CN116481791A - Steel structure connection stability monitoring system and method thereof - Google Patents
Steel structure connection stability monitoring system and method thereof Download PDFInfo
- Publication number
- CN116481791A CN116481791A CN202310517071.8A CN202310517071A CN116481791A CN 116481791 A CN116481791 A CN 116481791A CN 202310517071 A CN202310517071 A CN 202310517071A CN 116481791 A CN116481791 A CN 116481791A
- Authority
- CN
- China
- Prior art keywords
- vibration waveform
- context
- local vibration
- feature vectors
- waveform feature
- 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
- 229910000831 Steel Inorganic materials 0.000 title claims abstract description 146
- 239000010959 steel Substances 0.000 title claims abstract description 146
- 238000012544 monitoring process Methods 0.000 title claims abstract description 58
- 238000000034 method Methods 0.000 title claims abstract description 53
- 239000013598 vector Substances 0.000 claims abstract description 263
- 238000005070 sampling Methods 0.000 claims abstract description 107
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 45
- 238000012545 processing Methods 0.000 claims description 30
- 239000011159 matrix material Substances 0.000 claims description 14
- 230000007246 mechanism Effects 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 10
- 238000012546 transfer Methods 0.000 claims description 8
- 230000004913 activation Effects 0.000 claims description 7
- 238000011176 pooling Methods 0.000 claims description 7
- 238000003062 neural network model Methods 0.000 claims description 4
- 238000009826 distribution Methods 0.000 description 18
- 230000006870 function Effects 0.000 description 12
- 230000008569 process Effects 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 238000000605 extraction Methods 0.000 description 6
- 238000003860 storage Methods 0.000 description 5
- 229910001294 Reinforcing steel Inorganic materials 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000007774 longterm Effects 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000000306 recurrent effect Effects 0.000 description 3
- 230000003252 repetitive effect Effects 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000001149 cognitive effect Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000007477 logistic regression Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000015654 memory Effects 0.000 description 2
- 238000013508 migration Methods 0.000 description 2
- 230000005012 migration Effects 0.000 description 2
- 238000013139 quantization Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 229910000746 Structural steel Inorganic materials 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 239000004567 concrete Substances 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005272 metallurgy Methods 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
- 239000011150 reinforced concrete Substances 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000035939 shock Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Classifications
-
- 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
-
- 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
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Abstract
A steel structure connection stability monitoring system and method thereof are disclosed. Firstly, sliding window sampling based on sampling windows is carried out on vibration signals of connection steel when the vibration signals are impacted, which are acquired by a vibration sensor, so as to obtain a sequence of sampling windows, then, each sampling window in the sequence of sampling windows is processed through a convolutional neural network model so as to obtain a plurality of local vibration waveform characteristic vectors, then, the local vibration waveform characteristic vectors are input into a context encoder based on a converter so as to obtain upper and lower Wen Yuyi associated vibration waveform characteristic vectors, and finally, the upper and lower Wen Yuyi associated vibration waveform characteristic vectors are processed through a classifier so as to obtain a classification result for indicating whether the connection stability of the connection steel meets a preset standard. Thus, the loosening condition of the connecting piece can be accurately identified.
Description
Technical Field
The present application relates to the field of intelligent monitoring, and more particularly, to a system and method for monitoring connection stability of a steel structure.
Background
The steel structure building is a novel building system, breaks through the industrial boundaries among the house industry, the building industry and the metallurgy industry, and is integrated into a novel industry system, namely the steel structure building system which is commonly seen by the industry, compared with the traditional concrete building, the steel structure building has the advantages that steel plates or section steel are used for replacing reinforced concrete, the strength is higher, and the shock resistance is better. And the components can be manufactured in a factory and installed on site, so that the construction period is greatly reduced. Because the steel can be recycled, the construction waste can be greatly reduced, and the method is more environment-friendly.
Through retrieving, chinese patent application number CN201922237154.3 discloses an improved H-section steel connection structure, including two H-section steel I and H-section steel II that wait to connect, two H-section steel I, II wait to connect the tip, the preceding, the back both sides face of its web are welded mutual symmetrical angle steel that connects respectively, wherein one limb board of angle steel cooperatees with this side web face, another limb board terminal surface of angle steel and looks welded H-section steel terminal surface parallel and level.
An improved H-section steel connection structure in the above patent has the following disadvantages: the overall structure promotes the connection and fixation of H-shaped steel, but is not perfect in terms of the stability of the overall structure in the long term, and the connection degree with other structural steel plates is single.
Aiming at the technical problems, chinese patent No. 113944230B discloses a quick connecting assembly of a steel structure and a steel structure building thereof, during operation, each H-shaped steel is clamped into a U-shaped clamping groove of connecting steel, a limiting rod is inserted into a connecting hole and a guide hole, simultaneously, the top circumference and the bottom circumference of the limiting rod are fixed through a fastening nut, the connection and fixation of the H-shaped steel are ensured, meanwhile, a first reinforcing steel plate and a second reinforcing steel plate are inserted into slots of two adjacent connecting steels, the first reinforcing steel plate and the second reinforcing steel plate are relatively short, the connecting steels are connected through mortises and mortises, the problem that the connecting steels shake greatly when the connecting steels collide is avoided, the stability of the H-shaped steel on the adjacent connecting steels is ensured, and finally, the quick connecting mechanism is inserted into a connecting hole of a connecting column, so that other steel structures are assembled.
Although the steel structure connection strength can be enhanced at the structural level by the steel structure quick connection assembly, in the actual working process, the connection structure constructed by the steel structure quick connection assembly can be loosened along with the time, so that the connection stability fluctuates, and the structural safety risk is possibly caused.
Therefore, a steel structure connection stability monitoring scheme is desired.
Disclosure of Invention
The present application has been made in order to solve the above technical problems. Embodiments of the present application provide a steel structure connection stability monitoring system and method thereof. Firstly, sliding window sampling based on sampling windows is carried out on vibration signals of connection steel when the vibration signals are impacted, which are acquired by a vibration sensor, so as to obtain a sequence of sampling windows, then, each sampling window in the sequence of sampling windows is processed through a convolutional neural network model so as to obtain a plurality of local vibration waveform characteristic vectors, then, the local vibration waveform characteristic vectors are input into a context encoder based on a converter so as to obtain upper and lower Wen Yuyi associated vibration waveform characteristic vectors, and finally, the upper and lower Wen Yuyi associated vibration waveform characteristic vectors are processed through a classifier so as to obtain a classification result for indicating whether the connection stability of the connection steel meets a preset standard. Thus, the loosening condition of the connecting piece can be accurately identified.
According to one aspect of the present application, there is provided a steel structure connection stability monitoring method, comprising:
acquiring a vibration signal of the connecting steel acquired by the vibration sensor when impacted;
sampling the vibration signal by a sliding window based on a sampling window to obtain a sequence of sampling windows;
each sampling window in the sequence of sampling windows is passed through a convolutional neural network model serving as a filter to obtain a plurality of local vibration waveform feature vectors;
inputting the plurality of local vibration waveform feature vectors into a converter-based context encoder to obtain a context Wen Yuyi associated vibration waveform feature vector; and
and the upper and lower Wen Yuyi associated vibration waveform characteristic vectors are passed through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the connection stability of the connection steel meets a preset standard.
