CN114444187B - Bridge damage diagnosis method for fusion of vibration transmission big data and capsule network - Google Patents

Bridge damage diagnosis method for fusion of vibration transmission big data and capsule network Download PDF

Info

Publication number
CN114444187B
CN114444187B CN202210108414.0A CN202210108414A CN114444187B CN 114444187 B CN114444187 B CN 114444187B CN 202210108414 A CN202210108414 A CN 202210108414A CN 114444187 B CN114444187 B CN 114444187B
Authority
CN
China
Prior art keywords
damage
layer
transfer rate
image
capsule
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.)
Active
Application number
CN202210108414.0A
Other languages
Chinese (zh)
Other versions
CN114444187A (en
Inventor
曹茂森
李帅
朱华新
姜亚洲
王泽雨
苏玛拉.德拉戈斯拉夫
崔丽
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangsu Zhongji Engineering Technology Research Co ltd
Hohai University HHU
JSTI Group Co Ltd
Original Assignee
Jiangsu Zhongji Engineering Technology Research Co ltd
Hohai University HHU
JSTI Group Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Jiangsu Zhongji Engineering Technology Research Co ltd, Hohai University HHU, JSTI Group Co Ltd filed Critical Jiangsu Zhongji Engineering Technology Research Co ltd
Priority to CN202210108414.0A priority Critical patent/CN114444187B/en
Publication of CN114444187A publication Critical patent/CN114444187A/en
Priority to GB2218502.9A priority patent/GB2615402A/en
Application granted granted Critical
Publication of CN114444187B publication Critical patent/CN114444187B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/13—Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M5/00—Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings
    • G01M5/0008—Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings of bridges
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M5/00—Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings
    • G01M5/0033—Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings by determining damage, crack or wear
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M5/00—Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings
    • G01M5/0066—Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings by exciting or detecting vibration or acceleration
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
    • 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
    • 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
    • G06N3/0464—Convolutional networks [CNN, ConvNet]
    • 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
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation

Abstract

The invention provides a bridge damage diagnosis method for merging vibration transmission big data with a capsule network, which comprises the following steps: according to the vibration response data, calculating to obtain a vibration transfer rate function, establishing a transfer rate function matrix, and converting the transfer rate function matrix into an image form to obtain a transfer rate image; constructing a neural network model for positioning and quantifying structural damage based on a capsule network, wherein a feature extraction layer in the capsule network is changed into a dense layer, and a convolution attention module is added after a second convolution layer; and inputting the transmissibility image into a trained neural network model, and carrying out damage positioning and quantification of the structure to obtain a predicted result of the damage position and degree. According to the method, through introducing a dense layer and a convolution attention mode, more efficient damage image extraction is realized, vector neurons are output through dynamic routing of a capsule network layer, attitude information in vibration transmission data information is stored, and finally prediction of damage positions and magnitude is realized through a full-connection layer.

