KR102402966B1 - System and method for the failure mechanism detection of hybrid composite structures based on acoustic emission sensors and artificial intelligence algorithms - Google Patents

Abstract

Disclosed are a system for detecting a breakage mechanism of a hybrid composite structure based on an acoustic emission sensor and an artificial intelligence algorithm, and a method therefor. A breakage mechanism detection method of a hybrid composite material structure performed by a breakage mechanism detection system according to an embodiment includes: detecting signal information emitted from the hybrid composite material structure; performing clustering according to the identified breakage mechanism using the detected signal information; inputting new signal information into the learned learning model to classify the breakage mechanism using the signal information of the cluster generated by performing the clustering; and detecting a breakage mechanism of the hybrid composite material structure from the new signal information using the learned deep learning-based learning model.

Description

SYSTEM AND METHOD FOR THE FAILURE MECHANISM DETECTION OF HYBRID COMPOSITE STRUCTURES BASED ON ACOUSTIC EMISSION SENSORS AND ARTIFICIAL INTELLIGENCE ALGORITHMS

The description below relates to a technique for detecting failure mechanisms in hybrid composite structures.

Recently, the application of various lightweight materials such as ultra-high-strength steel, aluminum, and Fiber Reinforced Plastics (FRP) is expanding to reduce the weight of transportation means such as automobiles, aircraft, and ships. In particular, the hybrid composite material has excellent mechanical properties, such as specific strength, specific stiffness, fatigue resistance, and heat resistance, and thus is widely applied in various industrial fields. In addition, as the mechanical bonding method technology using rivets for bonding existing materials between different materials such as FRP and aluminum is advanced, lightweight and high-performance products, devices, or buildings that mix FRP with metal materials are further increasing.

As the use of hybrid composite materials increases, it is essential to secure the safety and durability of structures to which the hybrid composite material is applied. . However, the existing method, which relies only on acoustic emission factor analysis by attaching only the acoustic emission (AE) sensor to the structure, takes a long time to analyze, and it is difficult to analyze material damage and various failure mechanisms without specialized knowledge. In addition, there is a difficult limitation in identifying the breakage mechanism by the noise signal according to the environment to which the structure is exposed. At present, there is almost no system or method for identifying a breakage mechanism by applying an artificial intelligence algorithm to the acoustic emission signal data specialized for each hybrid composite structure.

The damage occurring in the hybrid composite material structure is measured using the acoustic emission sensor, and each signal characteristic is checked by applying the acoustic emission factor analysis method, and then the artificial intelligence algorithm is based on the data clustered according to the failure mechanism of each material of the structure. By applying classification learning, it is possible to provide a system and method for automatically classifying and detecting the damage mechanism of a structure by inputting only the sound emission signal data value without the user having expertise in the acoustic emission signal or material behavior and damage.

A fracture mechanism detection method of a hybrid composite material structure performed by the failure mechanism detection system, the method comprising the steps of: detecting signal information emitted from the composite material structure; performing clustering according to the identified breakage mechanism using the detected signal information; inputting new signal information to the learned deep learning-based learning model to classify the breakage mechanism using the signal information of the cluster generated by performing the clustering; and detecting a failure mechanism of the hybrid composite material structure from the new signal information using the learned learning model.

The detecting of the signal information includes attaching an acoustic emission sensor to a position where deformation or damage is expected in the hybrid composite material structure, and using the attached acoustic emission sensor to the hybrid composite material structure or the hybrid composite A method comprising: acquiring an acoustic wave emitted from a material constituting a material structure; and performing the clustering includes a Short time furrier transform (STFT) on the acquired acoustic wave. ) or processing a Fast furrier transform (FFT) to extract a plurality of sound emission factor characteristics for sound emission signal information including amplitude, holding time, energy, maximum frequency, and normalized cumulative energy according to time may include

The step of performing the clustering may include identifying a failure mechanism of a hybrid composite material or a hybrid composite material structure based on the extracted characteristics of a plurality of acoustic emission factors, and performing the clustering to identify the identified clusters. It may include the step of linking the breakage mechanism to database the signal information clustered according to the breakage mechanism.

The step of clustering according to the breakage mechanism using a plurality of factors according to time extracted from the detected seismic wave includes converting the sound emission signal information into k-Means, Genetic k-Means, Fuzzy C-Means, The method may include clustering using any one of a Self-Organizing Map (SOM), a Gaussian Mixture Model (GMM), and a hierarchical model.

