CN104316341A - Underground structure damage identification method based on BP neural network - Google Patents
Underground structure damage identification method based on BP neural network Download PDFInfo
- Publication number
- CN104316341A CN104316341A CN201410655876.XA CN201410655876A CN104316341A CN 104316341 A CN104316341 A CN 104316341A CN 201410655876 A CN201410655876 A CN 201410655876A CN 104316341 A CN104316341 A CN 104316341A
- Authority
- CN
- China
- Prior art keywords
- neural network
- damage
- input
- network model
- layer
- 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
Landscapes
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention discloses an underground structure damage identification method based on a BP neural network. The structure damage is predicted through the curvature change rate. The method includes the steps that a BP neural network model is established, wherein the BP neural network model comprises a three-layer network structure composed of an input layer, a hidden layer and an output layer; the curvature change rate is selected as the input quantity of the BP neural network model, the damage position and the damage degree serve as the output quantity, and the damage position and the damage degree are identified; network training is conducted on the BP neural network model, the weight and the threshold are continuously corrected, the input error of the BP neural network model reaches the set threshold after being adjusted, and iteration is conducted until network convergence; the damage position and the damage degree of an underground structure are determined according to the curvature change rate in the BP neural network model. By the adoption of the damage identification method, in the double-damage prediction process, the error between the theoretical value and the predicted value is small, and it is shown that the BP neural network can accurately predict double-damage nodes.
Description
Technical field
The invention discloses a kind of underground structure damnification recognition method based on BP neural network, relate to structural damage detection technical field.
Background technology
Underground structure in use, due to Fatigue Load, environmental corrosion, material aging, component defect and other factors, produce damage accumulation gradually, this damage accumulation makes the load-resisting capacity degradation of structure and the decline of opposing disaster ability, particularly may may occur serious damage by existing structure when experience geological process.Therefore analyze the dynamic damage degree of underground structure, and then evaluation is carried out to its seismic Damage characteristic seem particularly important.The Quantitative assessment of underground structure degree of injury under geological process is finally set up out, for rear reparation of seismic design and shake of underground structure provides theoretical foundation by seismic Damage evaluation.
Because the kinematic behavior of structure is directly related with structural physical parameter, the damage of structure will cause the change of corresponding kinematic behavior.But the simple damnification recognition method using kinetic parameter constructs signature for damage detection with kinetic parameter, often need to solve complicated mathematical inversion problem, make more difficult being applied in Practical Project.In view of neural network can make inverse problem direct problem with the non-linear mapping capability of its excellence, kinetic parameter analytical approach is combined with nerual network technique, with kinetic parameter structure signature for damage detection, as the characteristic parameter of neural network input, thus carry out Damage Assessment Method.
BP neural network becomes with its good non-linear mapping capability the neural network be most widely used in diagnosing structural damage field.BP neural network is by the vague generalization of W-H learning rules, non-linear differentiable function is carried out to the multitiered network of Weight Training, it is a kind of multilayer feed-forward neural network, its neuronic transforming function transformation function is S type function, therefore output quantity is the continuous quantity between 0 to 1, and it can realize the arbitrary Nonlinear Mapping from being input to output.
Summary of the invention
Technical matters to be solved by this invention is: for the defect of prior art, a kind of underground structure damnification recognition method based on BP neural network is provided, utilizing finite element ABAQUS and MATLAB program Modal Analysis Theory and BP neural network to be combined is applied in the non-destructive tests of construction of underground structure, realizes the identification to damage position and degree of injury simultaneously.
