CN115356397A - Steel pipe concrete structure void defect diagnosis method and device based on sound signals - Google Patents
Steel pipe concrete structure void defect diagnosis method and device based on sound signals Download PDFInfo
- Publication number
- CN115356397A CN115356397A CN202211017584.4A CN202211017584A CN115356397A CN 115356397 A CN115356397 A CN 115356397A CN 202211017584 A CN202211017584 A CN 202211017584A CN 115356397 A CN115356397 A CN 115356397A
- Authority
- CN
- China
- Prior art keywords
- steel pipe
- concrete structure
- pipe concrete
- void defect
- sound signals
- 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 96
- 239000010959 steel Substances 0.000 title claims abstract description 96
- 239000011800 void material Substances 0.000 title claims abstract description 82
- 230000007547 defect Effects 0.000 title claims abstract description 70
- 230000005236 sound signal Effects 0.000 title claims abstract description 61
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000003745 diagnosis Methods 0.000 title claims abstract description 24
- 238000012549 training Methods 0.000 claims abstract description 27
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 22
- 238000012360 testing method Methods 0.000 claims abstract description 15
- 238000001514 detection method Methods 0.000 claims abstract description 14
- 238000012795 verification Methods 0.000 claims abstract description 14
- 238000007781 pre-processing Methods 0.000 claims abstract description 9
- 230000006870 function Effects 0.000 claims description 37
- 238000011176 pooling Methods 0.000 claims description 23
- 238000010586 diagram Methods 0.000 claims description 19
- 230000009466 transformation Effects 0.000 claims description 17
- 230000008569 process Effects 0.000 claims description 10
- 210000002569 neuron Anatomy 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 7
- 230000008447 perception Effects 0.000 claims description 5
- 230000006835 compression Effects 0.000 claims description 3
- 238000007906 compression Methods 0.000 claims description 3
- 230000002093 peripheral effect Effects 0.000 claims description 3
- 230000004913 activation Effects 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 claims description 2
- 230000001419 dependent effect Effects 0.000 claims description 2
- 230000009286 beneficial effect Effects 0.000 abstract description 3
- 238000004590 computer program Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 238000006073 displacement reaction Methods 0.000 description 2
- 210000005069 ears Anatomy 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 230000003213 activating effect Effects 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 239000000919 ceramic Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000005553 drilling Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000002203 pretreatment Methods 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/04—Analysing solids
- G01N29/045—Analysing solids by imparting shocks to the workpiece and detecting the vibrations or the acoustic waves caused by the shocks
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/44—Processing the detected response signal, e.g. electronic circuits specially adapted therefor
- G01N29/4481—Neural 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
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/02—Indexing codes associated with the analysed material
- G01N2291/023—Solids
- G01N2291/0232—Glass, ceramics, concrete or stone
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Immunology (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Pathology (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Signal Processing (AREA)
- Acoustics & Sound (AREA)
- Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)
Abstract
The invention relates to a steel pipe concrete structure void defect diagnosis method and a device based on sound signals, wherein the method comprises the following steps: 1) Acquiring sound signals of the steel pipe concrete structure in different void states; 2) Reading information in the sound signals, performing continuous wavelet transform by adopting a CMOR wavelet basis function, and converting the one-dimensional sound signals into a two-dimensional wavelet time-frequency graph as image characteristics for void defect identification; 3) Preprocessing the acquired image characteristics of the steel pipe concrete structure void defect identification; 4) Dividing the image characteristics of void defect recognition into a training set, a verification set and a test set according to a set proportion, and training a convolutional neural network model for identifying void defects of the steel pipe concrete structure; 5) And inputting the image characteristics of the concrete-filled steel tube member to be detected into the trained convolutional neural network model to obtain a detection result. The method and the device are beneficial to accurately and efficiently automatically diagnosing the steel pipe concrete structure void defect.
Description
Technical Field
The invention belongs to the technical field of structure detection, and particularly relates to a steel pipe concrete structure void defect diagnosis method and device based on sound signals.
Background
The steel pipe concrete structure has excellent mechanical performance and may be used widely in large engineering. However, the steel pipe concrete structure has a void defect inside due to objective factors such as construction process, material characteristics, service environment, and the like. As the void defect can weaken and even destroy the 'combination effect', the mechanical performance indexes of the steel pipe concrete structure, such as the ultimate bearing capacity, the rigidity, the ductility and the like, are reduced, thereby causing structural damage and diseases, possibly causing structural collapse in severe cases, and causing serious economic loss and serious casualties.
The premise of realizing scientific control on the void defect is to accurately detect and identify the void defect. At present, the existing steel pipe concrete void defect detection technology, such as a manual knocking method, a core drilling sampling method, an ultrasonic method, a piezoelectric ceramic method and the like, has the defects of large empirical error, breakage, low efficiency, pre-embedding of sensors and the like. In particular, the manual knocking detection method commonly used in the engineering field adopts the ears to screen the concrete filled steel tube knocking vibration sound and judge whether the structure is empty or not, and the problem of erroneous judgment or missed judgment caused by insufficient experience due to the influence of human factors exists. Meanwhile, for the defect of void inside the steel pipe, the sound signal generated by knocking the outer steel pipe may pass through the reflection process of the steel pipe-air-concrete-steel pipe, or may be directly diffracted along the steel pipe and then collected by the sensor, so that the sound signal is very complex, and how to accurately extract the characteristics of the sound signal is the key to realize accurate diagnosis.
Disclosure of Invention
The invention aims to provide a steel pipe concrete structure void defect diagnosis method and device based on sound signals, which are beneficial to accurately and efficiently automatically diagnosing steel pipe concrete structure void defects.
In order to realize the purpose, the invention adopts the technical scheme that: a steel pipe concrete structure void defect diagnosis method based on sound signals comprises the following steps:
1) Acquiring sound signals of the steel pipe concrete structure in different void states;
2) Reading information in the sound signals, performing continuous wavelet transformation by adopting a CMOR wavelet basis function, and converting the one-dimensional sound signals into a two-dimensional wavelet time-frequency graph as an image characteristic for void defect identification;
3) Preprocessing the acquired image characteristics of the steel pipe concrete structure void defect identification;
4) Dividing the image characteristics of the void defect recognition into a training set, a verification set and a test set according to a set proportion, and training a convolutional neural network model for the void defect recognition of the steel pipe concrete structure;
5) And inputting the image characteristics of the concrete-filled steel tube member to be detected into the trained convolutional neural network model to obtain a detection result.
Further, in the step 1), an acoustic sensor is arranged at a set distance above the surface of the steel pipe concrete structure, the surface of the steel pipe concrete is knocked, and the sound signal is collected by a collection device through the acoustic sensor; the sound signals comprise sound signals collected through the reflection process of the steel pipe, the air, the concrete and the steel pipe and sound signals collected after the sound signals are directly diffracted along the steel pipe.
Further, in the step 2), the sound signal x (t) ∈ L 2 (R) the calculation formula for performing the continuous wavelet transform is as follows:
wherein x (t) is an acquired sound signal; a is a scale factor representing frequency dependent scaling;
representing a family function generated by shifting and stretching a base wavelet psi (t), which is called a wavelet base function; the method comprises the following steps: firstly, determining wavelet basis and scale, solving a wavelet coefficient coefs by virtue of cwt (), converting a scale sequence into an actual frequency sequence f by virtue of scal2frq (), and finally drawing a wavelet time-frequency graph by combining a time sequence t and imagesc (t, f, abs (coefs));
the CMOR wavelet is used as a basic wavelet and is defined as:
where t represents time, i is an imaginary number, fb is a bandwidth factor, and Fc is a center frequency factor.
Further, in the step 3), the preprocessing includes: and removing peripheral coordinates, characters and energy bars of the time-frequency graph from the image characteristics, and performing grid normalized compression processing to obtain a pixel format.
Further, in the step 4), the two-dimensional wavelet time-frequency graph and the void defect label corresponding to the two-dimensional wavelet time-frequency graph are used as a sample pair, and all samples are divided into a training set, a verification set and a test set according to a set proportion.
Further, in the step 4), a convolutional neural network model for identifying the void defects of the steel pipe concrete structure is trained, the activating function adopts a ReLU function, the loss function adopts a cross entropy function, the output layer adopts a softmax function, and model classification is completed, wherein the convolutional neural network model sequentially comprises a first convolutional layer, a first pooling layer, a second convolutional layer, a second pooling layer, a third convolutional layer, a third pooling layer, a fourth convolutional layer, a fourth pooling layer, a first full-connection layer, a second full-connection layer and a third full-connection layer.
Further, the first convolution layer adopts 64 convolution kernels of 3 × 3, the step length is 1, images of 128 × 128 × 3 pixels are convoluted to obtain a feature map of 128 × 128 × 64 pixels, data after convolution operation are subjected to ReLU transformation and then are pooled, and the size of the feature map is reduced to 64 × 64 × 64 pixels;
the second convolutional layer is used for convolving the pooled 64X 64pixel feature maps by adopting 128 3X 3 convolution kernels to obtain 64X 128pixel feature maps, reLU transformation is carried out on the data after convolution operation, and then after pooling is carried out, the size of the feature maps is reduced by half to be 32X 128pixel images;
the third convolutional layer performs convolution on the pooled 32 × 32 × 128pixel feature maps by adopting 256 3 × 3 convolution kernels to obtain 32 × 32 × 256pixel feature maps, performs ReLU transformation on the data after the convolution operation, and performs pooling on the data, wherein the size of the feature maps is reduced by half to be a 16 × 16 × 256pixel image;
the fourth convolutional layer is used for convolving the pooled 16 × 16 × 256pixel feature maps by adopting 512 3 × 3 convolution kernels to obtain 16 × 16 × 512pixel feature maps, performing ReLU transformation on the data after convolution operation, and then pooling the data, wherein the size of the feature maps is reduced to 8 × 8 × 512pixel images;
the number of the neurons of the first full connection layer is 1024, the number of the neurons of the second full connection layer is 512, and the number of the neurons of the third full connection layer is 11.
The invention also provides a steel pipe concrete structure void defect diagnosis device based on sound signals for realizing the method, which comprises the following steps:
the sound perception module is used for acquiring sound signals of the steel pipe concrete structure in different void states;
the characteristic acquisition module is used for reading information in the sound signals, performing continuous wavelet transformation by adopting a CMOR wavelet basis function, and converting the one-dimensional sound signals into a two-dimensional wavelet time-frequency diagram as image characteristics for void defect identification; the method is also used for preprocessing the acquired image characteristics for identifying the steel pipe concrete structure void defect;
the training module is used for dividing the image characteristics of void defect recognition into a training set, a verification set and a test set according to a set proportion and training a convolutional neural network model for identifying the void defects of the steel pipe concrete structure;
the diagnosis module is used for inputting the image characteristics of the concrete-filled steel tube member to be detected and detecting the image characteristics through the trained convolutional neural network model to obtain a detection result; and
and the display module is used for displaying the detection result, namely the probability of the steel pipe concrete structure void defect.
Compared with the prior art, the invention has the following beneficial effects: aiming at the sound signal characteristics of the concrete filled steel tube void defect, the perception-based void sound signal utilizes a continuous wavelet transform method to convert a one-dimensional time domain signal into a time-frequency graph form consisting of a time domain and a frequency domain, extracts image characteristics and reduces the complexity of noise and signal components of the sound signal; the method comprises the steps of inputting a time-frequency diagram into a steel pipe concrete void defect diagnosis model obtained through convolutional neural network training in advance, finishing void defect working condition classification by using a softmax classifier to obtain a steel pipe concrete void defect recognition result, and solving the problem that misjudgment or misjudgment is caused due to the fact that human factors influence when human ears are used for screening and judging the steel pipe concrete knocking vibration sound in the prior art. According to the invention, by utilizing acoustic collection and acoustic diagnosis to collect and analyze the vibration sound of the concrete filled steel tube knocking, the accurate and efficient automatic identification and judgment of the steel tube concrete structure void defect are realized.
Drawings
FIG. 1 is a flow chart of a method implementation of an embodiment of the present invention;
FIG. 2 is a diagram of a steel pipe concrete component according to an embodiment of the present invention;
FIG. 3 is a graph of non-nulled and nulled acoustic signals in an embodiment of the present invention;
FIG. 4 is a schematic diagram of a pre-treatment process in an embodiment of the present invention;
FIG. 5 is a schematic diagram of a convolutional neural network model in an embodiment of the present invention;
FIG. 6 is a wavelet time-frequency diagram of non-void and void embodiments of the invention.
Detailed Description
The invention is further explained below with reference to the drawings and the embodiments.
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the disclosure. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments according to the present application. As used herein, the singular forms "a", "an", and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
As shown in fig. 1, the present embodiment provides a steel pipe concrete structure void defect diagnosis method based on an acoustic signal, including the following steps:
1) And acquiring sound signals of the steel pipe concrete structure in different void states.
In the step 1), an acoustic sensor is arranged at a set distance (20 mm in the embodiment) above the surface of the steel pipe concrete structure, the surface of the steel pipe concrete is knocked, and sound signals are collected by collection equipment through the acoustic sensor. For the defect of void inside the steel pipe, the sound signal generated by knocking outside the steel pipe can be collected through the reflection process of the steel pipe, air, concrete and the steel pipe, and can also be directly diffracted along the steel pipe and then collected by a sensor, so that the sound signal is very complex.
One of the concrete filled steel tube members in this embodiment is shown in FIG. 2. The concrete-filled steel tube members are subjected to detection tests on 4 pieces in total, wherein the number of the square concrete-filled steel tube members and the number of the round concrete-filled steel tube test pieces are respectively 2, the number of the non-hollow members is respectively 1, the number of the crown-shaped hollow members is respectively 1, the debonding positions are located locally, the total length L =1200mm of the members, the diameter D =150mm of the round concrete-filled steel tube members, the side length B =150mm of the square concrete-filled steel tube members, the local hollowing range is 80mm, the hollowing value ds =8mm of the round concrete-filled steel tube members, and the hollowing value dt =8mm of the square concrete-filled steel tube members.
In the embodiment, 6 different working condition data are collected. The non-empty and empty sound signal diagram in this embodiment is shown in fig. 3.
2) Reading information in the sound signals, performing continuous wavelet transformation by adopting a CMOR wavelet basis function, and converting the one-dimensional sound signals into a two-dimensional wavelet time-frequency graph as an image characteristic for void defect identification.
Due to the complexity of sound signals, accurately extracting features of such sound signals is critical to achieving accurate diagnosis. On the basis of a large number of experiments, various feature extraction methods such as a Mel frequency feature matrix, a zero crossing rate and the like are tried and compared, and finally a wavelet theory is adopted.
The method for converting the sound signal into the wavelet time-frequency graph can accurately reflect the characteristics of the concrete filled steel tube void sound signal, the wavelet threshold has a certain noise reduction effect, and the data set constructed according to the wavelet time-frequency graph can be effectively used for deep neural network judgment.
In this embodiment, the implementation process of the continuous wavelet transform is as follows:
s21, firstly, selecting a proper wavelet basis function, and fixing a scale factor.
In the embodiment, the basic wavelet adopts CMOR wavelet, which is an abbreviation of complex wavelet Morlet; is defined as:
where t represents time, i is an imaginary number, fb is a bandwidth factor, and Fc is a center frequency factor. The CMOR wavelet can adjust the accuracy of the time-frequency analysis by changing Fb and Fc. The CMOR wavelet has certain properties of a Gaussian function in time domain and frequency domain forms, has good resolution, has good time-frequency aggregation reflected on a time-frequency diagram, and can perform self-adaptive decomposition on a target signal.
The fixed one-scale sequence length of this embodiment takes 256.
And S22, calculating a wavelet coefficient, wherein the wavelet coefficient reflects the similarity between the wavelet under the current scale and the corresponding signal segment.
Wherein, according to the determined wavelet basis function and the scale sequence, the wavelet coefficient coefs is solved by utilizing cwt (). Then, scale sequences are converted into actual frequency sequences f by scal2frq (), and finally, a wavelet time-frequency graph is drawn by combining the time sequences t and imagesc (t, f and abs (coefs)).
And S23, changing the translation factor to enable the wavelet to generate a certain displacement along the time axis, and repeating the steps to finish one-time transformation.
And S24, increasing the scale factor, and repeating the three steps to finish the second analysis.
And S25, circularly executing the steps until the analysis requirement is met.
Specifically, after the wavelet basis function ψ (t) is shifted by b, the wavelet basis function and the original sound signal x (t) are subjected to inner product at different scale a to complete the wavelet transform process, which can be expressed by the following formula:
in the formula, a represents a scale factor and a is greater than 0, and the function is to perform stretching change on the wavelet basis function; b represents a shifting factor reflecting the displacement change of the wavelet basis function psi (t), and the value can be positive or negative; in the embodiment of the present invention, a and b are both continuous variables, so that the method is called Continuous Wavelet Transform (CWT).
3) And preprocessing the acquired image characteristics for identifying the void defects of the steel pipe concrete structure.
In the step 3), the pretreatment comprises: and removing peripheral coordinates, characters and energy bars of the time-frequency image from the image characteristics, and performing grid normalized compression processing to obtain a pixel format. The pretreatment process is shown in FIG. 4.
4) Dividing the image characteristics of the void defect recognition into a training set, a verification set and a test set according to a certain proportion, and training a convolutional neural network model for the void defect recognition of the steel pipe concrete structure.
Specifically, a two-dimensional wavelet time-frequency graph and a void defect label corresponding to the two-dimensional wavelet time-frequency graph are used as a sample pair, and all samples are divided into a training set, a verification set and a test set according to a set proportion.
In the step 4), a convolutional neural network model for identifying the steel pipe concrete structure void defects is trained, the ReLU function is adopted as an activation function, the cross entropy function is adopted as a loss function, and the softmax function is adopted as an output layer, so that model classification is completed. As shown in fig. 5, the convolutional neural network model sequentially includes a first convolutional layer, a first pooling layer, a second convolutional layer, a second pooling layer, a third convolutional layer, a third pooling layer, a fourth convolutional layer, a fourth pooling layer, a first fully-connected layer, a second fully-connected layer, and a third fully-connected layer.
The first convolution layer adopts 64 convolution kernels of 3 × 3, the step length is 1, images of 128 × 128 × 3 pixels are convoluted to obtain feature maps of 128 × 128 × 64 pixels, data after convolution operation are subjected to ReLU transformation and then pooled, and the size of the feature maps is reduced by half to be images of 64 × 64 × 64 pixels.
The second convolutional layer uses 128 convolution kernels of 3 × 3 to convolve the pooled feature maps of 64 × 64 × 64 pixels to obtain feature maps of 64 × 64 × 128 pixels, performs ReLU transformation on the data after the convolution operation, and then performs pooling, so that the feature map size is reduced to 32 × 32 × 128 pixels.
The third convolutional layer performs convolution on the pooled 32 × 32 × 128pixel feature maps by using 256 3 × 3 convolution kernels to obtain 32 × 32 × 256pixel feature maps, performs ReLU transformation on the data after the convolution operation, and performs pooling to obtain an image with a feature map size reduced by half to 16 × 16 × 256 pixels.
The fourth convolution layer convolves the pooled 16 × 16 × 256pixel feature maps by using 512 3 × 3 convolution kernels to obtain 16 × 16 × 512pixel feature maps, performs ReLU transformation on the convolved data, and then performs pooling, thereby reducing the feature map size to an image of 8 × 8 × 512 pixels.
The number of the neurons of the first full connection layer is 1024, the number of the neurons of the second full connection layer is 512, and the number of the neurons of the third full connection layer is 11.
In this example, each condition included 800 samples for 4800 samples. After the data of the time-frequency diagrams under 6 working conditions are labeled, the data are divided into a training set, a verification set and a test set according to the proportion of 7. FIG. 6 is a wavelet time-frequency diagram of non-void and void.
Inputting training and verification samples into a network for forward propagation, obtaining a predicted value after the training and verification samples pass through a classification layer, reversely propagating a loss value through an optimization algorithm, and updating network parameters. Wherein the hyper-parameters of the model are set as: the learning rate was set to 0.002; the number of training rounds is 60; the optimization algorithm selects an Adam optimization algorithm.
In this embodiment, the model is composed of four convolution layers, four pooling layers and three full-connection layers. After the training and verification set is used for adjusting the network model, the model parameters are iterated to be optimal, and the feasibility of the whole model is verified by utilizing the test set void defect sample.
And inputting the test set into the optimal model to obtain a diagnosis result.
5) And inputting the image characteristics of the concrete-filled steel tube member to be detected into the trained convolutional neural network model to obtain a detection result.
The embodiment also provides a steel pipe concrete structure void defect diagnosis device based on sound signals for implementing the method, which comprises the following steps: the device comprises a sound perception module, a characteristic acquisition module, a training module, a diagnosis module and a display module.
The sound perception module is used for acquiring sound signals of the steel pipe concrete structures in different void states;
the characteristic acquisition module is used for reading information in the sound signals, performing continuous wavelet transformation by adopting a CMOR wavelet basis function, and converting the one-dimensional sound signals into a two-dimensional wavelet time-frequency diagram as image characteristics for void defect identification; the method is also used for preprocessing the acquired image characteristics for identifying the steel pipe concrete structure void defect;
the training module is used for dividing the image characteristics of the void defect recognition into a training set, a verification set and a test set according to a set proportion and training a convolutional neural network model for the void defect recognition of the steel pipe concrete structure;
the diagnosis module is used for inputting the image characteristics of the concrete-filled steel tube component to be detected and detecting the image characteristics through the trained convolutional neural network model to obtain a detection result; and
and the display module is used for displaying the detection result, namely the probability of the steel pipe concrete structure void defect.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and so forth) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The foregoing is directed to preferred embodiments of the present invention, other and further embodiments of the invention may be devised without departing from the basic scope thereof, and the scope thereof is determined by the claims that follow. However, any simple modification, equivalent change and modification of the above embodiments according to the technical essence of the present invention will still fall within the protection scope of the technical solution of the present invention.
Claims (8)
1. A steel pipe concrete structure void defect diagnosis method based on sound signals is characterized by comprising the following steps:
1) Acquiring sound signals of the steel pipe concrete structure in different void states;
2) Reading information in the sound signals, performing continuous wavelet transform by adopting a CMOR wavelet basis function, and converting the one-dimensional sound signals into a two-dimensional wavelet time-frequency graph as image characteristics for void defect identification;
3) Preprocessing the acquired image characteristics of the steel pipe concrete structure void defect identification;
4) Dividing the image characteristics of void defect recognition into a training set, a verification set and a test set according to a set proportion, and training a convolutional neural network model for identifying void defects of the steel pipe concrete structure;
5) And inputting the image characteristics of the concrete-filled steel tube member to be detected into the trained convolutional neural network model to obtain a detection result.
2. The steel pipe concrete structure void defect diagnosis method based on sound signals as set forth in claim 1, wherein in said step 1), an acoustic sensor is disposed at a set distance above the surface of the steel pipe concrete structure, the surface of the steel pipe concrete is tapped, and a collection device collects said sound signals through the acoustic sensor; the sound signals comprise sound signals collected through the reflection process of the steel pipe, the air, the concrete and the steel pipe and sound signals collected after the sound signals are directly diffracted along the steel pipe.
3. The method for diagnosing void defect of steel pipe concrete structure according to claim 1, wherein in the step 2), the sound signal x (t) e L 2 (R) the calculation formula for performing the continuous wavelet transform is as follows:
wherein x (t) is an acquired sound signal; a is a scale factor representing frequency dependent scaling;
representing a family function generated by shifting and stretching a base wavelet psi (t), which is called a wavelet base function;
the CMOR wavelet is used as a basic wavelet and is defined as:
where t represents time, i is an imaginary number, fb is a bandwidth factor, and Fc is a center frequency factor.
4. The method for diagnosing void defect of a steel pipe concrete structure according to claim 1, wherein the preprocessing in the step 3) comprises: and removing peripheral coordinates, characters and energy bars of the time-frequency graph from the image characteristics, and performing grid normalized compression processing to obtain a pixel format.
5. The steel pipe concrete structure void defect diagnosis method based on sound signals as claimed in claim 1, wherein in the step 4), the two-dimensional wavelet time-frequency diagram and the void defect label corresponding to the two-dimensional wavelet time-frequency diagram are used as a sample pair, and all samples are divided into a training set, a verification set and a test set according to a set proportion.
6. The steel pipe concrete structure void defect diagnosis method based on the acoustic signal as claimed in claim 1, wherein in the step 4), a convolutional neural network model for steel pipe concrete structure void defect recognition is trained, the activation function adopts a ReLU function, the loss function adopts a cross entropy function, the output layer adopts a softmax function, and model classification is completed, wherein the convolutional neural network model sequentially comprises a first convolutional layer, a first pooling layer, a second convolutional layer, a second pooling layer, a third convolutional layer, a third pooling layer, a fourth convolutional layer, a fourth pooling layer, a first fully-connected layer, a second fully-connected layer and a third fully-connected layer.
7. The steel pipe concrete structure void defect diagnosis method based on the acoustic signal as set forth in claim 6, wherein the first convolution layer adopts 64 convolution kernels of 3 x 3, the step length is 1, the images of 128 x 3 pixels are convoluted to obtain a feature map of 128 x 64 pixels, the data after convolution operation is subjected to ReLU transformation and then pooled, and the size of the feature map is reduced by half to be the images of 64 x 64 pixels;
the second convolutional layer is used for convolving the pooled 64X 64pixel feature maps by adopting 128 3X 3 convolution kernels to obtain 64X 128pixel feature maps, reLU transformation is carried out on the data after convolution operation, and then after pooling is carried out, the size of the feature maps is reduced by half to be 32X 128pixel images;
the third convolutional layer performs convolution on the pooled 32 × 32 × 128pixel feature maps by adopting 256 3 × 3 convolution kernels to obtain 32 × 32 × 256pixel feature maps, performs ReLU transformation on the data after the convolution operation, and performs pooling on the data, wherein the size of the feature maps is reduced by half to be a 16 × 16 × 256pixel image;
the fourth convolution layer is used for convolving the feature map of 16 multiplied by 256pixel after pooling by adopting 512 convolution kernels of 3 multiplied by 3 to obtain the feature map of 16 multiplied by 512pixel, reLU transformation is carried out on the data after convolution operation, and then after pooling is carried out, the size of the feature map is reduced to 8 multiplied by 512 pixel;
the number of the neurons of the first full connection layer is 1024, the number of the neurons of the second full connection layer is 512, and the number of the neurons of the third full connection layer is 11.
8. An acoustic signal-based steel pipe concrete structure void defect diagnosis apparatus for implementing the method according to any one of claims 1 to 7, comprising:
the sound perception module is used for acquiring sound signals of the steel pipe concrete structures in different void states;
the characteristic acquisition module is used for reading information in the sound signals, performing continuous wavelet transform by adopting a CMOR wavelet basis function, and converting the one-dimensional sound signals into a two-dimensional wavelet time-frequency graph as image characteristics for void defect identification; the method is also used for preprocessing the acquired image characteristics for identifying the steel pipe concrete structure void defect;
the training module is used for dividing the image characteristics of the void defect recognition into a training set, a verification set and a test set according to a set proportion and training a convolutional neural network model for the void defect recognition of the steel pipe concrete structure;
the diagnosis module is used for inputting the image characteristics of the concrete-filled steel tube component to be detected and detecting the image characteristics through the trained convolutional neural network model to obtain a detection result; and
and the display module is used for displaying the detection result, namely the probability of the steel pipe concrete structure void defect.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211017584.4A CN115356397A (en) | 2022-08-23 | 2022-08-23 | Steel pipe concrete structure void defect diagnosis method and device based on sound signals |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211017584.4A CN115356397A (en) | 2022-08-23 | 2022-08-23 | Steel pipe concrete structure void defect diagnosis method and device based on sound signals |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115356397A true CN115356397A (en) | 2022-11-18 |
Family
ID=84002498
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211017584.4A Pending CN115356397A (en) | 2022-08-23 | 2022-08-23 | Steel pipe concrete structure void defect diagnosis method and device based on sound signals |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115356397A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116609441A (en) * | 2023-07-14 | 2023-08-18 | 河北工业大学 | Quantitative identification method for single microcracks |
| CN117011295A (en) * | 2023-10-07 | 2023-11-07 | 西南科技大学 | UHPC prefabricated member quality detection method based on depth separable convolutional neural network |
| CN117129565A (en) * | 2023-08-23 | 2023-11-28 | 广西大学 | Concrete filled steel tube void knocking force detection method based on energy ratio and GWO-SVM |
| CN117349601A (en) * | 2023-12-06 | 2024-01-05 | 济南大学 | Concrete damage classification method and system based on piezoelectric vibration waves |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010169558A (en) * | 2009-01-23 | 2010-08-05 | Hitachi Constr Mach Co Ltd | Ultrasonic measuring device |
| CN110045015A (en) * | 2019-04-18 | 2019-07-23 | 河海大学 | A kind of concrete structure Inner Defect Testing method based on deep learning |
| CN110243934A (en) * | 2019-05-30 | 2019-09-17 | 中国计量大学上虞高等研究院有限公司 | A kind of ultrasonic weld seam detection method based on wavelet convolution neural network |
| CN110689011A (en) * | 2019-09-29 | 2020-01-14 | 河北工业大学 | Solar cell panel defect detection method of multi-scale combined convolution neural network |
| CN112229585A (en) * | 2020-09-18 | 2021-01-15 | 长沙理工大学 | Crack damage positioning method and system based on artificial intelligence and acoustic emission technology |
| CN113670610A (en) * | 2021-06-09 | 2021-11-19 | 广州大学 | Fault detection method, system and medium based on wavelet transformation and neural network |
| CN114048682A (en) * | 2021-11-22 | 2022-02-15 | 昆明理工大学 | Rolling bearing acoustic emission intelligent diagnosis method based on optimized wavelet basis and multi-dimensional depth feature fusion |
-
2022
- 2022-08-23 CN CN202211017584.4A patent/CN115356397A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010169558A (en) * | 2009-01-23 | 2010-08-05 | Hitachi Constr Mach Co Ltd | Ultrasonic measuring device |
| CN110045015A (en) * | 2019-04-18 | 2019-07-23 | 河海大学 | A kind of concrete structure Inner Defect Testing method based on deep learning |
| CN110243934A (en) * | 2019-05-30 | 2019-09-17 | 中国计量大学上虞高等研究院有限公司 | A kind of ultrasonic weld seam detection method based on wavelet convolution neural network |
| CN110689011A (en) * | 2019-09-29 | 2020-01-14 | 河北工业大学 | Solar cell panel defect detection method of multi-scale combined convolution neural network |
| CN112229585A (en) * | 2020-09-18 | 2021-01-15 | 长沙理工大学 | Crack damage positioning method and system based on artificial intelligence and acoustic emission technology |
| CN113670610A (en) * | 2021-06-09 | 2021-11-19 | 广州大学 | Fault detection method, system and medium based on wavelet transformation and neural network |
| CN114048682A (en) * | 2021-11-22 | 2022-02-15 | 昆明理工大学 | Rolling bearing acoustic emission intelligent diagnosis method based on optimized wavelet basis and multi-dimensional depth feature fusion |
Non-Patent Citations (2)
| Title |
|---|
| 卢超, 于润桥, 彭应秋, 王方明: "连续小波变换在焊接缺陷超声检测中的应用", 南昌大学学报(工科版), no. 02, 30 August 2002 (2002-08-30) * |
| 王森;伍星;张印辉;陈庆;: "基于深度学习的全卷积网络图像裂纹检测", 计算机辅助设计与图形学学报, no. 05, 15 May 2018 (2018-05-15) * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116609441A (en) * | 2023-07-14 | 2023-08-18 | 河北工业大学 | Quantitative identification method for single microcracks |
| CN116609441B (en) * | 2023-07-14 | 2023-09-26 | 河北工业大学 | Quantitative identification method for single microcracks |
| CN117129565A (en) * | 2023-08-23 | 2023-11-28 | 广西大学 | Concrete filled steel tube void knocking force detection method based on energy ratio and GWO-SVM |
| CN117129565B (en) * | 2023-08-23 | 2024-06-11 | 广西大学 | Concrete filled steel tube void knocking force detection method based on energy ratio and GWO-SVM |
| CN117011295A (en) * | 2023-10-07 | 2023-11-07 | 西南科技大学 | UHPC prefabricated member quality detection method based on depth separable convolutional neural network |
| CN117011295B (en) * | 2023-10-07 | 2024-01-23 | 西南科技大学 | UHPC prefabricated member quality detection method based on depth separable convolutional neural network |
| CN117349601A (en) * | 2023-12-06 | 2024-01-05 | 济南大学 | Concrete damage classification method and system based on piezoelectric vibration waves |
| CN117349601B (en) * | 2023-12-06 | 2024-03-05 | 济南大学 | Concrete damage classification method and system based on piezoelectric vibration waves |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115356397A (en) | Steel pipe concrete structure void defect diagnosis method and device based on sound signals | |
| CN112508901B (en) | Underwater structure disease identification method, system, device and storage medium | |
| CN110440148B (en) | Method, device and system for classifying and identifying leakage acoustic signals | |
| CN114445386B (en) | PVC pipe quality detection and evaluation method and system based on artificial intelligence | |
| CN116546023B (en) | Method and system for identifying violent behaviors of oil and gas operation area | |
| CN113569672A (en) | Lightweight target detection and fault identification method, device and system | |
| CN113822201B (en) | Deep learning method for underwater object shape recognition based on flow field velocity component time course | |
| CN116596326B (en) | Urban environment detection and comprehensive evaluation method based on remote sensing data | |
| Huang et al. | A deep learning framework based on improved self‐supervised learning for ground‐penetrating radar tunnel lining inspection | |
| CN118209305B (en) | Method, medium and system for detecting fatigue degree of iron core and clamping piece bolt fastener | |
| CN112349057A (en) | Deep learning-based indoor smoke and fire detection method | |
| CN113719764A (en) | Pipeline leakage detection method | |
| CN116051496A (en) | Real-time sewer defect detection method | |
| CN117849193A (en) | Online crack damage monitoring method for neodymium iron boron sintering | |
| CN114219998A (en) | Sonar image real-time detection method based on target detection neural network | |
| CN116778346B (en) | Pipeline identification method and system based on improved self-attention mechanism | |
| CN109945075B (en) | Method and device for detecting leakage degree of water supply pipeline | |
| CN117274355A (en) | Drainage pipeline flow intelligent measurement method based on acceleration guidance area convolutional neural network and parallel multi-scale unified network | |
| CN115356400B (en) | BiLSTM-based ultrasonic wood nondestructive testing method and BiLSTM-based ultrasonic wood nondestructive testing system | |
| CN116884435A (en) | Voice event detection method and device based on audio prompt learning | |
| CN108764311A (en) | A kind of shelter target detection method, electronic equipment, storage medium and system | |
| CN114199992A (en) | Oil storage tank wall corrosion detection method and system | |
| CN110322894A (en) | A kind of waveform diagram generation and giant panda detection method based on sound | |
| CN113962934A (en) | Defect detection method and system based on fast RCNN (radar cross-section network) | |
| Li et al. | Research on reading recognition of pointer meter based on improved U-net network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |