 and Fig. 9, it can be seen that match value is consistent with trained values variation tendency, to trained values and plan
Conjunction value carries out statistical analysis, and shown in statistical result table 2, showing the BP network models has good capability of fitting.Extrapolation
Ability is as shown in Figure 10 and Figure 11, it can be seen that test value and match value variation tendency are basically identical, and carry out statistical to it
Analysis, statistical result is as shown in table 3, and showing the BP network models has good extrapolability.Therefore, can from Fig. 8-Figure 11
Seeing the neutral net of foundation has good generalization ability.
The training sample of table 2 and fitting sample statistics
The test sample of table 3 and fitting sample statistics
Step 5, the Monitoring Data of bridge under the present situation is gathered, tries to achieve the natural frequency of vibration, and using the natural frequency of vibration as inspection
Test sample.
Step 6, according to the BP neural network model established in step 4, the bridge self-vibration under current influence factor is simulated
Frequency, draw fitting frequency values fs, as healthy sample.
The natural frequency of vibration value that the present embodiment chooses the actual measurement in a period of time is tested, sample number 100, according to the section
Influence factor measured value in time, using the neutral net established, obtain current fitting frequency values fs, Figure 12 is current
The fitting frequency values (i.e. match value) of the practical frequency value (i.e. measured value) of the single order natural frequency of vibration and neutral net under influence factor,
Figure 13 is the fitting frequency values of the practical frequency value (i.e. measured value) of the five rank natural frequencies of vibration and neutral net under current influence factor
(i.e. match value).
Step 7, healthy sample and test samples are subjected to both hypothesis testings, observation and whether there is significant difference:
If there is significant difference, then it is assumed that current bridge structure is damaged, and carries out damage alarming;
If there is no significant difference, then it is assumed that current bridge structure is in a safe condition, and by current inspection
Sample is extended in training sample database, and the BP neural network model established in step 4 is trained again.
Specifically include:
Step 701, the fitting frequency values f drawn up with the BP neural network pattern die trainedsAs healthy sample, health
The average of sample is μs;
Step 702, with the practical frequency value f of self excited vibrational frequency of bridge span under current influence factormAs test samples, sample is examined
This average is μm;
Step 703, hypothesis testing is carried out to healthy sample and test samples, null hypothesis is:
H0:μs=μm
H1:μs≠μm
If receive it is assumed that i.e. H0Set up, now μs=μm, then healthy sample and test samples do not have notable difference, i.e. bridge
Girder construction is without obvious damage, without carrying out the overall damage alarming of bridge structure;
If refusal is it is assumed that i.e. H1Set up, now μs≠μm, then healthy sample and test samples have notable difference, i.e. bridge
Structure, which has, substantially to be damaged, it is necessary to carry out the overall damage alarming of bridge structure;
Step 704, if bridge damages without obvious, test samples are extended in training sample database, and again
Neural network model is trained.
The present embodiment carries out hypothesis testing to healthy sample and test samples, it is assumed that the result of inspection can be examined by observing
In P values determine, wherein, P values refer to the observation result resulting when being assumed to be true or more extreme result occur it is general
Rate, the probability in the section as calculated when being assumed to be true, when too small (taking 0.01) the can refusal of P values is former false
It is as a result 0 when receiving null hypothesis if assay 1.As a result it is as shown in table 4.As a result show:Receive null hypothesis, i.e., two
Significant difference is not present in group sample, then it is assumed that bridge structure is in a safe condition.Practical frequency value and neutral net in table 4
Fitting frequency values are abbreviated as measured value, network match value respectively.
The fitting frequency values sample parameter of the practical frequency value of table 4 and neutral net
In order to verify application of this method to bridge damnification early warning, measured data is handled respectively, simulation damage
Generation.Due to the general reduction for all showing as rigidity of the generation of Bridge Structural Damage, therefore, actual measurement that can be to the natural frequency of vibration is frequently
Rate value is adjusted as follows:
F'=(1- α) fm
In formula, f' is that the frequency values after lesion mimic, f are carried out to practical frequency valuemFor the practical frequency value of the natural frequency of vibration,
α is frequency reduction rate, take herein scope be 0.001~0.01 between.