The present invention is for solving the problems of the technologies described above by the following technical solutions:
Based on a underground structure damnification recognition method for BP neural network, predicted by the damage of curvature variation to structure, concrete steps comprise:
Step one, set up BP neural network model, comprise Three Tiered Network Architecture input layer, hidden layer and output layer, setting:
X
jrepresent the input of an input layer jth node, j=1,2 ..., M; w
ijrepresent the weights between hidden layer i-th node to an input layer jth node; θ
jrepresent the threshold value of hidden layer i-th node; φ (x) represents the excitation function of hidden layer; w
kjrepresent the weights between an output layer kth node to a hidden layer jth node, j=1,2 ..., q; a
jrepresent the threshold value of an output layer jth node, j=1,2 ..., q;
represent the excitation function of output layer; O
jrepresent the output of an output layer jth node, j=1,2 ..., q;
The BP algorithm carried out based on above-mentioned BP neural network model comprises the propagated forward of signal and the backpropagation of error, undertaken, and the correction of weights and threshold is carried out from the direction outputting to input when calculating actual output by from the direction being input to output;
Step 2, select curvature variation as the input quantity of BP neural network model, using damage position and degree of injury as output quantity, come position and the degree of identification of damage, wherein, curvature variation is defined as: when supposing structure not damaged, the i-th rank curvature value of node k is δ
i,k, after damage, under i first order mode, curvature value is
then its curature variation amount
the relative changes rate defining its curvature is the rate of change sum of the curvature on front i rank
Step 3, carry out network training to BP neural network model, constantly revise weights and threshold, make the error originated from input of BP neural network model reach the threshold value of setting by adjustment, iteration is until network convergence;
Step 4, test sample book collection to be input in the BP neural network model that trains, to determine damage position and the degree of injury of underground structure according to BP neural network model mean curvature rate of change.
As present invention further optimization scheme, in step 3, described error originated from input is 0.0015.
As present invention further optimization scheme, in step one, the input layer number of described BP neural network model is 4, and hidden layer neuron number is 6, and output layer neuron number is 4.
As present invention further optimization scheme, the transport function of described hidden layer neuron adopts S type tan tansig, and output layer neural transferring function adopts S type logarithmic function logsig.
As present invention further optimization scheme, in step 3, the training sample of described network training is, the curvature variation data of tested point under a part of degree of injury.
As present invention further optimization scheme, in step 4, described test sample book integrates as training sample, adds the curvature variation of tested point when another part degree of injury.
The present invention adopts above technical scheme compared with prior art, there is following technique effect: adopt damnification recognition method disclosed in this invention, when two damage forecast, the error between theoretical value and predicted value is less, illustrates that BP neural network can predict two damage node more accurately.
Accompanying drawing explanation
Fig. 1 is BP network structure in the present invention.
Fig. 2 is BP algorithm flow schematic diagram in the present invention.
Fig. 3 is in one embodiment of the present of invention, the FEM meshing of foundation soil-underground tunnel structure system.
Fig. 4 is in one embodiment of the present of invention, underground tunnel structure node diagram.
Fig. 5 is the Curvature varying rate curve of each node.
Fig. 6 is the curvature variation of lower No. 3 nodes of Injured level.
Fig. 7 is the graph of errors between degree of injury predicted value and theoretical value.
Embodiment
Be described below in detail embodiments of the present invention, the example of described embodiment is shown in the drawings, and wherein same or similar label represents same or similar element or has element that is identical or similar functions from start to finish.Being exemplary below by the embodiment be described with reference to the drawings, only for explaining the present invention, and can not limitation of the present invention being interpreted as.
Below in conjunction with accompanying drawing, technical scheme of the present invention is described in further detail:
The present invention is on the basis of research domestic and international related structure non-destructive tests and neural network literature in a large number, according to the development prospect of Damage Assessment Method and neural network, utilizing finite element ABAQUS and MATLAB program Modal Analysis Theory and BP neural network to be combined is applied in the non-destructive tests of construction of underground structure, realize the identification to damage position and degree of injury simultaneously, form a set of Structural Damage Identification based on Modal Analysis Theory and BP neural network.