In the above method for monitoring connection stability of a steel structure, passing each sampling window in the sequence of sampling windows through a convolutional neural network model as a filter to obtain a plurality of local vibration waveform feature vectors, including:
and respectively performing two-dimensional convolution processing, mean pooling processing based on a feature matrix and nonlinear activation processing on input data in forward transfer of layers by using each layer of the convolutional neural network model as a filter to output the plurality of local vibration waveform feature vectors by the last layer of the convolutional neural network model as the filter, wherein the input of the first layer of the convolutional neural network model as the filter is each sampling window in the sequence of sampling windows.
In the above method for monitoring connection stability of a steel structure, inputting the plurality of local vibration waveform feature vectors into a context encoder based on a converter to obtain a vibration waveform feature vector associated with a context Wen Yuyi, the method comprises:
performing context semantic coding based on a self-attention mechanism on the plurality of local vibration waveform feature vectors by using the context encoder based on the converter to obtain a plurality of context local vibration waveform feature vectors;
respectively calculating Gaussian regression uncertainty factors of each contextual local vibration waveform feature vector in the contextual local vibration waveform feature vectors;
taking the Gaussian regression uncertainty factor of each context local vibration waveform characteristic vector as a weight, and respectively weighting each context local vibration waveform characteristic vector in the context local vibration waveform characteristic vectors to obtain a plurality of weighted context local vibration waveform characteristic vectors; and
the plurality of weighted contextual local vibration waveform feature vectors are concatenated to obtain the contextual lower Wen Yuyi associated vibration waveform feature vector.
In the above method for monitoring connection stability of a steel structure, performing context semantic coding on the plurality of local vibration waveform feature vectors based on a self-attention mechanism by using the context encoder based on a converter to obtain a plurality of context local vibration waveform feature vectors, including:
One-dimensional arrangement is carried out on the plurality of local vibration waveform characteristic vectors so as to obtain global vibration waveform characteristic vectors;
calculating the product between the global vibration waveform characteristic vector and the transpose vector of each local vibration waveform characteristic vector in the plurality of local vibration waveform characteristic vectors to obtain a plurality of self-attention correlation matrices;
respectively carrying out standardization processing on each self-attention correlation matrix in the plurality of self-attention correlation matrices to obtain a plurality of standardized self-attention correlation matrices;
obtaining a plurality of probability values by using a Softmax classification function through each normalized self-attention correlation matrix in the normalized self-attention correlation matrices; and
and weighting each local vibration waveform characteristic vector in the local vibration waveform characteristic vectors by taking each probability value in the probability values as a weight so as to obtain the context local vibration waveform characteristic vectors.
In the above method for monitoring connection stability of a steel structure, respectively calculating gaussian regression uncertainty factors of each of the plurality of context local vibration waveform feature vectors, including:
Calculating a gaussian regression uncertainty factor of each of the plurality of contextual local vibration waveform feature vectors, respectively, with the following factor calculation formula;
wherein, the factor calculation formula is:
wherein v is ij Is a feature value of a j-th position of each of the plurality of contextual local vibration waveform feature vectors, L is a length of the feature vector, μ i Sum sigma i 2 The mean and variance of the feature set, respectively, and log is the base 2 logarithm.
In the above method for monitoring connection stability of a steel structure, cascading the plurality of weighted context local vibration waveform feature vectors to obtain the upper and lower Wen Yuyi associated vibration waveform feature vectors includes:
cascading the plurality of weighted context local vibration waveform feature vectors with the following cascading formula to obtain the upper and lower Wen Yuyi associated vibration waveform feature vectors;
wherein, the cascade formula is:
V c ＝Concat[V 1 ，V 2 ，...V n ]
wherein V is 1 ，V 2 ，...V n Representing the plurality of weighted context local vibration waveform feature vectors, concat []Representing a cascade function, V c Representing the upper and lower Wen Yuyi associated vibration waveform feature vectors.
In the above method for monitoring connection stability of steel structure, the step of passing the vibration waveform feature vectors associated with the upper and lower Wen Yuyi through a classifier to obtain a classification result, where the classification result is used to indicate whether connection stability of the connection steel meets a predetermined standard, includes:
performing full-connection coding on the upper Wen Yuyi and lower Wen Yuyi associated vibration waveform feature vectors by using a full-connection layer of the classifier to obtain coded classification feature vectors; and
and inputting the coding classification feature vector into a Softmax classification function of the classifier to obtain the classification result.
According to another aspect of the present application, there is provided a steel structure connection stability monitoring system, comprising:
the signal acquisition module is used for acquiring the vibration signal of the connecting steel when impacted, which is acquired by the vibration sensor;
the sliding window sampling module is used for sampling the vibration signal based on a sliding window of a sampling window to obtain a sequence of the sampling window;
the convolution coding module is used for enabling each sampling window in the sequence of the sampling windows to pass through a convolution neural network model serving as a filter so as to obtain a plurality of local vibration waveform characteristic vectors;
a context encoding module for inputting the plurality of local vibration waveform feature vectors into a transducer-based context encoder to obtain a context Wen Yuyi associated vibration waveform feature vector; and
And the classification module is used for enabling the vibration waveform characteristic vectors associated with the upper Wen Yuyi and the lower Wen Yuyi to pass through a classifier to obtain classification results, wherein the classification results are used for indicating whether the connection stability of the connection steel meets a preset standard or not.
In the above steel structure connection stability monitoring system, the convolutional encoding module is configured to:
and respectively performing two-dimensional convolution processing, mean pooling processing based on a feature matrix and nonlinear activation processing on input data in forward transfer of layers by using each layer of the convolutional neural network model as a filter to output the plurality of local vibration waveform feature vectors by the last layer of the convolutional neural network model as the filter, wherein the input of the first layer of the convolutional neural network model as the filter is each sampling window in the sequence of sampling windows.
In the above steel structure connection stability monitoring system, the context encoding module is configured to:
performing context semantic coding based on a self-attention mechanism on the plurality of local vibration waveform feature vectors by using the context encoder based on the converter to obtain a plurality of context local vibration waveform feature vectors;
Respectively calculating Gaussian regression uncertainty factors of each contextual local vibration waveform feature vector in the contextual local vibration waveform feature vectors;
taking the Gaussian regression uncertainty factor of each context local vibration waveform characteristic vector as a weight, and respectively weighting each context local vibration waveform characteristic vector in the context local vibration waveform characteristic vectors to obtain a plurality of weighted context local vibration waveform characteristic vectors; and
the plurality of weighted contextual local vibration waveform feature vectors are concatenated to obtain the contextual lower Wen Yuyi associated vibration waveform feature vector.
Compared with the prior art, the steel structure connection stability monitoring system and the method thereof are provided. Firstly, sliding window sampling based on sampling windows is carried out on vibration signals of connection steel when the vibration signals are impacted, which are acquired by a vibration sensor, so as to obtain a sequence of sampling windows, then, each sampling window in the sequence of sampling windows is processed through a convolutional neural network model so as to obtain a plurality of local vibration waveform characteristic vectors, then, the local vibration waveform characteristic vectors are input into a context encoder based on a converter so as to obtain upper and lower Wen Yuyi associated vibration waveform characteristic vectors, and finally, the upper and lower Wen Yuyi associated vibration waveform characteristic vectors are processed through a classifier so as to obtain a classification result for indicating whether the connection stability of the connection steel meets a preset standard. Thus, the loosening condition of the connecting piece can be accurately identified.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly introduced below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art. The following drawings are not intended to be drawn to scale, with emphasis instead being placed upon illustrating the principles of the present application.