Using the practical frequency value f ' after damage as test samples, frequency values f will be fittedsAs healthy sample, Liang Zhejin
The P value changes difference of row hypothesis testing, the single order natural frequency of vibration and the five rank natural frequencies of vibration is as shown in FIG. 14 and 15.As a result show:It is right
In the single order natural frequency of vibration, when frequency reduction rate is more than 0.2%, P values are less than 0.01, it is believed that two groups of sample datas exist notable
Sex differernce, i.e. bridge damage;For the five rank natural frequencies of vibration, when frequency reduction rate is more than 0.4%, P values are less than 0.01, i.e.,
It is believed that bridge damages;As can be seen that change of this method to self excited vibrational frequency of bridge span has stronger sensitiveness.
The present invention is using bridge structure natural frequency of vibration progress non-destructive tests, can be with using neural network correlation model
The not influence of tested person environment, and according to existing health monitoring data, employ hypothesis testing in statistical analysis and damaged
Early warning, there is stronger sensitiveness and applicability to the change of bridge frequency.
The preferred embodiments of the present invention are the foregoing is only, are not intended to limit the invention, for the skill of this area
For art personnel, the present invention can have various modifications and variations.Within the spirit and principles of the invention, that is made any repaiies
Change, equivalent substitution, improvement etc., should be included in the scope of the protection.
Claims (5)
- A kind of 1. bridge entirety damage alarming method based on neutral net, it is characterised in that this method includes:Step 1, Monitoring Data of the bridge within a period of time under health status is gathered;Step 2, by the natural frequency of vibration of the Analysis on monitoring data bridge structure collected;Step 3, according to bridge site environment, several influence factors being had an impact to the natural frequency of vibration in step 2 are analyzed;Step 4, using the influence factor in the natural frequency of vibration and step 3 in step 2 as training sample, training sample data are established Storehouse, using BP neural network algorithm, the BP neural network model established between these influence factors and the natural frequency of vibration, and to BP god It is trained, tests through network model;Step 5, the Monitoring Data of bridge under the present situation is gathered, tries to achieve the natural frequency of vibration, and using the natural frequency of vibration as inspection sample This;Step 6, according to the BP neural network model established in step 4, the self excited vibrational frequency of bridge span under current influence factor is simulated, Draw fitting frequency values fs, as healthy sample;Step 7, healthy sample and test samples are subjected to both hypothesis testings, observation and whether there is significant difference:If there is significant difference, then it is assumed that current bridge structure is damaged, and carries out damage alarming;If there is no significant difference, then it is assumed that current bridge structure is in a safe condition, and by current test samples Extend in training sample database, the BP neural network model established in step 4 is trained again.
- 2. bridge entirety damage alarming method according to claim 1, it is characterised in that step 4 specifically includes:Step 401, the topological structure of BP neural network model is determined:Determine the neuron number that input layer, output layer, the hidden layer number of plies and each hidden layer of BP neural network model are included;Step 402, from input quantity of the influence factor as BP neural network model, using the natural frequency of vibration as output quantity, establish Training sample database;Step 403, using most of data in training sample database as training sample, using remaining sub-fraction as test Sample, BP neural network is trained using conjugate gradient method and BP neural network model surveyed using test sample Examination.
- 3. bridge entirety damage alarming method according to claim 2, it is characterised in that hidden layer is determined using empirical method The number of plies, the neuron number that hidden layer is included using object function determine, wherein, object function is defined as output valve and target Error sum of squares between value, when the object function reaches minimum, the neuron number that hidden layer is included is optimal.
- 4. bridge entirety damage alarming method according to claim 1, it is characterised in that step 7 specifically includes:Step 701, the fitting frequency values f drawn up with the BP neural network pattern die trainedsAs healthy sample, healthy sample Average be μs;Step 702, with the practical frequency value f of self excited vibrational frequency of bridge span under current influence factormAs test samples, test samples Average is μm;Step 703, hypothesis testing is carried out to healthy sample and test samples, null hypothesis is:H0:μs=μmH1:μs≠μmIf receive it is assumed that i.e. H0Set up, now μs=μm, then healthy sample and test samples do not have notable difference, i.e. bridge knot Structure is without obvious damage, without carrying out the overall damage alarming of bridge structure;If refusal is it is assumed that i.e. H1Set up, now μs≠μm, then healthy sample and test samples have notable difference, i.e. bridge structure Have and substantially damage, it is necessary to carry out the overall damage alarming of bridge structure;Step 704, if bridge damages without obvious, test samples are extended in training sample database, and again to god It is trained through network model.
- 5. bridge entirety damage alarming method according to claim 4, it is characterised in that frequency values f will be fittedsWith actual measurement Frequency values fmHypothesis testing is carried out, both whether there is significant difference for observation, judge that it is bright whether bridge structure occurs with this Aobvious damage.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711085667.6A CN107871161A (en) | 2017-11-07 | 2017-11-07 | A kind of bridge entirety damage alarming method based on neutral net |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711085667.6A CN107871161A (en) | 2017-11-07 | 2017-11-07 | A kind of bridge entirety damage alarming method based on neutral net |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107871161A true CN107871161A (en) | 2018-04-03 |
Family
ID=61753725
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711085667.6A Pending CN107871161A (en) | 2017-11-07 | 2017-11-07 | A kind of bridge entirety damage alarming method based on neutral net |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107871161A (en) |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108629144A (en) * | 2018-06-11 | 2018-10-09 | 湖北交投智能检测股份有限公司 | A kind of bridge health appraisal procedure |
CN109102016A (en) * | 2018-08-08 | 2018-12-28 | 北京新桥技术发展有限公司 | A kind of test method for bridge technology situation |
CN109614669A (en) * | 2018-11-23 | 2019-04-12 | 同济大学 | Net grade Bridge performance assessment prediction method |
CN110567745A (en) * | 2019-09-16 | 2019-12-13 | 中国铁道科学研究院集团有限公司铁道建筑研究所 | Bridge pier detection evaluation system under water |
CN110795780A (en) * | 2019-09-09 | 2020-02-14 | 杭州鲁尔物联科技有限公司 | XGboost algorithm-based cable-stayed bridge finite element correction method |
CN111160528A (en) * | 2019-12-28 | 2020-05-15 | 浙江大学 | Method for predicting service performance degradation of reinforced concrete bridge |
CN111625988A (en) * | 2020-03-10 | 2020-09-04 | 河北工程大学 | Bridge health management analysis and prediction system and method based on deep learning |
CN111738996A (en) * | 2020-06-09 | 2020-10-02 | 交通运输部公路科学研究所 | Bridge health monitoring and early warning system based on machine learning |
CN111967076A (en) * | 2020-07-15 | 2020-11-20 | 哈尔滨工业大学(深圳) | Method, device and equipment for separating influence of temperature action on engineering structure frequency |
CN113139691A (en) * | 2021-04-30 | 2021-07-20 | 北华大学 | High-speed rail bridge health monitoring system and method based on piezoelectric sensor |
CN113219048A (en) * | 2021-07-09 | 2021-08-06 | 西南交通大学 | Steel bridge damage detection system and method based on eddy current and digital twinning technology |
CN113642068A (en) * | 2021-07-07 | 2021-11-12 | 哈尔滨工业大学 | Bridge vortex-induced vibration amplitude prediction method based on decision tree and recurrent neural network |
CN114444366A (en) * | 2022-04-08 | 2022-05-06 | 深圳市城市交通规划设计研究中心股份有限公司 | Bridge digital twin model updating method and device based on finite element simulation |
CN116029555A (en) * | 2023-03-22 | 2023-04-28 | 西南科技大学 | Bridge risk identification early warning system based on lightweight neural network and application method |
CN117077897A (en) * | 2023-09-21 | 2023-11-17 | 四川省华地建设工程有限责任公司 | Method and system for deducing damage of earthquake disaster |
CN117216844A (en) * | 2023-09-12 | 2023-12-12 | 汕头大学 | Bridge structure damage detection method, system and storage medium |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101145214A (en) * | 2007-11-06 | 2008-03-19 | 东南大学 | Cable-stayed bridge cable damage positioning method based on modified reverse transmittance nerve network |
CN102508110A (en) * | 2011-10-10 | 2012-06-20 | 上海大学 | Texture-based insulator fault diagnostic method |
CN102565194A (en) * | 2012-02-09 | 2012-07-11 | 东南大学 | Method for carrying out early warning on damage to steel box girder of long span bridge in operation state |
CN104200005A (en) * | 2014-07-28 | 2014-12-10 | 东北大学 | Bridge damage identification method based on neural network |
CN104200004A (en) * | 2014-07-28 | 2014-12-10 | 东北大学 | Optimized bridge damage identification method based on neural network |
CN106092402A (en) * | 2016-05-31 | 2016-11-09 | 东南大学 | Total stress computational methods based on Monitoring Data and the large span steel beam bridge of analysis on temperature stress and safe early warning method |
WO2017130699A1 (en) * | 2016-01-26 | 2017-08-03 | 富士フイルム株式会社 | Crack information detection device, crack information detection method, and crack information detection program |
-
2017
- 2017-11-07 CN CN201711085667.6A patent/CN107871161A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101145214A (en) * | 2007-11-06 | 2008-03-19 | 东南大学 | Cable-stayed bridge cable damage positioning method based on modified reverse transmittance nerve network |
CN102508110A (en) * | 2011-10-10 | 2012-06-20 | 上海大学 | Texture-based insulator fault diagnostic method |
CN102565194A (en) * | 2012-02-09 | 2012-07-11 | 东南大学 | Method for carrying out early warning on damage to steel box girder of long span bridge in operation state |
CN104200005A (en) * | 2014-07-28 | 2014-12-10 | 东北大学 | Bridge damage identification method based on neural network |
CN104200004A (en) * | 2014-07-28 | 2014-12-10 | 东北大学 | Optimized bridge damage identification method based on neural network |
WO2017130699A1 (en) * | 2016-01-26 | 2017-08-03 | 富士フイルム株式会社 | Crack information detection device, crack information detection method, and crack information detection program |
CN106092402A (en) * | 2016-05-31 | 2016-11-09 | 东南大学 | Total stress computational methods based on Monitoring Data and the large span steel beam bridge of analysis on temperature stress and safe early warning method |
Non-Patent Citations (2)
Title |
---|
孙宗光 等: "基于新奇检测技术的桥梁结构损伤预警方法", 《公路交通科技》 * |
邓扬 等: "基于监测数据的大跨度悬索桥频率与环境条件的相关性模型", 《中南大学学报(自然科学版)》 * |
Cited By (25)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108629144A (en) * | 2018-06-11 | 2018-10-09 | 湖北交投智能检测股份有限公司 | A kind of bridge health appraisal procedure |
CN108629144B (en) * | 2018-06-11 | 2022-06-17 | 湖北交投智能检测股份有限公司 | Bridge health assessment method |
CN109102016A (en) * | 2018-08-08 | 2018-12-28 | 北京新桥技术发展有限公司 | A kind of test method for bridge technology situation |
CN109614669A (en) * | 2018-11-23 | 2019-04-12 | 同济大学 | Net grade Bridge performance assessment prediction method |
CN110795780A (en) * | 2019-09-09 | 2020-02-14 | 杭州鲁尔物联科技有限公司 | XGboost algorithm-based cable-stayed bridge finite element correction method |
CN110795780B (en) * | 2019-09-09 | 2023-02-10 | 杭州鲁尔物联科技有限公司 | XGboost algorithm-based cable-stayed bridge finite element correction method |
CN110567745B (en) * | 2019-09-16 | 2022-06-07 | 中国铁道科学研究院集团有限公司铁道建筑研究所 | Bridge pier detection evaluation system under water |
CN110567745A (en) * | 2019-09-16 | 2019-12-13 | 中国铁道科学研究院集团有限公司铁道建筑研究所 | Bridge pier detection evaluation system under water |
CN111160528A (en) * | 2019-12-28 | 2020-05-15 | 浙江大学 | Method for predicting service performance degradation of reinforced concrete bridge |
CN111160528B (en) * | 2019-12-28 | 2021-01-08 | 浙江大学 | Method for predicting service performance degradation of reinforced concrete bridge |
CN111625988A (en) * | 2020-03-10 | 2020-09-04 | 河北工程大学 | Bridge health management analysis and prediction system and method based on deep learning |
CN111738996B (en) * | 2020-06-09 | 2023-04-07 | 交通运输部公路科学研究所 | Bridge health monitoring and early warning system based on machine learning |
CN111738996A (en) * | 2020-06-09 | 2020-10-02 | 交通运输部公路科学研究所 | Bridge health monitoring and early warning system based on machine learning |
CN111967076A (en) * | 2020-07-15 | 2020-11-20 | 哈尔滨工业大学(深圳) | Method, device and equipment for separating influence of temperature action on engineering structure frequency |
CN113139691A (en) * | 2021-04-30 | 2021-07-20 | 北华大学 | High-speed rail bridge health monitoring system and method based on piezoelectric sensor |
CN113642068A (en) * | 2021-07-07 | 2021-11-12 | 哈尔滨工业大学 | Bridge vortex-induced vibration amplitude prediction method based on decision tree and recurrent neural network |
CN113219048A (en) * | 2021-07-09 | 2021-08-06 | 西南交通大学 | Steel bridge damage detection system and method based on eddy current and digital twinning technology |
CN113219048B (en) * | 2021-07-09 | 2021-09-14 | 西南交通大学 | Steel bridge damage detection system and method based on eddy current and digital twinning technology |
CN114444366A (en) * | 2022-04-08 | 2022-05-06 | 深圳市城市交通规划设计研究中心股份有限公司 | Bridge digital twin model updating method and device based on finite element simulation |
CN116029555A (en) * | 2023-03-22 | 2023-04-28 | 西南科技大学 | Bridge risk identification early warning system based on lightweight neural network and application method |
CN116029555B (en) * | 2023-03-22 | 2023-06-13 | 西南科技大学 | Bridge risk identification early warning system based on lightweight neural network and application method |
CN117216844A (en) * | 2023-09-12 | 2023-12-12 | 汕头大学 | Bridge structure damage detection method, system and storage medium |
CN117216844B (en) * | 2023-09-12 | 2024-03-26 | 汕头大学 | Bridge structure damage detection method, system and storage medium |
CN117077897A (en) * | 2023-09-21 | 2023-11-17 | 四川省华地建设工程有限责任公司 | Method and system for deducing damage of earthquake disaster |
CN117077897B (en) * | 2023-09-21 | 2024-03-19 | 四川省华地建设工程有限责任公司 | Method and system for deducing damage of earthquake disaster |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107871161A (en) | A kind of bridge entirety damage alarming method based on neutral net | |
Cho et al. | Structural health monitoring of a cable-stayed bridge using wireless smart sensor technology: data analyses | |
CN105300692B (en) | A kind of bearing failure diagnosis and Forecasting Methodology based on expanded Kalman filtration algorithm | |
CN107727338B (en) | A kind of bridge damnification diagnostic method based on Vehicle-Bridge Coupling System | |
CN104712542B (en) | A kind of reciprocating compressor sensitive features based on Internet of Things are extracted and method for diagnosing faults | |
CN105606499B (en) | Suspended particulate matter mass concentration real-time detection device, and measuring method | |
CN110702418A (en) | Aircraft engine fault prediction method | |
CN101900789B (en) | Tolerance analog circuit fault diagnosing method based on wavelet transform and fractal dimension | |
CN104200005A (en) | Bridge damage identification method based on neural network | |
CN106407633A (en) | Method and system for estimating ground PM2.5 based on space-time regression Kriging model | |
CN102879791A (en) | System for sensing activity data of elder person based on Beidou positioning terminal | |
CN104880217B (en) | A kind of fault sensor signal reconstruct method based on the measured value degree of association | |
CN109325263A (en) | Truss bridge damage position neural network based and damage extent identification method | |
CN106777909A (en) | Gestational period health risk assessment system | |
CN107423406A (en) | A kind of construction method of campus student relational network | |
CN106769030A (en) | A kind of bearing state tracking and Forecasting Methodology based on MEA BP neural network algorithms | |
CN106618609A (en) | Psychological test method and psychological test instrument | |
CN102682180A (en) | Evaluation method for performance degradation of rotary mechanical equipment | |
CN104458252A (en) | Method for monitoring running state of high-speed train gear box | |
CN109063290A (en) | A kind of flutter prediction technique based on nerual network technique | |
CN103913512B (en) | The damage reason location system of suspension cable periodic detection | |
CN104732098A (en) | Early warning method for deterioration of bearing capacity of railway steel truss arched bridge girder | |
CN108549847A (en) | A kind of girder structure crack damage recognition methods under the conditions of no reference data | |
CN107367552A (en) | Damage propatagtion method of real-time based on FCM algorithms under the influence of time varying temperature | |
Akintunde et al. | Singular value decomposition and unsupervised machine learning for virtual strain sensing: Application to an operational railway bridge |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180403 |
|
WD01 | Invention patent application deemed withdrawn after publication |