Basic BP algorithm comprises two aspects: the propagated forward of signal and the backpropagation of error.Undertaken by from the direction being input to output when namely calculating actual output, and the correction of weights and threshold is carried out from the direction outputting to input.Its network structure and algorithm flow are as Fig. 1 and Fig. 2.In figure: x
jrepresent the input of an input layer jth node, j=1,2 ..., M; w
ijrepresent the weights between hidden layer i-th node to an input layer jth node; θ
jrepresent the threshold value of hidden layer i-th node; φ (x) represents the excitation function of hidden layer; w
kjrepresent the weights between an output layer kth node to a hidden layer jth node, j=1,2 ..., q; a
jrepresent the threshold value of an output layer jth node, j=1,2 ..., q;
represent the excitation function of output layer; O
jrepresent the output of an output layer jth node, j=1,2 ..., q.
The selection of the input vector of neural network directly affects the recognition performance of network, and the physical quantity of damage reason location should be the parameter responsive to structure partial damage, and is the trend of monotone variation with the increase of structural damage degree.The present invention selects curvature variation as the input quantity of neural network, using damage position and degree of injury as output quantity, carrys out position and the degree of identification of damage.The present invention adopts curvature variation to be defined as: when supposing structure not damaged, the i-th rank curvature value of node k is δ
i,k, after damage, under i first order mode, curvature value is
then its curature variation amount
the rate of change sum of the curvature on i rank before this relative changes rate defining its curvature is
in the present invention, emphasis inquires into structural damage to the impact of its curvature variation, and is effectively predicted by the damage of curvature variation to structure.
The present invention is in non-destructive tests process, BP network architecture adopts three-layer network, input layer number is 4, hidden layer neuron number is 6, output layer neuron number is 4, the transport function of hidden layer neuron adopts S type tan tansig, and output layer neural transferring function adopts S type logarithmic function logsig.Network training process is a process constantly revising weights and threshold, and make the error originated from input of network reach minimum by adjustment, target training error of the present invention is 0.0015, and iteration is until network convergence.
Getting typical underground tunnel structure is research object, underground tunnel structure adopts beam element simulation, and foundation soil adopts solid element simulation, and the ground width of computational fields is 120m, as shown in Figure 3, underground tunnel structure node is as Fig. 4 for the FEM meshing of foundation soil-underground tunnel structure system.Structured material mechanics parameter: elastic modulus E=3 × 10
10pa, density p=2500kg/m
3, Poisson ratio μ=0.2; The mechanical parameters of the soil body: elastic modulus E=8 × 10
6pa, density p=1800kg/m
3, Poisson ratio μ=0.35.The input quantity of neural network is drawn by finite element modal analysis.
The damage essence of structure is exactly the loss of structure partial rigidity, therefore can carry out simplification to the Dynamic Damage Behavior analysis of structure and calculate, and supposes that structural damage can be simulated by the method reducing unitary elasticity modulus.The damage of this position is simulated, with the degree of injury of the reduction degree model configuration of unitary elasticity modulus with the reduction of the elastic modulus of a certain position.
The most serious in the position of about 45 °, tunnel plane according to the damage of the known circular tunnel structure of existing research, therefore Study on Damage Identification is carried out to No. 3, No. 7, No. 11 and No. 15 nodes in circular underground tunnel structure finite element model.Suppose that underground structure, at No. 3 node surrounding cells places, the damage of 30% occurs, the curve that before making structure, 4 rank curvature variations change with node as shown in Figure 5, can find out: the peak value of curve appears at No. 3 node location places, illustrate that curvature variation is very sensitive to damage position, not by the impact of symmetrical structure, therefore the damage position of underground structure can be determined according to 4 rank curvature variations before structure.
In order to determine the variation characteristic of the degree of injury of underground structure further, by the curvature variation of computing unit when different degree of impairment, Fig. 6 is the Curvature varying rate curve around No. 3 nodes when degree of injury is respectively 10%, 30%, 50%, 70% and 90%, can draw from figure, the curvature variation of structure is responsive to underground structural damage degree, namely at identical damage position place, along with the increase of degree of injury, the numerical value of curvature variation also increases gradually.
Mode Shape rate of change when being respectively 10%, 30%, 50%, 70%, 90% using 2,5,8, No. 11 Joint Damage degree totally 20 groups of data as the training sample of BP neural network.Represent the degree of impairment of 4 unit above respectively with the output of neural network (y1, y2, y3, y4), such as network output (0,0.5,0,0) represents No. 5 unit has the damage of 50% to occur.With training sample, curvature variation when adding 40% and 80% degree of injury, as the test sample book of neural network, is used for checking the adaptive faculty recalling memory, interpolation and extrapolation of neural network.
BP neural network neural metwork training error after 20000 training is 0.001398, and training stops.Be input in the neural network trained by test sample book collection, test result is as shown in table 1, and the graph of errors between theoretical value and predicted value as shown in Figure 7.Can find: the error between theoretical value and predicted value is less, the predicted value drawn based on BP neural network can reflect degree of injury completely.
Table 1 is based on the non-destructive tests result of BP neural network
Being called two degree of impairment when rigidity reduction occurs two unit of certain in underground structure simultaneously, by setting up BP neural network, carrying out two non-destructive tests Primary Location to underground structure, adopt two mode damaged to simulate, various damage regime is as table 2.
Table 2 pair damage regime
Curvature variation when being respectively 10%, 30%, 50%, 70%, 90% using 3 & 7,3 & 11,3 & 15,7 & 11,7 & 15,11 & No. 15 Joint Damage degree is as the training sample of BP neural network.With the output (y1 of neural network, y2, y3, y4) degree of impairment of 3 & 7,3 & 11,3 & 15,7 & 11,7 & 15,11 & No. 15 each nodes is represented respectively, such as network exports (0.5,0.5,0,0) No. 3 are represented and No. 7 nodes have the damage of 50% to occur.Test result is as shown in table 3, can find: when two damage forecast, and the error between theoretical value and predicted value is also less, illustrates that BP neural network can predict two damage node equally preferably.
Table 3 is based on two non-destructive tests results of BP neural network
Claims (6)
1. based on a underground structure damnification recognition method for BP neural network, it is characterized in that, predicted by the damage of curvature variation to structure, concrete steps comprise:
Step one, set up BP neural network model, comprise Three Tiered Network Architecture input layer, hidden layer and output layer, setting:
X
jrepresent the input of an input layer jth node, j=1,2 ..., M; w
ijrepresent the weights between hidden layer i-th node to an input layer jth node; θ
jrepresent the threshold value of hidden layer i-th node; φ (x) represents the excitation function of hidden layer; w
kjrepresent the weights between an output layer kth node to a hidden layer jth node, j=1,2 ..., q; a
jrepresent the threshold value of an output layer jth node, j=1,2 ..., q;
represent the excitation function of output layer; O
jrepresent the output of an output layer jth node, j=1,2 ..., q;
The BP algorithm carried out based on above-mentioned BP neural network model comprises the propagated forward of signal and the backpropagation of error, undertaken, and the correction of weights and threshold is carried out from the direction outputting to input when calculating actual output by from the direction being input to output;
Step 2, select curvature variation as the input quantity of BP neural network model, using damage position and degree of injury as output quantity, come position and the degree of identification of damage, wherein, curvature variation is defined as: when supposing structure not damaged, the i-th rank curvature value of node k is δ
i,k, after damage, under i first order mode, curvature value is
then its curature variation amount
the relative changes rate defining its curvature is the rate of change sum of the curvature on front i rank
Step 3, carry out network training to BP neural network model, constantly revise weights and threshold, make the error originated from input of BP neural network model reach the threshold value of setting by adjustment, iteration is until network convergence;
Step 4, test sample book collection to be input in the BP neural network model that trains, to determine damage position and the degree of injury of underground structure according to BP neural network model mean curvature rate of change.
2. a kind of underground structure damnification recognition method based on BP neural network as claimed in claim 1, it is characterized in that: in step 3, described error originated from input is 0.0015.
3. a kind of underground structure damnification recognition method based on BP neural network as claimed in claim 1 or 2, it is characterized in that: in step one, the input layer number of described BP neural network model is 4, and hidden layer neuron number is 6, and output layer neuron number is 4.