Description

Bridge damage diagnosis method for fusion of vibration transmission big data and capsule network
Technical Field
The invention relates to the technical field of structural damage diagnosis, in particular to a bridge damage diagnosis method for fusing vibration transmission big data with a capsule network.
Background
The bridge structure has complex structure, is subjected to various load types and severe environments, and is inevitably damaged in the long-term service process, and the structural damage seriously threatens the safe operation of the structure. Therefore, the damage diagnosis of the structure is carried out, and the damage of the structure is found early and the damage degree is mastered, so that the method is an important means for ensuring the safe operation of the bridge structure.
At present, a plurality of large-scale bridges are provided with structural health monitoring systems so as to grasp the running state of the bridges in real time and ensure the safe operation of the bridges. The structural health monitoring system collects and stores a large amount of monitoring data, the large amount of monitoring data contains abundant information representing structural damage, the large amount of data is fully utilized, and the conversion from a data warehouse to data productivity is still a challenge.
The key of structural damage diagnosis is to determine the damage position and quantify the damage degree. Structural damage diagnosis methods based on structural dynamic response signals and artificial intelligence algorithms are favored because of their efficient data processing capabilities, excellent feature extraction capabilities, and the ability to achieve real-time diagnosis of structural damage. The deep learning intelligent algorithm represented by the convolutional neural network achieves remarkable effect in the aspect of structural damage diagnosis, but the convolutional neural network loses much important information in the pooling process, and the scalar neurons used by the convolutional neural network cannot extract pose information such as positions, angles and the like when extracting features. However, in actual engineering, the bridge structure is subjected to the combined action of complex multi-source loads, the variation and uncertainty of the loads can cause similar characteristics of the same type of damage to be reduced, and the diagnostic performance of the convolutional neural network is poor due to the defects, particularly in the structural damage information needing to extract massive monitoring data. In addition, the traditional time-domain signals with time sequences or frequency-domain signals based on Fourier transformation are used as damage identification data, and high instability exists under excitation interference, so that the damage is not sensitive enough, and the damage characteristics of the structure are difficult to fully characterize. Therefore, there is a need for a more efficient structural damage diagnostic method that combines structural vibration response signals with artificial intelligence algorithms to achieve accurate localization of structural damage and quantification of the extent of damage.
Disclosure of Invention
In order to solve the problems, the invention provides a bridge intelligent damage diagnosis method based on the integration of vibration transmission big data and a capsule network by combining a capsule network with a transfer rate function and integrating a plurality of machine vision learning algorithms.
The invention provides the following technical scheme.
A bridge damage diagnosis method for merging vibration transmission big data with a capsule network comprises the following steps:
according to vibration response data of a bridge structure to be predicted, calculating to obtain a vibration transfer rate function, establishing a transfer rate function matrix, and converting the transfer rate function matrix into an image form to obtain a transfer rate image;
constructing a neural network model for positioning and quantifying structural damage based on a capsule network, wherein a feature extraction layer in the capsule network is changed into a dense layer, and a convolution attention module is added after a second convolution layer;
and inputting the transmissibility image into a trained neural network model, and carrying out damage positioning and quantification of the structure to obtain a predicted result of the damage position and degree.
Preferably, the training of the neural network model for structural damage localization and quantification comprises the following steps:
constructing a finite element model of a bridge structure for simulating various damage conditions of the bridge, and extracting vibration response data of the structure in various damage states;
according to the vibration response data, calculating to obtain vibration transfer rate functions under various damage states, establishing a transfer rate function matrix, and converting the transfer rate function matrix into an image form to obtain a transfer rate image;
classifying the transmissibility images, and attaching corresponding damage labels, wherein the label content comprises the damage positions and the damage degrees of the structures; constructing an image sample dataset from the transmissibility image;
and determining a loss function, and inputting an image sample data set to perform neural network model training to obtain a trained neural network model.
Preferably, the vibration response data includes acceleration and displacement, noted as:
x i ＝{x(t n )}，i＝1,2,...k，n＝1,2,...,N
wherein k is the number of data acquisition points, and N is the length of response data.
Preferably, the vibration transfer rate function is:
y i ＝{y(f m )}，m＝1,2,...,M
wherein M is the length of the transmissibility data;
the transfer rate function matrix is:
Y j ＝{y 1 ,y 2 ,...,y i }
wherein j represents a j-th damage condition;
the calculation of the vibration transfer rate function includes:
wherein, the liquid crystal display device comprises a liquid crystal display device,in degree of freedom i 1 Department sum i 2 Transmissibility between responses at +.>Is a cross spectrum，/>Is self-spectrum, x (omega) and x * (omega) is the vibration response x (t) n ) And its conjugate.
Preferably, the transfer rate function matrix converted image is in the form of a contour map.