Fig. 1 is an application scenario diagram of a method for monitoring connection stability of a steel structure according to an embodiment of the present application.
Fig. 2 is a flowchart of a method for monitoring connection stability of a steel structure according to an embodiment of the present application.
Fig. 3 is a schematic diagram of a steel structure connection stability monitoring method according to an embodiment of the present application.
Fig. 4 is a flowchart of substep S140 of the steel structure connection stability monitoring method according to an embodiment of the present application.
Fig. 5 is a flowchart of substep S141 of the steel structure connection stability monitoring method according to an embodiment of the present application.
Fig. 6 is a flowchart of substep S150 of the steel structure connection stability monitoring method according to an embodiment of the present application.
Fig. 7 is a block diagram of a steel structure connection stability monitoring system according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some, but not all embodiments of the present application. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments of the present application without making any inventive effort, are also within the scope of the present application.
As used in this application and in the claims, the terms "a," "an," "the," and/or "the" are not specific to the singular, but may include the plural, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
Although the present application makes various references to certain modules in a system according to embodiments of the present application, any number of different modules may be used and run on a user terminal and/or server. The modules are merely illustrative, and different aspects of the systems and methods may use different modules.
Flowcharts are used in this application to describe the operations performed by systems according to embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in order precisely. Rather, the various steps may be processed in reverse order or simultaneously, as desired. Also, other operations may be added to or removed from these processes.
Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some of the embodiments of the present application and not all of the embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
To above-mentioned technical problem, the technical conception of this application is through carrying out real-time supervision and aassessment to the vibration signal of connection steel when being impacted to vibration sensor gathers to connect stability early warning, can in time discover structural stability risk like this, thereby ensure connection structure's stability, reduce structural safety risk.
Specifically, first, a vibration signal of the connection steel at the time of being impacted, which is acquired by a vibration sensor, is acquired. As described above, the connection stability of the steel structure quick connector is an important guarantee for the safety of the steel structure. However, due to the influence of external factors or time, the connection member may be loosened, resulting in unstable connection structure, thereby affecting the safety of the whole steel structure. In order to detect the connection state of the connection element in time, a reliable and real-time monitoring means is needed. The vibration sensor can sensitively sense the vibration signal in the steel structure and convert the vibration signal into an electric signal to be output, so that the vibration sensor can be used for collecting the vibration signal of the connecting piece in the steel structure. When the connecting piece loosens, the characteristics of the vibration signal can be changed, analysis, processing and judgment are carried out on the characteristics, and the looseness condition of the connecting piece can be accurately identified, so that corresponding measures can be timely taken.
Then, the vibration signal is subjected to sliding window sampling based on a sampling window to obtain a sequence of sampling windows. For the connection monitoring task of the steel structure quick connecting piece, the vibration signal needs to be analyzed and processed so as to accurately judge whether abnormal conditions such as looseness exist in the connecting piece or not. The vibration signal in the steel structure is usually characterized by a time sequence, so that the original vibration signal can be sampled according to a certain time step by adopting a sliding window sampling method based on sampling windows, and the continuous vibration signal is divided into a plurality of sampling windows, so that a sequence of sampling windows is formed. The method can keep the time sequence information of the vibration signal, can reduce the data volume, and is convenient for subsequent processing and analysis. By adopting the sliding window sampling method, the monitored steel structure vibration signal can be fully utilized, and the characteristic information of the monitored steel structure vibration signal can be kept as much as possible. Meanwhile, the proper size of the sampling window and the sliding window step length are determined, the problems of signal overlapping, loss and the like can be avoided, and the processing efficiency and accuracy of the vibration signal are further improved.
Then, each sampling window in the sequence of sampling windows is passed through a convolutional neural network model as a filter to obtain a plurality of local vibration waveform feature vectors. Vibration signals in steel structures often have a variety of complex characteristics, such as frequency, amplitude, phase, etc., which may change under different connection conditions. In order to be able to accurately monitor the connection state in the steel structure, it is necessary to extract its key features from the vibration signal in order to perform effective analysis and judgment. Convolutional Neural Networks (CNNs) are a deep learning model that can automatically extract local features in input data by convolution operations and combine them into higher-level feature representations, and thus CNNs are widely used in the fields of image processing and signal processing.
And each sampling window in the sequence of sampling windows is subjected to characteristic extraction through a convolutional neural network model serving as a filter, and a plurality of local vibration waveform characteristic vectors are obtained from the characteristic extraction. These eigenvectors can reflect different characteristics of the vibration signal, such as vibration frequency, vibration amplitude, etc., in different states of the connection in the steel structure.
The plurality of local vibration waveform feature vectors are then input to a transducer-based context encoder to derive a context Wen Yuyi associated vibration waveform feature vector. In the technical solution of the present application, the local vibration waveform feature vectors are used to represent vibration waveform local features within a predetermined time window, and the vibration signal is a waveform distribution that is continuous in a time sequence space, so in the technical solution of the present application, the local vibration waveform feature vectors are further input into a context encoder based on a converter, where the context encoder based on a converter uses a self-attention mechanism to perform context semantic encoding on the local vibration waveform feature vectors to obtain a plurality of context local vibration waveform feature vectors, and the context local vibration waveform feature vectors are cascaded to obtain the context Wen Yuyi-associated vibration waveform feature vector.
Next, the upper and lower Wen Yuyi associated vibration waveform feature vectors are passed through a classifier to obtain a classification result for indicating whether the connection stability of the connection steel meets a predetermined criterion. The classifier is a machine learning model that can match input data to predefined categories and output category labels corresponding thereto.
In the technical scheme of the application, the upper and lower Wen Yuyi associated vibration waveform feature vectors are passed through a classifier to obtain a classification result, wherein the classifier can be trained according to known connection state information in monitoring data, so that different characteristics of vibration signals in different connection states, such as frequency, amplitude and the like, can be identified, and a class label corresponding to the characteristics can be generated, and indicates whether the connection stability of the connection steel meets a predetermined standard. It should be noted that the classification labels of the classifier can be used to indicate the operational status of the connectors in the steel structure, and also to determine if further inspection and maintenance measures are required.
Specifically, in consideration of source signal noise introduced in the process of signal acquisition of the vibration sensor by the vibration signal, after sampling by the sliding window based on the sampling window, source signal noise exists in each sampling local signal, after extracting signal image semantic features by a convolution neural network model serving as a filter and performing context-dependent encoding of the signal image semantic features by a context encoder based on a converter, gaussian distribution error uncertainty of respective feature distribution is further introduced into a plurality of context local vibration waveform feature vectors obtained by the context encoder based on the converter, so that in consideration of the fact that the context Wen Yuyi associated vibration waveform feature vectors are obtained by directly cascading the plurality of context local vibration waveform feature vectors, direct superposition of the gaussian distribution error uncertainty also causes classification regression errors of the context Wen Yuyi associated vibration waveform feature vectors, and accuracy of classification results obtained by the context Wen Yuyi associated vibration waveform feature vectors by the classifier is affected.
Based on this, in the technical solution of the present application, each of the plurality of contextual local vibration waveform feature vectors is calculated separately, for example denoted as V i Is expressed as:
l is the length of the feature vector, μ i Sum sigma i 2 Respectively the feature sets v ij ∈V i Mean and variance of (v), where v ij Is the feature vector V i Is the eigenvalue of the j-th position of (c), and log is the base 2 logarithm.