4. a kind of underground structure damnification recognition method based on BP neural network as claimed in claim 3, it is characterized in that: the transport function of described hidden layer neuron adopts S type tan tansig, and output layer neural transferring function adopts S type logarithmic function logsig.
5. a kind of underground structure damnification recognition method based on BP neural network as claimed in claim 1, it is characterized in that: in step 3, the training sample of described network training is, the curvature variation data of tested point under a part of degree of injury.
6. a kind of underground structure damnification recognition method based on BP neural network as claimed in claim 5, it is characterized in that: in step 4, described test sample book integrates as training sample, adds the curvature variation of tested point when another part degree of injury.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410655876.XA CN104316341A (en) | 2014-11-17 | 2014-11-17 | Underground structure damage identification method based on BP neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410655876.XA CN104316341A (en) | 2014-11-17 | 2014-11-17 | Underground structure damage identification method based on BP neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104316341A true CN104316341A (en) | 2015-01-28 |
Family
ID=52371603
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410655876.XA Pending CN104316341A (en) | 2014-11-17 | 2014-11-17 | Underground structure damage identification method based on BP neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104316341A (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105184364A (en) * | 2015-09-28 | 2015-12-23 | 华中科技大学 | Tension compensation method for damage identification |
| CN105912822A (en) * | 2016-05-10 | 2016-08-31 | 东南大学 | Method for simulating desulfurization efficiency of seawater desulfurizer through neural network |
| CN108896649A (en) * | 2018-04-28 | 2018-11-27 | 南京航空航天大学 | Method for estimating yield strength of ferromagnetic material based on electromagnetic ultrasound |
| CN109030132A (en) * | 2018-06-14 | 2018-12-18 | 国网安徽省电力有限公司电力科学研究院 | A kind of creep impairment reference block preparation method, damage detecting method and system |
| CN109240885A (en) * | 2018-08-30 | 2019-01-18 | 郑州云海信息技术有限公司 | A kind of method for monitoring performance, system and electronic equipment and storage medium |
| CN109271828A (en) * | 2017-07-17 | 2019-01-25 | 国网江苏省电力公司泰州供电公司 | The method and system of construction segregator barriers condition intelligent detection based on deep learning |
| CN109361539A (en) * | 2018-10-16 | 2019-02-19 | 郑州云海信息技术有限公司 | A kind of performance alarm method and apparatus based on development network |
| CN110146678A (en) * | 2019-05-30 | 2019-08-20 | 广东工业大学 | A kind of structural damage detection system, method, apparatus and readable storage medium storing program for executing |
| CN110909503A (en) * | 2019-11-25 | 2020-03-24 | 中国船舶重工集团公司第七一九研究所 | Method for predicting leakage of pipeline system flange |
| CN112380631A (en) * | 2020-12-02 | 2021-02-19 | 黑龙江科技大学 | Novel iterative hybrid test method based on neural network |
| CN112613219A (en) * | 2020-12-28 | 2021-04-06 | 长沙理工大学 | Floating plate track steel spring damage positioning and identifying method |
| CN113341459A (en) * | 2021-05-12 | 2021-09-03 | 北京大学 | Earthquake positioning method and device based on machine learning and dynamics calculation fusion |
| CN113361170A (en) * | 2021-06-11 | 2021-09-07 | 金陵科技学院 | Unsaturated soil hydraulic characteristic parameter numerical value inversion method based on evaporation process |
| CN113742983A (en) * | 2021-10-09 | 2021-12-03 | 福州大学 | Long-span structural damage identification method based on depth self-encoder neural network |
| CN113868750A (en) * | 2021-10-20 | 2021-12-31 | 清华大学 | Building earthquake damage and residual capacity evaluation method based on deep machine learning |
| CN116123988A (en) * | 2023-04-14 | 2023-05-16 | 国家石油天然气管网集团有限公司 | Pipeline geometric deformation detection method and system based on data processing |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5774376A (en) * | 1995-08-07 | 1998-06-30 | Manning; Raymund A. | Structural health monitoring using active members and neural networks |
| 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 |
| CN103018338A (en) * | 2012-12-05 | 2013-04-03 | 河海大学 | Concrete lossless detection method based on sound emission and neural network |
| JP5384429B2 (en) * | 2010-05-21 | 2014-01-08 | 日本電信電話株式会社 | Crack detection apparatus, crack detection method and program for concrete structure image |
| CN103575807A (en) * | 2013-10-24 | 2014-02-12 | 河海大学 | Method for detecting structural damage of beam based on Teager energy operator-wavelet transformation curvature mode |
-
2014
- 2014-11-17 CN CN201410655876.XA patent/CN104316341A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5774376A (en) * | 1995-08-07 | 1998-06-30 | Manning; Raymund A. | Structural health monitoring using active members and neural networks |
| 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 |
| JP5384429B2 (en) * | 2010-05-21 | 2014-01-08 | 日本電信電話株式会社 | Crack detection apparatus, crack detection method and program for concrete structure image |
| CN103018338A (en) * | 2012-12-05 | 2013-04-03 | 河海大学 | Concrete lossless detection method based on sound emission and neural network |
| CN103575807A (en) * | 2013-10-24 | 2014-02-12 | 河海大学 | Method for detecting structural damage of beam based on Teager energy operator-wavelet transformation curvature mode |
Non-Patent Citations (2)
| Title |
|---|
| 魏锦辉: "基于神经网络的桥梁损伤识别", 《中国优秀硕士学位论文全文数据库工程科技II辑》 * |
| 黄汉江: "《全国投资与建设研究论文集》", 30 June 2009 * |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105184364B (en) * | 2015-09-28 | 2017-12-12 | 华中科技大学 | A kind of pulling force compensation method for non-destructive tests |
| CN105184364A (en) * | 2015-09-28 | 2015-12-23 | 华中科技大学 | Tension compensation method for damage identification |
| CN105912822A (en) * | 2016-05-10 | 2016-08-31 | 东南大学 | Method for simulating desulfurization efficiency of seawater desulfurizer through neural network |
| CN105912822B (en) * | 2016-05-10 | 2019-08-20 | 东南大学 | The neuron network simulation method of seawater desulfurization device desulfuration efficiency |
| CN109271828A (en) * | 2017-07-17 | 2019-01-25 | 国网江苏省电力公司泰州供电公司 | The method and system of construction segregator barriers condition intelligent detection based on deep learning |
| CN108896649B (en) * | 2018-04-28 | 2021-05-07 | 南京航空航天大学 | Method for estimating yield strength of ferromagnetic material based on electromagnetic ultrasound |
| CN108896649A (en) * | 2018-04-28 | 2018-11-27 | 南京航空航天大学 | Method for estimating yield strength of ferromagnetic material based on electromagnetic ultrasound |
| CN109030132A (en) * | 2018-06-14 | 2018-12-18 | 国网安徽省电力有限公司电力科学研究院 | A kind of creep impairment reference block preparation method, damage detecting method and system |
| CN109240885A (en) * | 2018-08-30 | 2019-01-18 | 郑州云海信息技术有限公司 | A kind of method for monitoring performance, system and electronic equipment and storage medium |
| CN109361539A (en) * | 2018-10-16 | 2019-02-19 | 郑州云海信息技术有限公司 | A kind of performance alarm method and apparatus based on development network |
| CN110146678A (en) * | 2019-05-30 | 2019-08-20 | 广东工业大学 | A kind of structural damage detection system, method, apparatus and readable storage medium storing program for executing |
| CN110909503A (en) * | 2019-11-25 | 2020-03-24 | 中国船舶重工集团公司第七一九研究所 | Method for predicting leakage of pipeline system flange |
| CN110909503B (en) * | 2019-11-25 | 2023-08-18 | 中国船舶重工集团公司第七一九研究所 | Prediction method for flange leakage of pipeline system |
| CN112380631A (en) * | 2020-12-02 | 2021-02-19 | 黑龙江科技大学 | Novel iterative hybrid test method based on neural network |
| CN112380631B (en) * | 2020-12-02 | 2023-02-14 | 黑龙江科技大学 | Novel iterative hybrid test method based on neural network |
| CN112613219A (en) * | 2020-12-28 | 2021-04-06 | 长沙理工大学 | Floating plate track steel spring damage positioning and identifying method |
| CN113341459A (en) * | 2021-05-12 | 2021-09-03 | 北京大学 | Earthquake positioning method and device based on machine learning and dynamics calculation fusion |
| CN113361170A (en) * | 2021-06-11 | 2021-09-07 | 金陵科技学院 | Unsaturated soil hydraulic characteristic parameter numerical value inversion method based on evaporation process |
| CN113742983A (en) * | 2021-10-09 | 2021-12-03 | 福州大学 | Long-span structural damage identification method based on depth self-encoder neural network |
| CN113868750A (en) * | 2021-10-20 | 2021-12-31 | 清华大学 | Building earthquake damage and residual capacity evaluation method based on deep machine learning |
| CN113868750B (en) * | 2021-10-20 | 2022-05-17 | 清华大学 | Building earthquake damage and residual capacity evaluation method based on deep machine learning |
| CN116123988A (en) * | 2023-04-14 | 2023-05-16 | 国家石油天然气管网集团有限公司 | Pipeline geometric deformation detection method and system based on data processing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104316341A (en) | Underground structure damage identification method based on BP neural network | |
| Mansour et al. | Predicting the shear strength of reinforced concrete beams using artificial neural networks | |
| Chang et al. | Structural damage detection using an iterative neural network | |
| CN111339488B (en) | Slope system failure probability calculation method based on Kriging | |
| Koeppe et al. | An efficient Monte Carlo strategy for elasto-plastic structures based on recurrent neural networks | |
| CN104200005A (en) | Bridge damage identification method based on neural network | |
| CN104200004A (en) | Optimized bridge damage identification method based on neural network | |
| CN104834808A (en) | Back propagation (BP) neural network based method for predicting service life of rubber absorber | |
| CN114117840B (en) | Structural performance prediction method based on simulation and test data hybrid drive | |
| JP2012518834A5 (en) | ||
| Lu et al. | Data correlation analysis for optimal sensor placement using a bond energy algorithm | |
| Arsava et al. | Modeling of magnetorheological dampers under various impact loads | |
| Hasançebi et al. | Detailed load rating analyses of bridge populations using nonlinear finite element models and artificial neural networks | |
| CN103884593A (en) | Structural damage location and damage degree identification method based on small amount of static measurement data | |
| CN111950180A (en) | Reliability analysis method and system for cable bearing bridge structure system | |
| Wootton et al. | Structural health monitoring of a footbridge using Echo State Networks and NARMAX | |
| Strauss et al. | Inverse statistical nonlinear FEM analysis of concrete structures | |
| Liu et al. | Total solution for structural mechanics problems | |
| Zhang et al. | A Bayesian neural network approach for probabilistic model updating using incomplete modal data | |
| Koeppe et al. | An intelligent meta‐element for linear elastic continua | |
| Zar et al. | Vibration-based damage detection of arch dams using least-square support vector machines and salp swarm algorithms | |
| Rao et al. | Fuzzy logic-based expert system to predict the results of finite element analysis | |
| Hasheminejad et al. | Predicting the collapsibility potential of unsaturated soils using adaptive neural fuzzy inference system and particle swarm optimization | |
| Monsalve-Giraldo et al. | Hybrid Parabolic Interpolation–Artificial Neural Network Method (HPI-ANNM) for long-term extreme response estimation of steel risers | |
| De et al. | A hybrid probabilistic framework for model validation with application to structural dynamics modeling |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150128 |
|
| RJ01 | Rejection of invention patent application after publication |