Preferably, the neural network model for structural damage localization and quantification comprises a first convolution layer, a dense layer, a second convolution layer, an attention module, a capsule layer and a fully connected layer; the dense layer comprises a plurality of groups of dense blocks, and the dense blocks are connected by a transition layer; the capsule layers include a Primary-Caps layer and a second capsule layer.
Preferably, the inputting the transmissibility image into the trained neural network model, performing the damage positioning and quantification of the structure, includes the following steps:
inputting a transmissibility image with a size of c×m×n, c being the number of finger channels, m×n being the image size;
the transmissibility image enters a dense layer after the features are primarily extracted by a first convolution layer, and deep feature extraction is carried out;
the network computation of the dense layer can be expressed as:
x l ＝H l ([x 0 ,x 1 ,...,x l-1 ])
wherein the H () function consists of BN+ReLU+3.3Conv, [ x ] 0 ,x 1 ,...,x l-1 ]The output characteristic diagrams of the layers 0 to l-1 are connected, namely, the combination of channels is performed;
after deep feature extraction is carried out through the dense layer, entering a second convolution layer and initializing capsule input;
adding a convolution attention module after the second convolution, wherein the convolution attention module comprises a channel attention module and a space attention module; different information is respectively utilized by adopting two modes of global average pooling and maximum pooling, spatial characteristic information is summarized, and a spatial characteristic diagram is output;
the space feature map is input into a Primary-Caps layer in a capsule network capsule layer to output a plurality of neurons of a lower capsule network, and a dynamic routing algorithm is utilized to connect the internal space relation between a network deep learning coding part and the whole of information between capsules, so that the update of the plurality of neurons of a higher capsule network is realized;
flattening the high-level features of the second capsule layer, transmitting the flattened high-level features to two full-connection layers, activating the two full-connection layers by adopting a ReLU function, and finally outputting the damage degree values of all units;
and (5) positioning and quantifying the damage of the bridge structure according to the damage degree magnitude of each unit.
The invention has the beneficial effects that:
(1) The invention adopts the transmissibility data to replace the traditional time domain and frequency domain data, constructs a mass transmissibility data set, and converts the signals into images, so that an excellent machine vision learning algorithm can be fully utilized.
(2) According to the invention, vector neurons in the capsule network are utilized to replace scalar neurons in the traditional convolutional neural network, so that the inherent spatial relationship and semantic information in the data can be effectively considered, the gesture information contained in the transmissibility signal can be fully utilized, and the accuracy of the network is improved.
(3) The invention improves the feature extraction layer of the standard capsule network, adopts the Dense Block (Dense Block) to extract the features, and compared with the feature extraction in the standard capsule network by only using a single-layer convolution kernel, the Dense Block can directly connect all layers to each other, ensures the maximum information flow among the layers in the network, removes the original pooling operation in the Dense Block, avoids the loss of information and can extract more key information in the data set;
(4) Introducing a convolution attention module, the convolution attention module will follow two independent dimensions: the channel and space, in turn, infer an attention map, which is then multiplied by the input feature map for adaptive feature optimization, enhancing the extraction of useful features.
Drawings
FIG. 1 is a flow chart of a method of an embodiment of the present invention;
FIG. 2 is a schematic diagram of a neural network framework for structural damage localization and quantification in accordance with an embodiment of the present invention;
FIG. 3 is a diagram of a monorail through-type steel truss bridge model in accordance with an embodiment of the present invention;
FIG. 4 is a raw metrology structure acceleration response signal of an embodiment of the present invention;
FIG. 5 is a graph of the transfer rate function of all points when a structural unit is damaged according to an embodiment of the present invention;
FIG. 6 is a contour plot of transfer rate function transformation for an embodiment of the present invention;
FIG. 7 is a graph showing the localization and quantification of structural damage by a network model according to an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Example 1
The bridge damage diagnosis method for merging vibration transmission big data with a capsule network, as shown in figures 1-7, comprises the following steps:
s1: constructing a finite element model of a bridge structure for simulating various damage conditions of the bridge, and extracting vibration response data of the structure in various damage states; the vibration response data includes acceleration and displacement, noted as: x is x i ＝{x(t n ) I=1, 2,..k, n=1, 2,..n; wherein k is the number of data acquisition points, and N is the length of response data.
S2: according to the vibration response data, vibration transfer rate functions under various damage states are obtained through calculation, a transfer rate function matrix is established, the transfer rate function matrix is converted into an image form, a transfer rate image is obtained, and the image form converted by the transfer rate function matrix is a contour map. Wherein the vibration transfer rate function is: y is i ＝{y(f m ) M=1, 2,; wherein M is the length of the transmissibility data; the transfer rate function matrix is: y is Y j ＝{y 1 ,y 2 ,...,y i -a }; wherein j is a tableShowing a j-th injury condition; calculation of a vibration transfer rate function, comprising:
wherein, the liquid crystal display device comprises a liquid crystal display device,in degree of freedom i 1 Department sum i 2 Transmissibility between responses at +.>Is a cross-spectrum, ->Is self-spectrum, x (omega) and x * (omega) is the vibration response x (t) n ) And its conjugate.