Here, for the agnostic regression of the integrated feature set of each context local vibration waveform feature vector in the plurality of context local vibration waveform feature vectors, which may be caused by the distribution uncertainty information of the integrated feature set, scalar measurement of statistical characteristics of the feature set is performed by using the mean value and the variance of the statistical quantization parameter, so that a normal distribution cognitive mode represented by a feature error is expanded to an unknown distribution regression mode, and migration learning based on natural distribution transfer on the feature set scale is realized, so that the classification result obtained by the classifier of the context Wen Yuyi association vibration waveform feature vector is improved by weighting each context local vibration waveform feature vector with the gaussian regression uncertainty factor and cascading the context Wen Yuyi association vibration waveform feature vector, so that the uncertainty correction of the time base of forming the context Wen Yuyi association vibration waveform feature vector is realized, and the classification regression error existing in the context Wen Yuyi association vibration waveform feature vector is corrected.
Fig. 1 is an application scenario diagram of a method for monitoring connection stability of a steel structure according to an embodiment of the present application. As shown in fig. 1, in this application scenario, first, a vibration signal (e.g., D illustrated in fig. 1) of a connection steel (e.g., N illustrated in fig. 1) at the time of impact acquired by a vibration sensor (e.g., C illustrated in fig. 1) is acquired, and then the vibration signal is input to a server (e.g., S illustrated in fig. 1) in which a steel structure connection stability monitoring algorithm is deployed, wherein the server can process the vibration signal using the steel structure connection stability monitoring algorithm to obtain a classification result for indicating whether or not connection stability of the connection steel meets a predetermined criterion.
Having described the basic principles of the present application, various non-limiting embodiments of the present application will now be described in detail with reference to the accompanying drawings.
Fig. 2 is a flowchart of a method for monitoring connection stability of a steel structure according to an embodiment of the present application. As shown in fig. 2, a method for monitoring connection stability of a steel structure according to an embodiment of the present application includes the steps of: s110, acquiring vibration signals of the connecting steel during the collision, which are acquired by a vibration sensor; s120, sampling the vibration signal by a sliding window based on a sampling window to obtain a sequence of sampling windows; s130, passing each sampling window in the sequence of sampling windows through a convolutional neural network model serving as a filter to obtain a plurality of local vibration waveform feature vectors; s140, inputting the plurality of local vibration waveform feature vectors into a context encoder based on a converter to obtain a context Wen Yuyi associated vibration waveform feature vector; and S150, passing the vibration waveform characteristic vectors associated with the upper and lower Wen Yuyi through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the connection stability of the connection steel meets a preset standard.
Fig. 3 is a schematic diagram of a steel structure connection stability monitoring method according to an embodiment of the present application. As shown in fig. 3, in the network architecture, first, a vibration signal of the connection steel at the time of being impacted, which is acquired by a vibration sensor, is acquired; then, sliding window sampling based on a sampling window is carried out on the vibration signal so as to obtain a sequence of sampling windows; then, each sampling window in the sequence of sampling windows is passed through a convolutional neural network model serving as a filter to obtain a plurality of local vibration waveform feature vectors; next, inputting the plurality of local vibration waveform feature vectors into a converter-based context encoder to obtain a context Wen Yuyi associated vibration waveform feature vector; finally, the vibration waveform characteristic vectors associated with the upper and lower Wen Yuyi are passed through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the connection stability of the connection steel meets a preset standard.
More specifically, in step S110, a vibration signal of the connection steel at the time of being impacted, which is acquired by the vibration sensor, is acquired. The vibration sensor can sensitively sense the vibration signal in the steel structure and convert the vibration signal into an electric signal to be output, so that the vibration sensor can be used for collecting the vibration signal of the connecting piece in the steel structure. When the connecting piece loosens, the characteristics of the vibration signal can be changed, analysis, processing and judgment are carried out on the characteristics, and the looseness condition of the connecting piece can be accurately identified, so that corresponding measures can be timely taken.
More specifically, in step S120, the vibration signal is sampled by a sliding window based on a sampling window to obtain a sequence of sampling windows. For the connection monitoring task of the steel structure quick connecting piece, the vibration signal needs to be analyzed and processed so as to accurately judge whether abnormal conditions such as looseness exist in the connecting piece or not. The vibration signal in the steel structure is usually characterized by a time sequence, so that the original vibration signal can be sampled according to a certain time step by adopting a sliding window sampling method based on sampling windows, and the continuous vibration signal is divided into a plurality of sampling windows, so that a sequence of sampling windows is formed. The method can keep the time sequence information of the vibration signal, can reduce the data volume, and is convenient for subsequent processing and analysis. By adopting the sliding window sampling method, the monitored steel structure vibration signal can be fully utilized, and the characteristic information of the monitored steel structure vibration signal can be kept as much as possible. Meanwhile, the proper size of the sampling window and the sliding window step length are determined, the problems of signal overlapping, loss and the like can be avoided, and the processing efficiency and accuracy of the vibration signal are further improved.
More specifically, in step S130, each sampling window in the sequence of sampling windows is passed through a convolutional neural network model as a filter to obtain a plurality of local vibration waveform feature vectors. Vibration signals in steel structures often have a variety of complex characteristics, such as frequency, amplitude, phase, etc., which may change under different connection conditions. In order to be able to accurately monitor the connection state in the steel structure, it is necessary to extract its key features from the vibration signal in order to perform effective analysis and judgment. And each sampling window in the sequence of sampling windows is subjected to characteristic extraction through a convolutional neural network model serving as a filter, and a plurality of local vibration waveform characteristic vectors are obtained from the characteristic extraction. These eigenvectors can reflect different characteristics of the vibration signal, such as vibration frequency, vibration amplitude, etc., in different states of the connection in the steel structure.
It should be appreciated that convolutional neural network (Convolutional Neural Network, CNN) is an artificial neural network and has wide application in the fields of image recognition and the like. The convolutional neural network may include an input layer, a hidden layer, and an output layer, where the hidden layer may include a convolutional layer, a pooling layer, an activation layer, a full connection layer, etc., where the previous layer performs a corresponding operation according to input data, outputs an operation result to the next layer, and obtains a final result after the input initial data is subjected to a multi-layer operation.
Accordingly, in one specific example, passing each sampling window in the sequence of sampling windows through a convolutional neural network model as a filter to obtain a plurality of local vibration waveform feature vectors, comprising: and respectively performing two-dimensional convolution processing, mean pooling processing based on a feature matrix and nonlinear activation processing on input data in forward transfer of layers by using each layer of the convolutional neural network model as a filter to output the plurality of local vibration waveform feature vectors by the last layer of the convolutional neural network model as the filter, wherein the input of the first layer of the convolutional neural network model as the filter is each sampling window in the sequence of sampling windows.
More specifically, in step S140, the plurality of local vibration waveform feature vectors are input to a context encoder based on a converter to obtain a context Wen Yuyi associated vibration waveform feature vector. In the technical solution of the present application, the local vibration waveform feature vectors are used to represent vibration waveform local features within a predetermined time window, and the vibration signal is a waveform distribution that is continuous in a time sequence space, so in the technical solution of the present application, the local vibration waveform feature vectors are further input into a context encoder based on a converter, where the context encoder based on a converter uses a self-attention mechanism to perform context semantic encoding on the local vibration waveform feature vectors to obtain a plurality of context local vibration waveform feature vectors, and the context local vibration waveform feature vectors are cascaded to obtain the context Wen Yuyi-associated vibration waveform feature vector.
It should be appreciated that by the context encoder, the relationship between a certain word segment and other word segments in the vector representation sequence may be analyzed to obtain corresponding feature information. The context encoder aims to mine for hidden patterns between contexts in the word sequence, optionally the encoder comprises: CNN (Convolutional Neural Network ), recurrent NN (RecursiveNeural Network, recurrent neural network), language Model (Language Model), and the like. The CNN-based method has a better extraction effect on local features, but has a poor effect on Long-Term Dependency (Long-Term Dependency) problems in sentences, so Bi-LSTM (Long Short-Term Memory) based encoders are widely used. The repetitive NN processes sentences as a tree structure rather than a sequence, has stronger representation capability in theory, but has the weaknesses of high sample marking difficulty, deep gradient disappearance, difficulty in parallel calculation and the like, so that the repetitive NN is less in practical application. The transducer has a network structure with wide application, has the characteristics of CNN and RNN, has a better extraction effect on global characteristics, and has a certain advantage in parallel calculation compared with RNN (recurrent neural network).
Accordingly, in one specific example, as shown in fig. 4, inputting the plurality of local vibration waveform feature vectors into a context encoder based on a converter to obtain a context Wen Yuyi associated vibration waveform feature vector includes: s141, performing context semantic coding based on a self-attention mechanism on the local vibration waveform feature vectors by using the context encoder based on the converter to obtain a plurality of context local vibration waveform feature vectors; s142, respectively calculating Gaussian regression uncertainty factors of each of the plurality of context local vibration waveform feature vectors; s143, taking Gaussian regression uncertainty factors of the characteristic vectors of each context local vibration waveform as weights, and respectively weighting each of the characteristic vectors of the context local vibration waveform to obtain a plurality of weighted characteristic vectors of the context local vibration waveform; and S144, cascading the plurality of weighted context local vibration waveform feature vectors to obtain the context Wen Yuyi associated vibration waveform feature vector.
Accordingly, in one specific example, as shown in fig. 5, performing context semantic encoding based on a self-attention mechanism on the plurality of local vibration waveform feature vectors using the converter-based context encoder to obtain a plurality of context local vibration waveform feature vectors, including: s1411, performing one-dimensional arrangement on the plurality of local vibration waveform characteristic vectors to obtain global vibration waveform characteristic vectors; s1412, calculating the product between the global vibration waveform feature vector and the transpose vector of each of the plurality of local vibration waveform feature vectors to obtain a plurality of self-attention correlation matrices; s1413, respectively carrying out standardization processing on each self-attention correlation matrix in the plurality of self-attention correlation matrices to obtain a plurality of standardized self-attention correlation matrices; s1414, obtaining a plurality of probability values by using a Softmax classification function through each normalized self-attention correlation matrix in the normalized self-attention correlation matrices; and S1415, weighting each local vibration waveform characteristic vector in the local vibration waveform characteristic vectors by taking each probability value in the probability values as a weight so as to obtain the context local vibration waveform characteristic vectors.
Specifically, in consideration of source signal noise introduced in the process of signal acquisition of the vibration sensor by the vibration signal, after sampling by the sliding window based on the sampling window, source signal noise exists in each sampling local signal, after extracting signal image semantic features by a convolution neural network model serving as a filter and performing context-dependent encoding of the signal image semantic features by a context encoder based on a converter, gaussian distribution error uncertainty of respective feature distribution is further introduced into a plurality of context local vibration waveform feature vectors obtained by the context encoder based on the converter, so that in consideration of the fact that the context Wen Yuyi associated vibration waveform feature vectors are obtained by directly cascading the plurality of context local vibration waveform feature vectors, direct superposition of the gaussian distribution error uncertainty also causes classification regression errors of the context Wen Yuyi associated vibration waveform feature vectors, and accuracy of classification results obtained by the context Wen Yuyi associated vibration waveform feature vectors by the classifier is affected. Based on this, in the technical solution of the present application, a gaussian regression uncertainty factor of each of the plurality of contextual local vibration waveform feature vectors is calculated separately.
Accordingly, in one specific example, separately computing the gaussian regression uncertainty factor for each of the plurality of contextual local vibration waveform feature vectors comprises: calculating a gaussian regression uncertainty factor of each of the plurality of contextual local vibration waveform feature vectors, respectively, with the following factor calculation formula; wherein, the factor calculation formula is:
wherein v is ij Is the plurality of up and downThe eigenvalue of the j-th position of each contextual local vibration waveform eigenvector in the contextual local vibration waveform eigenvectors, L is the length of the eigenvector, μ i Sum sigma i 2 The mean and variance of the feature set, respectively, and log is the base 2 logarithm.
Here, for the agnostic regression of the integrated feature set of each context local vibration waveform feature vector in the plurality of context local vibration waveform feature vectors, which may be caused by the distribution uncertainty information of the integrated feature set, scalar measurement of statistical characteristics of the feature set is performed by using the mean value and the variance of the statistical quantization parameter, so that a normal distribution cognitive mode represented by a feature error is expanded to an unknown distribution regression mode, and migration learning based on natural distribution transfer on the feature set scale is realized, so that the classification result obtained by the classifier of the context Wen Yuyi association vibration waveform feature vector is improved by weighting each context local vibration waveform feature vector with the gaussian regression uncertainty factor and cascading the context Wen Yuyi association vibration waveform feature vector, so that the uncertainty correction of the time base of forming the context Wen Yuyi association vibration waveform feature vector is realized, and the classification regression error existing in the context Wen Yuyi association vibration waveform feature vector is corrected.
Accordingly, in one specific example, concatenating the plurality of weighted contextual local vibration waveform feature vectors to obtain the contextual lower Wen Yuyi associated vibration waveform feature vectors includes: cascading the plurality of weighted context local vibration waveform feature vectors with the following cascading formula to obtain the upper and lower Wen Yuyi associated vibration waveform feature vectors; wherein, the cascade formula is:
V c ＝Concat[V 1 ，V 2 ，...V n ]
wherein V is 1 ，V 2 ，...V n Representing the plurality of weighted contextual local vibration waveform featuresSyndrome vector, concay [. Cndot.]Representing a cascade function, V c Representing the upper and lower Wen Yuyi associated vibration waveform feature vectors.
More specifically, in step S150, the upper and lower Wen Yuyi associated vibration waveform feature vectors are passed through a classifier to obtain a classification result indicating whether or not the connection stability of the connection steel meets a predetermined criterion. The classifier can be trained according to the known connection state information in the monitoring data, so that different characteristics of vibration signals in different connection states, such as frequency, amplitude and the like, can be identified, and class labels corresponding to the characteristics are generated, and whether the connection stability of the connection steel meets the preset standard is indicated. It should be noted that the classification labels of the classifier can be used to indicate the operational status of the connectors in the steel structure, and also to determine if further inspection and maintenance measures are required.
That is, in the technical solution of the present application, the tag of the classifier includes that the connection stability of the connection steel meets a predetermined criterion (first tag), and that the connection stability of the connection steel does not meet a predetermined criterion (second tag), wherein the classifier determines to which classification tag the classification feature vector belongs through a soft maximum function. It should be noted that the first tag p1 and the second tag p2 do not include a concept set by human, and in fact, during the training process, the computer model does not have a concept of "whether the connection stability of the connection steel meets a predetermined criterion", which is simply that there are two kinds of classification tags and the probability that the output feature is under the two kinds of classification tags, that is, the sum of p1 and p2 is one. Therefore, the classification result of whether the connection stability of the connection steel meets the predetermined standard is actually converted into a classification probability distribution conforming to the natural rule through classifying the tag, and the physical meaning of the natural probability distribution of the tag is essentially used instead of the language text meaning of whether the connection stability of the connection steel meets the predetermined standard.
