scholarly journals Artificial Intelligence-Based Bolt Loosening Diagnosis Using Deep Learning Algorithms for Laser Ultrasonic Wave Propagation Data

Sensors ◽  
2020 ◽  
Vol 20 (18) ◽  
pp. 5329
Author(s):  
Dai Quoc Tran ◽  
Ju-Won Kim ◽  
Kassahun Demissie Tola ◽  
Wonkyu Kim ◽  
Seunghee Park
Keyword(s):  
Neural Network ◽  
Signal Processing ◽  
Wave Propagation ◽  
Deep Learning ◽  
Ultrasonic Wave ◽  
Ultrasonic Waves ◽  
Support Vector ◽  
Laser Ultrasonic ◽  
Data Set ◽  
Ultrasonic Wave Propagation

The application of deep learning (DL) algorithms to non-destructive evaluation (NDE) is now becoming one of the most attractive topics in this field. As a contribution to such research, this study aims to investigate the application of DL algorithms for detecting and estimating the looseness in bolted joints using a laser ultrasonic technique. This research was conducted based on a hypothesis regarding the relationship between the true contact area of the bolt head-plate and the guided wave energy lost while the ultrasonic waves pass through it. First, a Q-switched Nd:YAG pulsed laser and an acoustic emission sensor were used as exciting and sensing ultrasonic signals, respectively. Then, a 3D full-field ultrasonic data set was created using an ultrasonic wave propagation imaging (UWPI) process, after which several signal processing techniques were applied to generate the processed data. By using a deep convolutional neural network (DCNN) with a VGG-like architecture based regression model, the estimated error was calculated to compare the performance of a DCNN on different processed data set. The proposed approach was also compared with a K-nearest neighbor, support vector regression, and deep artificial neural network for regression to demonstrate its robustness. Consequently, it was found that the proposed approach shows potential for the incorporation of laser-generated ultrasound and DL algorithms. In addition, the signal processing technique has been shown to have an important impact on the DL performance for automatic looseness estimation.

scholarly journals High Frequency Ultrasonic Wave Propagation in  Anisotropic Materials

2021 ◽  
Author(s):  
◽  
Andrew Paul Dawson
Keyword(s):  
Wave Propagation ◽  
Ultrasonic Wave ◽  
High Frequency ◽  
Guided Wave ◽  
Ultrasonic Waves ◽  
Tissue Characterisation ◽  
Ultrasonic Wave Propagation ◽  
Matching Layer ◽  
Fibrous Tissues ◽  
Wave Modes

<p>The influence of highly regular, anisotropic, microstructured materials on high frequency ultrasonic wave propagation was investigated in this work. Microstructure, often only treated as a source of scattering, significantly influences high frequency ultrasonic waves, resulting in unexpected guided wave modes. Tissues, such as skin or muscle, are treated as homogeneous by current medical ultrasound systems, but actually consist of highly anisotropic micron-sized fibres. As these systems increase towards 100 MHz, these fibres will significantly influence propagating waves leading to guided wave modes. The effect of these modes on image quality must be considered. However, before studies can be undertaken on fibrous tissues, wave propagation in more ideal structures must be first understood. After the construction of a suitable high frequency ultrasound experimental system, finite element modelling and experimental characterisation of high frequency (20-200 MHz) ultrasonic waves in ideal, collinear, nanostructured alumina was carried out. These results revealed interesting waveguiding phenomena, and also identified the potential and significant advantages of using a microstructured material as an alternative acoustic matching layer in ultrasonic transducer design. Tailorable acoustic impedances were achieved from 4-17 MRayl, covering the impedance range of 7-12 MRayl most commonly required by transducer matching layers. Attenuation coefficients as low as 3.5 dBmm-1 were measured at 100 MHz, which is excellent when compared with 500 dBmm-1 that was measured for a state of the art loaded epoxy matching layer at the same frequency. Reception of ultrasound without the restriction of critical angles was also achieved, and no dispersion was observed in these structures (unlike current matching layers) until at least 200 MHz. In addition, to make a significant step forward towards high frequency tissue characterisation, novel microstructured poly(vinyl alcohol) tissue-mimicking phantoms were also developed. These phantoms possessed acoustic and microstructural properties representative of fibrous tissues, much more realistic than currently used homogeneous phantoms. The attenuation coefficient measured along the direction of PVA alignment in an example phantom was 8 dBmm-1 at 30 MHz, in excellent agreement with healthy human myocardium. This method will allow the fabrication of more realistic and repeatable phantoms for future high frequency tissue characterisation studies.</p>

Probabilistic characterization of ultrasonic wave propagation in wood poles

2012 ◽  
Vol 39 (4) ◽  
pp. 484-493 ◽  
Author(s):  
Fernando Tallavo ◽  
Mahesh D. Pandey ◽  
Giovanni Cascante
Keyword(s):  
Wave Propagation ◽  
Ultrasonic Wave ◽  
Cross Section ◽  
Wave Velocity ◽  
Orthotropic Material ◽  
Ultrasonic Waves ◽  
Reference Wave ◽  
Ultrasonic Wave Propagation ◽  
Probabilistic Characterization

Wood poles are widely used in North America to support power electric transmission and distribution lines. Wood poles are continuously exposed to wide ranging temperature and moisture conditions, making them vulnerable to internal decay and rotting. The resulting loss of strength makes the poles vulnerable to failure under adverse weather conditions, such as wind and snow storms. These failures can result in forced outages and customer disruptions with significant economic losses. Ultrasonic testing is a non-destructive method that has been used for detection of internal deterioration of in-service wood poles, which is based on the comparison of the measured wave velocity with a reference wave velocity associated with sound wood. The current ultrasonic methods assume that the reference wave velocity for a given wood species is constant in a pole cross section. This approach is simplistic because wood is an orthotropic material with highly variable material properties. This paper presents a method for probabilistic characterization of ultrasonic wave propagation in wood poles considering wood as an orthotropic material. A better understanding and characterization of ultrasonic wave propagation in a pole cross section will contribute to improve the condition assessment of in-service wood poles based on ultrasonic tests. As an example, P-wave velocity, surface waves, frequency response function, and magnitude spectrum area are used to characterize the propagation of ultrasonic waves at a cross section of a Douglas-fir pole of 25 cm diameter.

scholarly journals Out-Of-Plane Permeability Evaluation of Carbon Fiber Preforms by Ultrasonic Wave Propagation

Materials ◽  
2020 ◽  
Vol 13 (12) ◽  
pp. 2684 ◽  
Author(s):  
Francesca Lionetto ◽  
Francesco Montagna ◽  
Alfonso Maffezzoli
Keyword(s):  
Wave Propagation ◽  
Carbon Fiber ◽  
Ultrasonic Wave ◽  
Ultrasonic Transducer ◽  
Gravimetric Method ◽  
Ultrasonic Waves ◽  
Ultrasonic Wave Propagation ◽  
Out Of Plane ◽  
Set Up ◽  
Fiber Preforms

Out-of-plane permeability of reinforcement preforms is of crucial importance in the infusion of large and thick composite panels, but so far, there are no standard experimental methods for its determination. In this work, an experimental set-up for the measurement of unsaturated through thickness permeability based on the ultrasonic wave propagation in pulse echo mode is presented. A single ultrasonic transducer, working both as emitter and receiver of ultrasonic waves, was used to monitor the through thickness flow front during a vacuum assisted resin infusion experiment. The set-up was tested on three thick carbon fiber preforms, obtained by stacking thermal bonding of balanced or unidirectional plies either by automated fiber placement either by hand lay-up of unidirectional plies. The ultrasonic data were used to calculate unsaturated out-of-plane permeability using Darcy’s law. The permeability results were compared with saturated out-of-plane permeability, determined by a traditional gravimetric method, and validated by some analytical models. The results demonstrated the feasibility and potential of the proposed set-up for permeability measurements thanks to its noninvasive character and the one-side access.

Comparing Machine Learning and Deep Learning Methods for Real-Time Crash Prediction

2019 ◽  
Vol 2673 (8) ◽  
pp. 169-178 ◽  
Author(s):  
Athanasios Theofilatos ◽  
Cong Chen ◽  
Constantinos Antoniou
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Deep Learning ◽  
Real Time ◽  
Naive Bayes ◽  
Naïve Bayes ◽  
Support Vector ◽  
Data Set ◽  
Real Time Traffic ◽  
The Impact

Although there are numerous studies examining the impact of real-time traffic and weather parameters on crash occurrence on freeways, to the best of the authors’ knowledge there are no studies which have compared the prediction performances of machine learning (ML) and deep learning (DL) models. The present study adds to current knowledge by comparing and validating ML and DL methods to predict real-time crash occurrence. To achieve this, real-time traffic and weather data from Attica Tollway in Greece were linked with historical crash data. The total data set was split into training/estimation (75%) and validation (25%) subsets, which were then standardized. First, the ML and DL prediction models were trained/estimated using the training data set. Afterwards, the models were compared on the basis of their performance metrics (accuracy, sensitivity, specificity, and area under curve, or AUC) on the test set. The models considered were k-nearest neighbor, Naïve Bayes, decision tree, random forest, support vector machine, shallow neural network, and, lastly, deep neural network. Overall, the DL model seems to be more appropriate, because it outperformed all other candidate models. More specifically, the DL model managed to achieve a balanced performance among all metrics compared with other models (total accuracy = 68.95%, sensitivity = 0.521, specificity = 0.77, AUC = 0.641). It is surprising though that the Naïve Bayes model achieved a good performance despite being far less complex than other models. The study findings are particularly useful, because they provide a first insight into performance of ML and DL models.

Ultrasonic Wave Propagation in CdS Phonon Maser Structures

1973 ◽  
Vol 51 (12) ◽  
pp. 1350-1358 ◽  
Author(s):  
J. Vrba ◽  
R. R. Haering
Keyword(s):  
Wave Propagation ◽  
Ultrasonic Wave ◽  
Mode Conversion ◽  
Ultrasonic Waves ◽  
Coupling Constant ◽  
Velocity Of Sound ◽  
Angular Variation ◽  
Ultrasonic Wave Propagation ◽  
Piezoelectric Coupling ◽  
Maser Action

An analysis of phonon maser action in CdS is given which includes the complications arising from the presence of off-axis ultrasonic waves. The treatment includes the angular variation of the velocity of sound and of the piezoelectric coupling constant and takes account of mode conversion at the cavity walls. Numerical results are given for CdS maser structures.

scholarly journals High Frequency Ultrasonic Wave Propagation in  Anisotropic Materials

2021 ◽  
Author(s):  
◽  
Andrew Paul Dawson
Keyword(s):  
Wave Propagation ◽  
Ultrasonic Wave ◽  
High Frequency ◽  
Guided Wave ◽  
Ultrasonic Waves ◽  
Tissue Characterisation ◽  
Ultrasonic Wave Propagation ◽  
Matching Layer ◽  
Fibrous Tissues ◽  
Wave Modes

<p>The influence of highly regular, anisotropic, microstructured materials on high frequency ultrasonic wave propagation was investigated in this work. Microstructure, often only treated as a source of scattering, significantly influences high frequency ultrasonic waves, resulting in unexpected guided wave modes. Tissues, such as skin or muscle, are treated as homogeneous by current medical ultrasound systems, but actually consist of highly anisotropic micron-sized fibres. As these systems increase towards 100 MHz, these fibres will significantly influence propagating waves leading to guided wave modes. The effect of these modes on image quality must be considered. However, before studies can be undertaken on fibrous tissues, wave propagation in more ideal structures must be first understood. After the construction of a suitable high frequency ultrasound experimental system, finite element modelling and experimental characterisation of high frequency (20-200 MHz) ultrasonic waves in ideal, collinear, nanostructured alumina was carried out. These results revealed interesting waveguiding phenomena, and also identified the potential and significant advantages of using a microstructured material as an alternative acoustic matching layer in ultrasonic transducer design. Tailorable acoustic impedances were achieved from 4-17 MRayl, covering the impedance range of 7-12 MRayl most commonly required by transducer matching layers. Attenuation coefficients as low as 3.5 dBmm-1 were measured at 100 MHz, which is excellent when compared with 500 dBmm-1 that was measured for a state of the art loaded epoxy matching layer at the same frequency. Reception of ultrasound without the restriction of critical angles was also achieved, and no dispersion was observed in these structures (unlike current matching layers) until at least 200 MHz. In addition, to make a significant step forward towards high frequency tissue characterisation, novel microstructured poly(vinyl alcohol) tissue-mimicking phantoms were also developed. These phantoms possessed acoustic and microstructural properties representative of fibrous tissues, much more realistic than currently used homogeneous phantoms. The attenuation coefficient measured along the direction of PVA alignment in an example phantom was 8 dBmm-1 at 30 MHz, in excellent agreement with healthy human myocardium. This method will allow the fabrication of more realistic and repeatable phantoms for future high frequency tissue characterisation studies.</p>

scholarly journals Detection of Malicious Software by Analyzing Distinct Artifacts Using Machine Learning and Deep Learning Algorithms

Electronics ◽  
2021 ◽  
Vol 10 (14) ◽  
pp. 1694
Author(s):  
Mathew Ashik ◽  
A. Jyothish ◽  
S. Anandaram ◽  
P. Vinod ◽  
Francesco Mercaldo ◽  
...  
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Deep Learning ◽  
Support Vector ◽  
Malware Analysis ◽  
Learning Approaches ◽  
Dynamic Features ◽  
System Calls ◽  
Prevention Methods ◽  
Structural Aspects

Malware is one of the most significant threats in today’s computing world since the number of websites distributing malware is increasing at a rapid rate. Malware analysis and prevention methods are increasingly becoming necessary for computer systems connected to the Internet. This software exploits the system’s vulnerabilities to steal valuable information without the user’s knowledge, and stealthily send it to remote servers controlled by attackers. Traditionally, anti-malware products use signatures for detecting known malware. However, the signature-based method does not scale in detecting obfuscated and packed malware. Considering that the cause of a problem is often best understood by studying the structural aspects of a program like the mnemonics, instruction opcode, API Call, etc. In this paper, we investigate the relevance of the features of unpacked malicious and benign executables like mnemonics, instruction opcodes, and API to identify a feature that classifies the executable. Prominent features are extracted using Minimum Redundancy and Maximum Relevance (mRMR) and Analysis of Variance (ANOVA). Experiments were conducted on four datasets using machine learning and deep learning approaches such as Support Vector Machine (SVM), Naïve Bayes, J48, Random Forest (RF), and XGBoost. In addition, we also evaluate the performance of the collection of deep neural networks like Deep Dense network, One-Dimensional Convolutional Neural Network (1D-CNN), and CNN-LSTM in classifying unknown samples, and we observed promising results using APIs and system calls. On combining APIs/system calls with static features, a marginal performance improvement was attained comparing models trained only on dynamic features. Moreover, to improve accuracy, we implemented our solution using distinct deep learning methods and demonstrated a fine-tuned deep neural network that resulted in an F1-score of 99.1% and 98.48% on Dataset-2 and Dataset-3, respectively.

Sign in / Sign up

Export Citation Format

Share Document