S3: classifying the transmissibility images, and attaching corresponding damage labels, wherein the label content comprises the damage positions and the damage degrees of the structures; constructing an image sample dataset from the transmissibility image; the tag is set according to an example, specifically:
Label＝[0,0,0,15,0,0,30,0,0,0,0,0]
the label indicates that the 4 th unit was 15% damaged, the 7 th unit was 30% damaged, and a total of 12 units were involved in damage diagnosis.
S4: determining a loss function, inputting an image sample data set to perform neural network model training, and obtaining a trained neural network model; the neural network model for positioning and quantifying the structural damage comprises a first convolution layer, a dense layer, a second convolution layer, an attention module, a capsule layer and a full connection layer; the dense layer comprises a plurality of groups of dense blocks, and the dense blocks are connected by a transition layer; the capsule layers include a Primary-Caps layer and a second capsule layer.
S5: and converting vibration response data of the bridge structure to be predicted into a transmissibility image through S2, inputting the transmissibility image into a trained neural network model, and carrying out damage positioning and quantification of the structure to obtain a prediction result of the damage position and degree.
Specific:
s5.1: inputting a transmissibility image with a size of c×m×n, c being the number of finger channels, m×n being the image size; the transitivity image enters a Dense layer after the features are primarily extracted by a first convolution layer, deep feature extraction is carried out, and the size of the feature image can be adjusted and the growth rate in a Dense block Dense block can be determined at the step. The network computation of the dense layer can be expressed as:
x l ＝H l ([x 0 ,x 1 ,...,x l-1 ])
wherein the H () function consists of BN+ReLU+3.3Conv, [ x ] 0 ,x 1 ,...,x l-1 ]The output characteristic diagrams of the layers 0 to l-1 are connected, namely, the combination of channels is performed;
s5.2: after deep feature extraction through the dense layer, enter the second convolution layer and initialize capsule input.
S5.3: adding a convolution attention module after the second convolution, wherein the convolution attention module comprises a channel attention module and a space attention module; different information is respectively utilized by adopting two modes of global average pooling and maximum pooling:
M c (F)＝σ(MLP(AvgPool(F))+MLP(MaxPool(F)))
M s (F)＝σ(f 7*7 ([AvgPool(F),MaxPool(F)]))
and summarizing the spatial feature information and outputting a spatial feature map.
S5.4: the primy-Caps layer inputs the space feature map into the capsule layer of the capsule network to output a plurality of neurons of the capsule network of the lower layer, and the inner space relation between the network deep learning coding part and the whole of the information between the capsules is connected by using a dynamic routing algorithm, so that the update of the plurality of neurons of the capsule network of the higher layer is realized.
S5.5: and flattening the high-level features of the second capsule layer, transmitting the flattened high-level features to two full-connection layers, activating the two full-connection layers by adopting a ReLU function, and finally outputting the damage degree values of all units.
S5.6: and (5) obtaining the positioning and quantification of the damage of the bridge structure according to the damage degree magnitude of each unit, and making a corresponding maintenance decision.
In this embodiment:
in order to verify the effectiveness of the bridge intelligent damage diagnosis method with the integration of vibration transmission big data and a capsule network, an embodiment is provided for explanation.
In this embodiment, a finite element model is built according to structural parameters of an actual bridge, acceleration response of the bridge structure is selected and extracted as vibration analysis data, and analysis is performed according to a transmissibility image database based on vibration transmission big data in S2 and S3.
Referring to fig. 1, in this embodiment, a single-track through-type steel truss bridge is used as the analysis object, and the steel truss bridge is composed of two parallel triangular trusses with a distance of 5.75m and a height of 11m, and the span of the steel truss bridge is 64m, as shown in fig. 3. The vibration load condition of the actual running of the train is simulated by adopting the moving load consisting of 1 DF4D locomotive and 4C 80 trailers to run at the speed of 100km/h, the vertical acceleration response data are extracted, the sampling frequency is 200Hz, the extracted acceleration response time course curve is shown in figure 4, the transfer rate function is calculated according to the transfer rate calculation formula, as shown in figure 5, and the transfer rate function is converted into a corresponding high-line diagram, as shown in figure 6.
In this embodiment, three data sets of single damage, two-rod damage and three-rod damage are established, rod damage is simulated in a form of rigidity reduction, 8640 images are obtained, corresponding damage labels are set for all the images, the data sets divide training sets and testing sets according to the proportion of 0.85:0.15, and training times are set to be 50 times.
After training, the regression effect of the network model is tested by using the test set, the accuracy of the fitting effect of the network to the test set result is high, the average absolute percentage error is within 3%, and the damage degree of the structure can be well predicted. Fig. 7 shows predicted values of three injuries, and as can be seen from fig. 7, the intelligent damage diagnosis system for the bridge well predicts damage components and damage degree values of the structure.
Finally, depending on the damage level of the structure, a suitable maintenance decision is given, for example when the damage level is greater than a certain limit value, the replacement of the component can be taken into account.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, and alternatives falling within the spirit and principles of the invention.

Claims (4)

1. The bridge damage diagnosis method for fusing vibration transmission big data with a capsule network is characterized by comprising the following steps of:
according to vibration response data of a bridge structure to be predicted, calculating to obtain a vibration transfer rate function, establishing a transfer rate function matrix, and converting the transfer rate function matrix into an image form to obtain a transfer rate image;
constructing a neural network model for positioning and quantifying structural damage based on a capsule network, wherein a feature extraction layer in the capsule network is changed into a dense layer, and a convolution attention module is added after a second convolution layer;
inputting the transmissibility image into a trained neural network model, and carrying out damage positioning and quantification of the structure to obtain a predicted result of the damage position and degree;
the vibration transfer rate function is:
y i ＝{y(f m )}，m＝1,2,...,M
wherein M is the length of the transmissibility data;
the transfer rate function matrix is:
Y j ＝{y 1 ,y 2 ,...,y i }
wherein j represents a j-th damage condition;
the calculation of the vibration transfer rate function includes:
wherein, the liquid crystal display device comprises a liquid crystal display device,is free ofDegree i 1 Department sum i 2 Transmissibility between responses at +.>Is a cross-spectrum, ->Is self-spectrum, x (omega) and x * (omega) is the vibration response x (t) n ) Fourier transform of (a) and its conjugate;
the neural network model for positioning and quantifying the structural damage comprises a first convolution layer, a dense layer, a second convolution layer, an attention module, a capsule layer and a full connection layer; the dense layer comprises a plurality of groups of dense blocks, and the dense blocks are connected by a transition layer; the capsule layer comprises a Primary-Caps layer and a second capsule layer;
inputting the transmissibility image into a trained neural network model to perform damage positioning and quantization of the structure, wherein the method comprises the following steps of:
inputting a transmissibility image with a size of c×m×n, c being the number of finger channels, m×n being the image size;
the transmissibility image enters a dense layer after the features are primarily extracted by a first convolution layer, and deep feature extraction is carried out;
the network computation of the dense layer can be expressed as:
x l ＝H l ([x 0 ,x 1 ,...,x l-1 ])
wherein the H () function consists of BN+ReLU+3.3Conv, [ x ] 0 ,x 1 ,...,x l-1 ]The output characteristic diagrams of the layers 0 to l-1 are connected, namely, the combination of channels is performed;
after deep feature extraction is carried out through the dense layer, entering a second convolution layer and initializing capsule input;
adding a convolution attention module after the second convolution, wherein the convolution attention module comprises a channel attention module and a space attention module; different information is respectively utilized by adopting two modes of global average pooling and maximum pooling, spatial characteristic information is summarized, and a spatial characteristic diagram is output;
the space feature map is input into a Primary-Caps layer in a capsule network capsule layer to output a plurality of neurons of a lower capsule network, and a dynamic routing algorithm is utilized to connect the internal space relation between a network deep learning coding part and the whole of information between capsules, so that the update of the plurality of neurons of a higher capsule network is realized;
flattening the high-level features of the second capsule layer, transmitting the flattened high-level features to two full-connection layers, activating the two full-connection layers by adopting a ReLU function, and finally outputting the damage degree values of all units;
and (5) positioning and quantifying the damage of the bridge structure according to the damage degree magnitude of each unit.
2. The bridge damage diagnosis method for merging vibration transmission big data with capsule network according to claim 1, wherein training of the neural network model for structural damage localization and quantification comprises the following steps:
constructing a finite element model of a bridge structure for simulating various damage conditions of the bridge, and extracting vibration response data of the structure in various damage states;
according to the vibration response data, calculating to obtain vibration transfer rate functions under various damage states, establishing a transfer rate function matrix, and converting the transfer rate function matrix into an image form to obtain a transfer rate image;
classifying the transmissibility images, and attaching corresponding damage labels, wherein the label content comprises the damage positions and the damage degrees of the structures; constructing an image sample dataset from the transmissibility image;
and determining a loss function, and inputting an image sample data set to perform neural network model training to obtain a trained neural network model.
3. The bridge damage diagnosis method of claim 1, wherein the vibration response data includes acceleration and displacement, recorded as:
x i ＝{x(t n )}，i＝1,2,...k，n＝1,2,...,N
wherein k is the number of data acquisition points, and N is the length of response data.
4. The bridge damage diagnosis method for merging vibration transmission big data with capsule network according to claim 1, wherein the image form of transfer rate function matrix conversion is a contour map.
CN202210108414.0A 2022-01-28 2022-01-28 Bridge damage diagnosis method for fusion of vibration transmission big data and capsule network Active CN114444187B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN202210108414.0A CN114444187B (en) 2022-01-28 2022-01-28 Bridge damage diagnosis method for fusion of vibration transmission big data and capsule network
GB2218502.9A GB2615402A (en) 2022-01-28 2022-12-08 Bridge damage diagnosis method based on fusion of vibration transfer big data and capsule network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210108414.0A CN114444187B (en) 2022-01-28 2022-01-28 Bridge damage diagnosis method for fusion of vibration transmission big data and capsule network