It should be appreciated that the role of the classifier is to learn the classification rules and classifier using a given class, known training data, and then classify (or predict) the unknown data. Logistic regression (logistics), SVM, etc. are commonly used to solve the classification problem, and for multi-classification problems (multi-class classification), logistic regression or SVM can be used as well, but multiple bi-classifications are required to compose multiple classifications, but this is error-prone and inefficient, and the commonly used multi-classification method is the Softmax classification function.
Accordingly, in one specific example, as shown in fig. 6, the upper and lower Wen Yuyi associated vibration waveform feature vectors are passed through a classifier to obtain a classification result for indicating whether the connection stability of the connection steel meets a predetermined criterion, including: s151, performing full-connection coding on the upper and lower Wen Yuyi associated vibration waveform feature vectors by using a full-connection layer of the classifier to obtain coded classification feature vectors; and S152, inputting the coding classification feature vector into a Softmax classification function of the classifier to obtain the classification result.
In summary, according to the method for monitoring connection stability of a steel structure according to the embodiment of the application, firstly, sliding window sampling based on sampling windows is performed on vibration signals of connection steel when the connection steel is impacted, which are acquired by a vibration sensor, so as to obtain a sequence of sampling windows, then, each sampling window in the sequence of sampling windows is processed through a convolutional neural network model to obtain a plurality of local vibration waveform feature vectors, then, the plurality of local vibration waveform feature vectors are input into a context encoder based on a converter to obtain context Wen Yuyi associated vibration waveform feature vectors, and finally, the context Wen Yuyi associated vibration waveform feature vectors are processed through a classifier to obtain a classification result for indicating whether connection stability of the connection steel meets a predetermined standard. Thus, the loosening condition of the connecting piece can be accurately identified.
Fig. 7 is a block diagram of a steel structure connection stability monitoring system 100 according to an embodiment of the present application. As shown in fig. 7, a steel structure connection stability monitoring system 100 according to an embodiment of the present application includes: a signal acquisition module 110 for acquiring a vibration signal of the connection steel at the time of being impacted, which is acquired by the vibration sensor; a sliding window sampling module 120, configured to perform sliding window sampling based on a sampling window on the vibration signal to obtain a sequence of sampling windows; a convolutional encoding module 130, configured to pass each sampling window in the sequence of sampling windows through a convolutional neural network model that is a filter to obtain a plurality of local vibration waveform feature vectors; a context encoding module 140 for inputting the plurality of local vibration waveform feature vectors into a transducer-based context encoder to obtain a context Wen Yuyi associated vibration waveform feature vector; and a classification module 150, configured to pass the vibration waveform feature vectors associated with the upper and lower Wen Yuyi through a classifier to obtain a classification result, where the classification result is used to indicate whether the connection stability of the connection steel meets a predetermined criterion.
In one example, in the steel structure connection stability monitoring system 100, the convolutional encoding module 130 is configured to: and respectively performing two-dimensional convolution processing, mean pooling processing based on a feature matrix and nonlinear activation processing on input data in forward transfer of layers by using each layer of the convolutional neural network model as a filter to output the plurality of local vibration waveform feature vectors by the last layer of the convolutional neural network model as the filter, wherein the input of the first layer of the convolutional neural network model as the filter is each sampling window in the sequence of sampling windows.
In one example, in the steel structure connection stability monitoring system 100 described above, the context encoding module 140 is configured to: performing context semantic coding based on a self-attention mechanism on the plurality of local vibration waveform feature vectors by using the context encoder based on the converter to obtain a plurality of context local vibration waveform feature vectors; respectively calculating Gaussian regression uncertainty factors of each contextual local vibration waveform feature vector in the contextual local vibration waveform feature vectors; taking the Gaussian regression uncertainty factor of each context local vibration waveform characteristic vector as a weight, and respectively weighting each context local vibration waveform characteristic vector in the context local vibration waveform characteristic vectors to obtain a plurality of weighted context local vibration waveform characteristic vectors; and concatenating the plurality of weighted contextual local vibration waveform feature vectors to obtain the contextual Wen Yuyi associated vibration waveform feature vector.
In one example, in the above steel structure connection stability monitoring system 100, performing context semantic encoding of the plurality of local vibration waveform feature vectors based on a self-attention mechanism using the converter-based context encoder to obtain a plurality of context local vibration waveform feature vectors, comprising: one-dimensional arrangement is carried out on the plurality of local vibration waveform characteristic vectors so as to obtain global vibration waveform characteristic vectors; calculating the product between the global vibration waveform characteristic vector and the transpose vector of each local vibration waveform characteristic vector in the plurality of local vibration waveform characteristic vectors to obtain a plurality of self-attention correlation matrices; respectively carrying out standardization processing on each self-attention correlation matrix in the plurality of self-attention correlation matrices to obtain a plurality of standardized self-attention correlation matrices; obtaining a plurality of probability values by using a Softmax classification function through each normalized self-attention correlation matrix in the normalized self-attention correlation matrices; and weighting each local vibration waveform feature vector in the local vibration waveform feature vectors by taking each probability value in the probability values as a weight so as to obtain the context local vibration waveform feature vectors.
In one example, in the above steel structure connection stability monitoring system 100, calculating the gaussian regression uncertainty factor of each of the plurality of contextual local vibration waveform feature vectors, respectively, includes: calculating a gaussian regression uncertainty factor of each of the plurality of contextual local vibration waveform feature vectors, respectively, with the following factor calculation formula; wherein, the factor calculation formula is:
wherein v is ij Is each contextual local vibration waveform feature vector of the plurality of contextual local vibration waveform feature vectorsThe eigenvalue of the j-th position, L is the length of the eigenvector, μ i Sum sigma i 2 The mean and variance of the feature set, respectively, and log is the base 2 logarithm.
In one example, in the steel structure connection stability monitoring system 100, cascading the plurality of weighted contextual local vibration waveform feature vectors to obtain the upper and lower Wen Yuyi associated vibration waveform feature vectors includes: cascading the plurality of weighted context local vibration waveform feature vectors with the following cascading formula to obtain the upper and lower Wen Yuyi associated vibration waveform feature vectors; wherein, the cascade formula is:
V c ＝Concat[V 1 ，V 2 ，...V n ]
Wherein V is 1 ，V 2 ，...V n Representing the plurality of weighted context local vibration waveform feature vectors, concat []Representing a cascade function, V c Representing the upper and lower Wen Yuyi associated vibration waveform feature vectors.
In one example, in the steel structure connection stability monitoring system 100 described above, the classification module 150 is configured to: performing full-connection coding on the upper Wen Yuyi and lower Wen Yuyi associated vibration waveform feature vectors by using a full-connection layer of the classifier to obtain coded classification feature vectors; and inputting the coding classification feature vector into a Softmax classification function of the classifier to obtain the classification result.
Here, it will be understood by those skilled in the art that the specific functions and operations of the respective modules in the above-described steel structure connection stability monitoring system 100 have been described in detail in the above description of the steel structure connection stability monitoring method with reference to fig. 1 to 6, and thus, repetitive descriptions thereof will be omitted.
As described above, the steel structure connection stability monitoring system 100 according to the embodiment of the present application may be implemented in various wireless terminals, such as a server or the like having a steel structure connection stability monitoring algorithm. In one example, the steel structure connection stability monitoring system 100 according to embodiments of the present application may be integrated into a wireless terminal as one software module and/or hardware module. For example, the steel structure connection stability monitoring system 100 may be a software module in the operating system of the wireless terminal, or may be an application developed for the wireless terminal; of course, the steel structure connection stability monitoring system 100 may also be one of many hardware modules of the wireless terminal.
Alternatively, in another example, the steel structure connection stability monitoring system 100 and the wireless terminal may be separate devices, and the steel structure connection stability monitoring system 100 may be connected to the wireless terminal through a wired and/or wireless network and transmit interactive information in a agreed data format.
According to another aspect of the present application, there is also provided a non-volatile computer-readable storage medium having stored thereon computer-readable instructions which, when executed by a computer, can perform a method as described above.
Program portions of the technology may be considered to be "products" or "articles of manufacture" in the form of executable code and/or associated data, embodied or carried out by a computer readable medium. A tangible, persistent storage medium may include any memory or storage used by a computer, processor, or similar device or related module. Such as various semiconductor memories, tape drives, disk drives, or the like, capable of providing storage functionality for software.
All or a portion of the software may sometimes communicate over a network, such as the internet or other communication network. Such communication may load software from one computer device or processor to another. For example: a hardware platform loaded from a server or host computer of the video object detection device to a computer environment, or other computer environment implementing the system, or similar functioning system related to providing information needed for object detection. Thus, another medium capable of carrying software elements may also be used as a physical connection between local devices, such as optical, electrical, electromagnetic, etc., propagating through cable, optical cable, air, etc. Physical media used for carrier waves, such as electrical, wireless, or optical, may also be considered to be software-bearing media. Unless limited to a tangible "storage" medium, other terms used herein to refer to a computer or machine "readable medium" mean any medium that participates in the execution of any instructions by a processor.
This application uses specific words to describe embodiments of the application. Reference to "a first/second embodiment," "an embodiment," and/or "some embodiments" means that a particular feature, structure, or characteristic is associated with at least one embodiment of the present application. Thus, it should be emphasized and should be appreciated that two or more references to "an embodiment" or "one embodiment" or "an alternative embodiment" in various positions in this specification are not necessarily referring to the same embodiment. Furthermore, certain features, structures, or characteristics of one or more embodiments of the present application may be combined as suitable.
Furthermore, those skilled in the art will appreciate that the various aspects of the invention are illustrated and described in the context of a number of patentable categories or circumstances, including any novel and useful procedures, machines, products, or materials, or any novel and useful modifications thereof. Accordingly, aspects of the present application may be performed entirely by hardware, entirely by software (including firmware, resident software, micro-code, etc.) or by a combination of hardware and software. The above hardware or software may be referred to as a "data block," module, "" engine, "" unit, "" component, "or" system. Furthermore, aspects of the present application may take the form of a computer product, comprising computer-readable program code, embodied in one or more computer-readable media.
Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
The foregoing is illustrative of the present invention and is not to be construed as limiting thereof. Although a few exemplary embodiments of this invention have been described, those skilled in the art will readily appreciate that many modifications are possible in the exemplary embodiments without materially departing from the novel teachings and advantages of this invention. Accordingly, all such modifications are intended to be included within the scope of this invention as defined in the following claims. It is to be understood that the foregoing is illustrative of the present invention and is not to be construed as limited to the specific embodiments disclosed, and that modifications to the disclosed embodiments, as well as other embodiments, are intended to be included within the scope of the appended claims. The invention is defined by the claims and their equivalents.
Claims (10)
1. A method for monitoring connection stability of a steel structure, comprising:
acquiring a vibration signal of the connecting steel acquired by the vibration sensor when impacted;
sampling the vibration signal by a sliding window based on a sampling window to obtain a sequence of sampling windows;
each sampling window in the sequence of sampling windows is passed through a convolutional neural network model serving as a filter to obtain a plurality of local vibration waveform feature vectors;
inputting the plurality of local vibration waveform feature vectors into a converter-based context encoder to obtain a context Wen Yuyi associated vibration waveform feature vector; and
and the upper and lower Wen Yuyi associated vibration waveform characteristic vectors are passed through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the connection stability of the connection steel meets a preset standard.
2. The method of claim 1, wherein passing each sampling window in the sequence of sampling windows through a convolutional neural network model as a filter to obtain a plurality of local vibration waveform feature vectors, comprises:
and respectively performing two-dimensional convolution processing, mean pooling processing based on a feature matrix and nonlinear activation processing on input data in forward transfer of layers by using each layer of the convolutional neural network model as a filter to output the plurality of local vibration waveform feature vectors by the last layer of the convolutional neural network model as the filter, wherein the input of the first layer of the convolutional neural network model as the filter is each sampling window in the sequence of sampling windows.
3. The method of claim 2, wherein inputting the plurality of local vibration waveform feature vectors into a transducer-based context encoder to obtain a context Wen Yuyi-associated vibration waveform feature vector, comprises:
performing context semantic coding based on a self-attention mechanism on the plurality of local vibration waveform feature vectors by using the context encoder based on the converter to obtain a plurality of context local vibration waveform feature vectors;
respectively calculating Gaussian regression uncertainty factors of each contextual local vibration waveform feature vector in the contextual local vibration waveform feature vectors;
taking the Gaussian regression uncertainty factor of each context local vibration waveform characteristic vector as a weight, and respectively weighting each context local vibration waveform characteristic vector in the context local vibration waveform characteristic vectors to obtain a plurality of weighted context local vibration waveform characteristic vectors; and
the plurality of weighted contextual local vibration waveform feature vectors are concatenated to obtain the contextual lower Wen Yuyi associated vibration waveform feature vector.
4. The method of claim 3, wherein performing context semantic encoding of the plurality of local vibration waveform feature vectors using the transducer-based context encoder to obtain a plurality of context local vibration waveform feature vectors based on a self-attention mechanism comprises:
one-dimensional arrangement is carried out on the plurality of local vibration waveform characteristic vectors so as to obtain global vibration waveform characteristic vectors;
calculating the product between the global vibration waveform characteristic vector and the transpose vector of each local vibration waveform characteristic vector in the plurality of local vibration waveform characteristic vectors to obtain a plurality of self-attention correlation matrices;
respectively carrying out standardization processing on each self-attention correlation matrix in the plurality of self-attention correlation matrices to obtain a plurality of standardized self-attention correlation matrices;
obtaining a plurality of probability values by using a Softmax classification function through each normalized self-attention correlation matrix in the normalized self-attention correlation matrices; and
and weighting each local vibration waveform characteristic vector in the local vibration waveform characteristic vectors by taking each probability value in the probability values as a weight so as to obtain the context local vibration waveform characteristic vectors.
5. The method of claim 4, wherein separately calculating gaussian regression uncertainty factors for each of the plurality of contextual local vibration waveform feature vectors comprises:
calculating a gaussian regression uncertainty factor of each of the plurality of contextual local vibration waveform feature vectors, respectively, with the following factor calculation formula;
wherein, the factor calculation formula is:
wherein v is ij Is each contextual local vibration waveform of the plurality of contextual local vibration waveform feature vectorsThe eigenvalue of the j-th position of the eigenvector, L is the length of the eigenvector, μ i Sum sigma i 2 The mean and variance of the feature set, respectively, and log is the base 2 logarithm.
6. The method of claim 5, wherein cascading the plurality of weighted contextual local vibration waveform feature vectors to obtain the upper and lower Wen Yuyi associated vibration waveform feature vectors comprises:
cascading the plurality of weighted context local vibration waveform feature vectors with the following cascading formula to obtain the upper and lower Wen Yuyi associated vibration waveform feature vectors;
Wherein, the cascade formula is:
V c ＝Concat[V 1 ，V 2 ，...V n ]
wherein V is 1 ，V 2 ，...V n Representing the plurality of weighted context local vibration waveform feature vectors, concat []Representing a cascade function, V c Representing the upper and lower Wen Yuyi associated vibration waveform feature vectors.
7. The method of claim 6, wherein the step of passing the upper and lower Wen Yuyi-associated vibration waveform feature vectors through a classifier to obtain a classification result indicating whether the connection stability of the connection steel meets a predetermined criterion, comprises:
performing full-connection coding on the upper Wen Yuyi and lower Wen Yuyi associated vibration waveform feature vectors by using a full-connection layer of the classifier to obtain coded classification feature vectors; and
and inputting the coding classification feature vector into a Softmax classification function of the classifier to obtain the classification result.
8. A steel structure connection stability monitoring system, comprising:
the signal acquisition module is used for acquiring the vibration signal of the connecting steel when impacted, which is acquired by the vibration sensor;
the sliding window sampling module is used for sampling the vibration signal based on a sliding window of a sampling window to obtain a sequence of the sampling window;
The convolution coding module is used for enabling each sampling window in the sequence of the sampling windows to pass through a convolution neural network model serving as a filter so as to obtain a plurality of local vibration waveform characteristic vectors;
a context encoding module for inputting the plurality of local vibration waveform feature vectors into a transducer-based context encoder to obtain a context Wen Yuyi associated vibration waveform feature vector; and
and the classification module is used for enabling the vibration waveform characteristic vectors associated with the upper Wen Yuyi and the lower Wen Yuyi to pass through a classifier to obtain classification results, wherein the classification results are used for indicating whether the connection stability of the connection steel meets a preset standard or not.
9. The steel structure connection stability monitoring system of claim 8, wherein the convolutional encoding module is configured to:
and respectively performing two-dimensional convolution processing, mean pooling processing based on a feature matrix and nonlinear activation processing on input data in forward transfer of layers by using each layer of the convolutional neural network model as a filter to output the plurality of local vibration waveform feature vectors by the last layer of the convolutional neural network model as the filter, wherein the input of the first layer of the convolutional neural network model as the filter is each sampling window in the sequence of sampling windows.
10. The steel structure connection stability monitoring system of claim 9, wherein the context encoding module is configured to:
performing context semantic coding based on a self-attention mechanism on the plurality of local vibration waveform feature vectors by using the context encoder based on the converter to obtain a plurality of context local vibration waveform feature vectors;
respectively calculating Gaussian regression uncertainty factors of each contextual local vibration waveform feature vector in the contextual local vibration waveform feature vectors;
taking the Gaussian regression uncertainty factor of each context local vibration waveform characteristic vector as a weight, and respectively weighting each context local vibration waveform characteristic vector in the context local vibration waveform characteristic vectors to obtain a plurality of weighted context local vibration waveform characteristic vectors; and
the plurality of weighted contextual local vibration waveform feature vectors are concatenated to obtain the contextual lower Wen Yuyi associated vibration waveform feature vector.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310517071.8A CN116481791A (en) | 2023-05-09 | 2023-05-09 | Steel structure connection stability monitoring system and method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310517071.8A CN116481791A (en) | 2023-05-09 | 2023-05-09 | Steel structure connection stability monitoring system and method thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116481791A true CN116481791A (en) | 2023-07-25 |
Family
ID=87219480
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310517071.8A Pending CN116481791A (en) | 2023-05-09 | 2023-05-09 | Steel structure connection stability monitoring system and method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116481791A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117034093A (en) * | 2023-10-10 | 2023-11-10 | 尚宁智感(北京)科技有限公司 | Intrusion signal identification method based on optical fiber system |
| CN117649154A (en) * | 2024-01-29 | 2024-03-05 | 新疆三联工程建设有限责任公司 | Concrete test block manufacturing whole process management system and method based on digitization |
| CN117649154B (en) * | 2024-01-29 | 2024-04-19 | 新疆三联工程建设有限责任公司 | Concrete test block manufacturing whole process management system and method based on digitization |
-
2023
- 2023-05-09 CN CN202310517071.8A patent/CN116481791A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117034093A (en) * | 2023-10-10 | 2023-11-10 | 尚宁智感(北京)科技有限公司 | Intrusion signal identification method based on optical fiber system |
| CN117649154A (en) * | 2024-01-29 | 2024-03-05 | 新疆三联工程建设有限责任公司 | Concrete test block manufacturing whole process management system and method based on digitization |
| CN117649154B (en) * | 2024-01-29 | 2024-04-19 | 新疆三联工程建设有限责任公司 | Concrete test block manufacturing whole process management system and method based on digitization |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109034368B (en) | DNN-based complex equipment multiple fault diagnosis method | |
| CN113642754B (en) | Complex industrial process fault prediction method based on RF noise reduction self-coding information reconstruction and time convolution network | |
| Yu et al. | Automated damage diagnosis of concrete jack arch beam using optimized deep stacked autoencoders and multi-sensor fusion | |
| CN114757309B (en) | Multi-physical-field monitoring data collaborative fusion engineering disaster early warning method and system | |
| CN113778894B (en) | Method, device, equipment and storage medium for constructing test cases | |
| CN111275198A (en) | Bearing abnormity detection method and system | |
| CN109298633A (en) | Chemical production process fault monitoring method based on adaptive piecemeal Non-negative Matrix Factorization | |
| CN116481791A (en) | Steel structure connection stability monitoring system and method thereof | |
| Arul et al. | Data anomaly detection for structural health monitoring of bridges using shapelet transform | |
| CN116992226A (en) | Water pump motor fault detection method and system | |
| CN116482524A (en) | Power transmission and distribution switch state detection method and system | |
| Ye et al. | A deep learning-based method for automatic abnormal data detection: Case study for bridge structural health monitoring | |
| CN111290953B (en) | Method and device for analyzing test logs | |
| CN117041017A (en) | Intelligent operation and maintenance management method and system for data center | |
| CN117034123A (en) | Fault monitoring system and method for fitness equipment | |
| CN116842379A (en) | Mechanical bearing residual service life prediction method based on DRSN-CS and BiGRU+MLP models | |
| CN115017015B (en) | Method and system for detecting abnormal behavior of program in edge computing environment | |
| CN113255771B (en) | Fault diagnosis method and system based on multi-dimensional heterogeneous difference analysis | |
| CN111160419B (en) | Deep learning-based electronic transformer data classification prediction method and device | |
| CN113919540A (en) | Method for monitoring running state of production process and related equipment | |
| CN113157561A (en) | Defect prediction method for numerical control system software module | |
| CN111883226A (en) | Information processing and model training method, device, equipment and storage medium | |
| Cabrera et al. | Combining reservoir computing and variational inference for efficient one-class learning on dynamical systems | |
| CN116232761B (en) | Method and system for detecting abnormal network traffic based on shapelet | |
| Sony | Bridge damage identification using deep learning-based Convolutional Neural Networks (CNNs) |
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 |