The step of performing the clustering includes distance information, Euclidean distance or statistical and stochastic distance from the center of the cluster generated by performing clustering according to the identified breakage mechanism, in the x-y coordinate system, distance information, time It may include the step of databaseizing the signal information according to the.

The hybrid composite material structure is a single FRP material, or a structure joined by a rivet or bonding method between an FRP material and a metal material, and a rivet or bonding method between different FRP materials. It may include a structure.

The step of inputting the new signal information includes performing classification learning on the acoustic emission signal information according to the time at which the clustering is performed using a deep learning-based learning model built to classify the breakage mechanism, and , The step of detecting the breakage mechanism may include inputting a plurality of sound emission factors extracted from the new signal information into the learning model on which the classification learning is performed to detect the breakage mechanism of the hybrid composite material structure. have.

Breakage mechanism detection system, a signal detection unit for detecting the signal information emitted from the material constituting the hybrid composite structure; a clustering unit configured to perform clustering according to the identified breakage mechanism using the detected signal information; an input unit for inputting new signal information into the deep learning-based learning model learned in order to classify the breakage mechanism using the signal information of the cluster generated by performing the clustering; And it may include a failure mechanism detection unit for detecting the failure mechanism of the hybrid composite structure from the new signal information using the learned learning model.

The breakage mechanism detection system according to an embodiment applies the clustering signal data and artificial intelligence learning algorithm for each breakage mechanism occurring in the hybrid composite material structure in an environment where an arbitrary load is applied to simply and easily detect the failure mechanism of the hybrid composite material structure. can be checked

The breakage mechanism detection system according to an embodiment can be easily used by a worker who lacks professional knowledge for sound emission signal analysis.

The breakage mechanism detection system according to an embodiment can be applied to various products or devices to which the hybrid composite material structure is applied, so that stability and reliability can be immediately secured, and the breakage mechanism can be identified simply and accurately.

The breakage mechanism detection system according to an embodiment can respond to damage to the composite material structure in a relatively short time, thereby reducing cost and time required for breakage analysis.

1 is a block diagram for explaining the configuration of a breakage mechanism detection system according to an embodiment.
2 is a flowchart for explaining a method of detecting a breakage mechanism of a hybrid composite material structure based on an acoustic emission sensor and an artificial intelligence algorithm in a breakage mechanism detection system according to an embodiment.
3 is an example for explaining the SPR-bonded CFRP-A16061 specimen used in the overlap shear test in the breakage mechanism detection system according to an embodiment.
4 is a view for explaining an overlap shear test t-AE sensor linkage test in a breakage mechanism detection system according to an embodiment.
5 is an example for explaining the optical microscopic analysis results for each breakage mechanism of the CFRP damaged in the overlap shear test in the breakage mechanism detection system according to an embodiment.
6 is an example for explaining the maximum frequency clustering result of the sound emission signal over time of the CFRP in the breakage mechanism detection system according to an embodiment.
7 is an example for explaining the maximum frequency clustering result of the sound emission signal over time of A16061 in the breakage mechanism detection system according to an embodiment.
8 is an example for explaining a network constituting a learning model in the breakage mechanism detection system according to an embodiment.
9 is an example for explaining the results of classifying the breakage mechanism of the CFRP-A16061 specimen in the breakage mechanism detection system according to an embodiment.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings.

In the embodiment, after applying various loads to a hybrid composite material structure rather than a single composite material, a clustering algorithm such as k -means is applied to the acoustic emission signal measured at this time to link the failure mechanism according to each cluster. And after making a database of signal information clustered according to each breakage mechanism, an operation of analyzing the breakage mechanism for the acoustic emission signal measured in real time by applying and learning an artificial intelligence-based classification learning algorithm will be described. Accordingly, it is applied to all structures to which hybrid composite materials that require confirmation of stability and reliability based on artificial intelligence algorithms, such as automobiles, aircraft, ships, and building structures, are applied, and provides users with an immediate understanding of the risk of damage and countermeasures according to the damage mechanism. At the same time, it can be used to create a safe working and operating environment.

1 is a block diagram illustrating the configuration of a breakage mechanism detection system according to an embodiment, and FIG. 2 is a breakage of a hybrid composite material structure based on an acoustic emission sensor and an artificial intelligence algorithm in a breakage mechanism detection system according to an embodiment It is a flowchart for explaining the instrument detection method.

The breakage mechanism detection system 100 may include a signal detection unit 110 , a clustering unit 120 , an input unit 130 , and a breakage mechanism detection unit 140 . The components of the breakage mechanism detection system 100 may be representations of different functions performed by the processor according to instructions provided by the program code stored in the breakage mechanism detection system 100 . The components of the breakage mechanism detection system 100 and the breakage mechanism detection system 100 are steps 210 to 240 including the breakage mechanism detection method of the hybrid composite material structure based on the acoustic emission sensor and artificial intelligence algorithm of FIG. The breakage mechanism detection system can be controlled to perform In this case, the breakage mechanism detection system 100 and the components of the breakage mechanism detection system 100 may be implemented to execute instructions according to the code of the operating system and at least one program code included in the memory.

The breakage mechanism detection system may load the program code stored in the file of the program for the method for detecting the breakage mechanism of the hybrid composite material structure based on the acoustic emission sensor and the artificial intelligence algorithm into the memory. For example, when a program is executed in the breakage detection system, the processor may control the breakage mechanism detection system to load a program code from a file of the program into the memory according to the control of the operating system. At this time, each of the breakage mechanism detection system 100 and the breakage mechanism detection system 100 includes the signal detection unit 110 , the clustering unit 120 , the input unit 130 , and the breakage mechanism detection unit 140 loaded into the memory. may be different functional representations of the breakage mechanism detection system 100 for executing subsequent steps 210 to 240 by executing instructions of a corresponding part of the program code.

In step 210, the signal detector 110 may detect signal information emitted from the hybrid composite material structure. First, a hybrid composite material or a hybrid composite material structure may be prepared. At this time, the hybrid composite material structure is a single FRP material, or a structure joined by a rivet or bonding method between an FRP material and a metal material, and a rivet or bonding method between different FRP materials. structure may be included. The signal detection unit 110 attaches an acoustic emission sensor to a location (eg, a structure/material of interest) where deformation or damage is expected in the hybrid composite material or hybrid composite material structure, and the attached acoustic emission sensor can be used to acquire a hybrid composite material structure or an elastic wave emitted from a material constituting the hybrid composite material structure.

In step 220, the clustering unit 120 may perform clustering according to the identified breakage mechanism using the detected signal information. The clustering unit 120 processes a short time furrier transform (STFT) or a fast furrier transform (FFT) for the obtained acoustic wave to obtain amplitude over time, duration time) , energy, a peak frequency, and a plurality of sound emission factor characteristics for sound emission signal information including normalized cumulative energy may be extracted. At this time, the clustering unit 120 identifies the breakage mechanism of the hybrid composite material or hybrid composite material structure based on the extracted characteristics of the plurality of sound emission factors, and performs clustering to link the identified breakage mechanism to the generated cluster to damage According to the mechanism, the clustered signal information can be databased. The clustering unit 120 stores distance information including distance in the x-y coordinate system, Euclidean distance or statistical and stochastic distance from the center of the cluster generated by performing clustering according to the identified breakage mechanism, and signal information according to time database. can be stored in

In step 230, the input unit 130 may input new signal information to the deep learning-based learning model learned in order to classify the breakage mechanism using the signal information of the cluster generated by performing the clustering. The input unit 130 may perform classification learning on the acoustic emission signal information according to the time when the clustering is performed using a deep learning-based learning model built to classify the breakage mechanism. The input unit 130 may classify and learn the acoustic emission signal information according to time clustered by using a deep learning algorithm such as ANN, DNN, and RNN. The input unit 130 includes hyperparameters such as the ratio of training data and test data, the number of nodes and layers, activation function, and loss function in the signal information database. After tuning, classification learning is performed to build a deep learning-based learning model that can accurately classify the failure mechanism when a new sound emission signal due to the failure of the hybrid composite structure is input. The input unit 130 reads the new sound emission signal on the x-axis (time) and y-axis (a plurality of acoustic emission factor characteristics such as signal holding time, amplitude, maximum frequency, energy, normalized cumulative energy, etc.) data on the x-y coordinate system graph. New signal information can be input to the learned deep learning-based learning model in order to classify the breaking mechanism by the signal information data that appears when it is inserted.

In step 240 , the failure mechanism detection unit 140 may detect the failure mechanism of the hybrid composite structure from new signal information using the learned learning model. The breakage mechanism detector 140 may detect the breakage mechanism of the hybrid composite material structure by inputting a plurality of sound emission factors extracted from the new signal information into the learning model on which classification learning is performed.

Furthermore, the wavefront of the hybrid composite material structure can be observed using a scanning electron microscope or an optical microscope. For example, a fracture surface of the hybrid composite material structure may be observed using an image signal received from a scanning electron microscope or an optical microscope, and a damage state for the observed fracture surface may be output.

3 is an example for explaining the SPR-bonded CFRP-A16061 specimen used in the overlap shear test in the breakage mechanism detection system according to an embodiment.

The failure mechanism detection system can be used in the hybrid composite material or the hybrid composite material structure to identify the failure mechanism of the composite material structure. For example, a hybrid composite structure is a single FRP material, or a structure joined by a rivet or bonding method between an FRP material and a metal material, or a rivet or bonding method between different FRP materials It can be composed of a structure joined by Referring to FIG. 3 , a specimen obtained by bonding woven carbon fiber reinforced plastic (CFRP) and Al6061 with a self pierced rivet (SPR) will be described as an example.

4 is a view for explaining an overlap shear test t-AE sensor linkage test in a breakage mechanism detection system according to an embodiment.

In order to check the breakage mechanism and acoustic emission signal of the CFRP-Al6061 specimen in an arbitrary displacement control method, after installing acoustic emission sensors on CFRP and Al6061, respectively, a lap shear test can be performed using a tensile tester. In addition, it is possible to obtain signal information by connecting an acoustic emission sensor to a computer.

The breakage mechanism detection system may set the threshold voltage value to a preset value (eg, 40 dB) for signal processing obtained from the sound emission sensor, and perform short time fourier transform (STFT) processing. Alternatively, FFT processing may be applied in addition to STFT. In order to analyze the acoustic emission signal of the CFRP-Al6061 specimen under various loads, the tensile speed is increased by a preset unit (for example, from 1.0 mm/min to 0.5 mm/min unit) as shown in Table 1 to a maximum of 10.0 mm/min. After performing the overlap shear test in a plurality (eg, 19) cases, it is possible to analyze the maximum frequency data of the acoustic emission signal over time.

Table 1:

5 is an example for explaining the optical microscopic analysis results for each breakage mechanism of the CFRP damaged in the overlap shear test in the breakage mechanism detection system according to an embodiment.

The breakage mechanism detection system can classify various (for example, three in total) breakage mechanisms such as base metal breakage, fiber/base metal separation, and fiber breakage through optical microscopic analysis based on the performed overlap shear test. In Al6061, the breakage mechanism due to plastic deformation can be confirmed.

Figure 6 is an example for explaining the maximum frequency clustering result of the acoustic emission signal over time of CFRP in the breakage mechanism detection system according to an embodiment, Figure 7 is a breakage mechanism detection system according to an embodiment at the time of A16061 This is an example for explaining the maximum frequency clustering result of the acoustic emission signal according to the

For example, the breakage mechanism detection system reads the signal information about the acoustic emission signal on the x-y coordinate system graph, and then provides various artificial intelligence such as k-Means, Genetic k-Means, Fuzzy C-Means, SOM, GMM, hierarchical model, etc. It can be clustered based on an algorithm. Specifically, the breakage mechanism detection system applies an artificial intelligence algorithm to cluster the acoustic emission signals according to the breakage mechanism of the CFRP-Al6061 specimen. By applying the k-Means method to , the acoustic emission signals generated from the specimen can be clustered.

Among them, the maximum frequency clustering results of 2,062 and 1,470 time-dependent acoustic emission signals obtained from CFRP and Al6061 of the overlap shear test conducted for 45.2 seconds in Case 5 of Table 1 can be shown in the graphs of FIGS. 6 and 7, respectively. and the failure mechanisms belonging to the cluster can be represented as shown in Table 2.

Table 2:

After that, the breakage mechanism detection system applies the deep learning algorithm to perform classification learning, and the Euclidean distance from each cluster center to the signal data according to the breakage mechanism of CFRP and Al6061 and the signal information according to time are shown in Tables 3 and 4 and You can organize them together to build a database.

Table 3:

Table 4:

8 is an example for explaining a network constituting a learning model in the breakage mechanism detection system according to an embodiment.

The breakage mechanism detection system can perform classification learning based on artificial intelligence algorithms. As an example, in an environment such as Python, C, C++, or matlab, using a deep learning algorithm such as ANN, DNN, RNN, etc. can be done

In an embodiment, classification learning may be performed by applying the ANN-based Keras deep learning algorithm in a Python environment. In this case, the network may be configured as shown in FIG. 8 by setting the ratio of the number of training data to the number of test data for hyperparameter tuning. For example, for hyperparameter tuning, the ratio of the number of training data to the number of test data was set to 7:3, and the network was composed of 1000 nodes and 5 layers, and the logistic loss function and By building a deep learning model with the Relu activation function, classification learning can be performed 1000 times. Accordingly, a high level (eg, 94%) of learning accuracy may be derived.

9 is an example for explaining the results of classifying the breakage mechanism of the CFRP-A16061 specimen in the breakage mechanism detection system according to an embodiment.

The breakage mechanism detection system can present a learning model for the method of detecting breakage and breakage mechanism of hybrid composite structures based on acoustic emission sensors and artificial intelligence learning algorithms. When the AE signal newly acquired by the breakage detection system was placed in the coordinate information of 2119 signal information (time, maximum frequency), as shown in Table 5, the Euclidean distance from the cluster center to the signal data and the signal information data according to time were completed. You can check the damage mechanism intuitively by inputting it into the trained learning model.

Table 5:

The device described above may be implemented as a hardware component, a software component, and/or a combination of the hardware component and the software component. For example, devices and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA). , a programmable logic unit (PLU), microprocessor, or any other device capable of executing and responding to instructions, may be implemented using one or more general purpose or special purpose computers. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. A processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For convenience of understanding, although one processing device is sometimes described as being used, one of ordinary skill in the art will recognize that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that can include For example, the processing device may include a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as parallel processors.

Software may comprise a computer program, code, instructions, or a combination of one or more thereof, which configures a processing device to operate as desired or is independently or collectively processed You can command the device. The software and/or data may be any kind of machine, component, physical device, virtual equipment, computer storage medium or apparatus, to be interpreted by or to provide instructions or data to the processing device. may be embodied in The software may be distributed over networked computer systems and stored or executed in a distributed manner. Software and data may be stored in one or more computer-readable recording media.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be specially designed and configured for the embodiment, or may be known and available to those skilled in the art of computer software. Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic such as floppy disks. - includes magneto-optical media, and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like.

As described above, although the embodiments have been described with reference to the limited embodiments and drawings, various modifications and variations are possible from the above description by those skilled in the art. For example, the described techniques are performed in an order different from the described method, and/or the described components of the system, structure, apparatus, circuit, etc. are combined or combined in a different form than the described method, or other components Or substituted or substituted by equivalents may achieve an appropriate result.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (8)

In the failure mechanism detection method of the hybrid composite structure carried out by the failure mechanism detection system,
detecting signal information emitted from the hybrid composite structure;
performing clustering according to the identified breakage mechanism using the detected signal information;
inputting new signal information to the learned deep learning-based learning model to classify the breakage mechanism using the signal information of the cluster generated by performing the clustering; and
Detecting the failure mechanism of the hybrid composite material structure from the new signal information using the learned learning model
including,
The hybrid composite material structure,
It includes a single FRP material, or a bonded structure between an FRP material and a metallic material, and a bonded structure between different FRP materials,
The step of detecting the signal information,
As a lap shear test in which a load is increased in a preset unit using a tensile tester is performed, the hybrid composite material structure or the hybrid composite material structure using an acoustic emission sensor attached to the hybrid composite material structure obtaining an elastic wave emitted from the material constituting the
including,
The step of performing the clustering is,
Extracting a plurality of acoustic emission factor characteristics for the acoustic emission signal information including amplitude, holding time, energy, maximum frequency, and normalized cumulative energy according to time from the acquired acoustic wave, and extracting the plurality of acoustic emission factor characteristics Identifying the failure mechanism of the hybrid composite material or hybrid composite material structure based on the clustering, and linking the identified failure mechanism to the cluster generated by performing the clustering to database the signal information clustered according to the failure mechanism
including,
The step of inputting the new signal information includes:
Performing classification learning on the acoustic emission signal information according to the time at which the clustering was performed using a deep learning-based learning model built to classify the breakage mechanism
including,
The step of detecting the breakage mechanism,
A plurality of sound emission factors including amplitude, holding time, maximum frequency, energy, and normalized cumulative energy for time information extracted from the new signal information are input to the learning model in which the classification learning is performed, and Steps to detect breakage mechanism
including,
The deep learning-based learning model is
According to the ratio of the number of training data and the number of test data set for hyperparameter tuning, the network is composed of multiple nodes and multiple layers, and a learning model is built using a logistic loss function and Relu activation function to build Python ( Iteratively learned to classify the breakage mechanism from the clustered time-dependent acoustic emission signal information using the ANN-based Keras deep learning algorithm in the Python) environment.
A method of detecting a failure mechanism of a hybrid composite material structure comprising a. According to claim 1,
The hybrid composite material structure,
Including that an acoustic emission sensor is attached to a location (attached to a structure/material of interest) where deformation or damage is expected in the hybrid composite structure,
The step of performing the clustering is,
By processing the short time Fourier transform (STFT) or Fast furrier transform (FFT) on the obtained acoustic wave, amplitude, holding time, energy, maximum frequency and normalized cumulative energy according to time Extracting a plurality of acoustic emission factor characteristics for the acoustic emission signal information
A method of detecting a breakage mechanism of a hybrid composite material structure comprising a. delete According to claim 1,
The step of performing clustering according to the breakage mechanism using a plurality of factors according to time extracted from the detected seismic wave,
Clustering the acoustic emission signal information using any one of k-Means, Genetic k-Means, Fuzzy C-Means, Self-Organizing Map (SOM), Gaussian Mixture Model (GMM), and a hierarchical model
A method of detecting a breakage mechanism of a hybrid composite material structure comprising a. According to claim 1,
The step of performing the clustering is,
Databaseizing distance information including distance in the xy coordinate system, Euclidean distance or statistical and stochastic distance from the center of the cluster generated by performing clustering according to the identified breakage mechanism, and signal information according to time
A method of detecting a breakage mechanism of a hybrid composite material structure comprising a. delete delete In the broken mechanism detection system,
a signal detector for detecting signal information emitted from a material constituting the hybrid composite structure;
a clustering unit configured to perform clustering according to the identified breakage mechanism using the detected signal information;
an input unit for inputting new signal information into the deep learning-based learning model learned in order to classify the breakage mechanism using the signal information of the cluster generated by performing the clustering; and
A breakage mechanism detector for detecting a breakage mechanism of the hybrid composite material structure from the new signal information using the learned learning model
including,
The hybrid composite material structure,
It includes a single FRP material, or a bonded structure between an FRP material and a metallic material, and a bonded structure between different FRP materials,
The signal detection unit,
As a lap shear test in which a load is increased in a preset unit using a tensile tester is performed, the hybrid composite material structure or the hybrid composite material structure using an acoustic emission sensor attached to the hybrid composite material structure Including obtaining an elastic wave emitted from the material constituting the
The clustering unit,
Extracting a plurality of acoustic emission factor characteristics for the acoustic emission signal information including amplitude, holding time, energy, maximum frequency, and normalized cumulative energy according to time from the acquired acoustic wave, and extracting the plurality of acoustic emission factor characteristics Identifying the failure mechanism of the hybrid composite material or hybrid composite material structure based on the clustering, and linking the identified failure mechanism to the cluster generated by performing the clustering to database the signal information clustered according to the failure mechanism,
The input unit,
Including performing classification learning on the acoustic emission signal information according to the time when the clustering is performed using a deep learning-based learning model built to classify the breakage mechanism,
The breakage mechanism detection unit,
A plurality of sound emission factors including amplitude, holding time, maximum frequency, energy, and normalized cumulative energy for time information extracted from the new signal information are input to the learning model in which the classification learning is performed, and detecting breakage mechanisms;
The deep learning-based learning model is
According to the ratio of the number of training data and the number of test data set for hyperparameter tuning, a network is composed of multiple nodes and multiple layers, and a learning model is built using the logistic loss function and Relu activation function to build Python ( In Python) environment, using the ANN-based Keras deep learning algorithm, iteratively learned to classify the breakage mechanism from the clustered time-dependent acoustic emission signal information.
Breakage mechanism detection system.

Images