Publications (2)

Publication Number Publication Date
CN114444187A CN114444187A (en) 2022-05-06
CN114444187B true CN114444187B (en) 2023-07-18

Family

ID=81371107

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210108414.0A Active CN114444187B (en) 2022-01-28 2022-01-28 Bridge damage diagnosis method for fusion of vibration transmission big data and capsule network

Country Status (2)

Country Link
CN (1) CN114444187B (en)
GB (1) GB2615402A (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115410004B (en) * 2022-09-22 2023-06-16 安吉县科声磁性器材有限公司 Intelligent screening system and method for iron powder

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108875705A (en) * 2018-07-12 2018-11-23 广州麦仑信息科技有限公司 A kind of vena metacarpea feature extracting method based on Capsule
CN111461063A (en) * 2020-04-24 2020-07-28 武汉大学 Behavior identification method based on graph convolution and capsule neural network
CN111595541A (en) * 2020-04-27 2020-08-28 河海大学 Multi-dimensional structure damage identification method based on mass vibration transmissibility data convolutional neural network processing
CN111612030A (en) * 2020-03-30 2020-09-01 华电电力科学研究院有限公司 Wind turbine generator blade surface fault identification and classification method based on deep learning
KR20210047832A (en) * 2020-05-08 2021-04-30 베이징 바이두 넷컴 사이언스 앤 테크놀로지 코., 엘티디. Processing method and apparatus of neural network model
CN113255882A (en) * 2021-04-30 2021-08-13 南通大学 Bearing fault diagnosis method based on improved convolution capsule network

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108875705A (en) * 2018-07-12 2018-11-23 广州麦仑信息科技有限公司 A kind of vena metacarpea feature extracting method based on Capsule
CN111612030A (en) * 2020-03-30 2020-09-01 华电电力科学研究院有限公司 Wind turbine generator blade surface fault identification and classification method based on deep learning
CN111461063A (en) * 2020-04-24 2020-07-28 武汉大学 Behavior identification method based on graph convolution and capsule neural network
CN111595541A (en) * 2020-04-27 2020-08-28 河海大学 Multi-dimensional structure damage identification method based on mass vibration transmissibility data convolutional neural network processing
KR20210047832A (en) * 2020-05-08 2021-04-30 베이징 바이두 넷컴 사이언스 앤 테크놀로지 코., 엘티디. Processing method and apparatus of neural network model
CN113255882A (en) * 2021-04-30 2021-08-13 南通大学 Bearing fault diagnosis method based on improved convolution capsule network

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于改进胶囊神经网络的示功图诊断模型;周航等;《电子科技大学学报》;全文 *

Also Published As

Publication number Publication date
GB2615402A (en) 2023-08-09
CN114444187A (en) 2022-05-06
GB202218502D0 (en) 2023-01-25

Similar Documents

Publication Publication Date Title
Sun et al. Review of bridge structural health monitoring aided by big data and artificial intelligence: From condition assessment to damage detection
CN110598736A (en) Power equipment infrared image fault positioning, identifying and predicting method
CN111832814A (en) Air pollutant concentration prediction method based on graph attention machine mechanism
Si et al. Unsupervised deep transfer learning with moment matching: A new intelligent fault diagnosis approach for bearings
CN109215380B (en) Effective parking space prediction method
CN113971660B (en) Computer vision method for bridge health diagnosis and intelligent camera system
CN114548375B (en) Cable-stayed bridge girder dynamic deflection monitoring method based on two-way long-short-term memory neural network
CN116596151B (en) Traffic flow prediction method and computing device based on time-space diagram attention
CN114444187B (en) Bridge damage diagnosis method for fusion of vibration transmission big data and capsule network
CN116307152A (en) Traffic prediction method for space-time interactive dynamic graph attention network
Rosso et al. Review on deep learning in structural health monitoring
CN114818826A (en) Fault diagnosis method based on lightweight Vision Transformer module
Shaowu et al. Study on the health condition monitoring method of hydraulic pump based on convolutional neural network
Chamangard et al. Transfer learning for CNN-based damage detection in civil structures with insufficient data
CN111383273A (en) High-speed rail contact net part positioning method based on improved structure reasoning network
CN116842379A (en) Mechanical bearing residual service life prediction method based on DRSN-CS and BiGRU+MLP models
CN113570111B (en) Bridge health state on-chip monitoring method based on lightweight network
CN113554070B (en) Bearing fault diagnosis method based on transition probability matrix and lightweight network
CN114689351A (en) Equipment fault predictive diagnosis system and method
Kou Wireless communication system and its application in big data remote monitoring and decision-making
CN111462483A (en) Bridge overweight vehicle dynamic identification method based on deep learning
Zhang et al. Structural Vibration Data Anomaly Detection Based on Multiple Feature Information Using CNN-LSTM Model
CN111291490B (en) Nonlinear mapping intelligent modeling method for structure multi-scale heterogeneous response
CN113570478B (en) Bridge health state intelligent evaluation method based on edge calculation
CN114118140B (en) Multi-view intelligent fault diagnosis method and system for explosion-proof motor bearing

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant