Causal dilated convolutional neural networks for automatic inspection of ultrasonic signals in non-destructive evaluation and structural health monitoring

Mariani, Stefano; Rendu, Quentin; Urbani, Matteo; Sbarufatti, Claudio

doi:10.1016/j.ymssp.2021.107748

Cited by 45 publications

(25 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several ML methods have been developed in the last few years to solve various SHM and damage detection problems, especially by using neural networks (NN) [ 1 , 2 , 3 , 4 , 5 ]. Even though ML methods are already well established in vibration-based SHM [ 6 ], their use in guided wave-based SHM is currently rising [ 7 , 8 , 9 ]. For instance, Roy et al [ 7 ] described an unsupervised learning approach for structural damage identification under varying temperatures based on an NN.…”

Section: Introductionmentioning

confidence: 99%

“…They used a circular array of transducers on an isotropic metal plate with through-holes of different sizes modelled at different locations. Mariani et al [ 9 ] showed improvements in automatic damage detection when using a causal dilated convolutional NN without the need for feature engineering by a human operator. Qiu [ 1 ] studied Gaussian mixture models for GW in SHM systems using measurements from a full-scale fatigue test.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Towards Interpretable Machine Learning for Automated Damage Detection Based on Ultrasonic Guided Waves

Schnur

Goodarzi

Lugovtsova

et al. 2022

Sensors

View full text Add to dashboard Cite

Data-driven analysis for damage assessment has a large potential in structural health monitoring (SHM) systems, where sensors are permanently attached to the structure, enabling continuous and frequent measurements. In this contribution, we propose a machine learning (ML) approach for automated damage detection, based on an ML toolbox for industrial condition monitoring. The toolbox combines multiple complementary algorithms for feature extraction and selection and automatically chooses the best combination of methods for the dataset at hand. Here, this toolbox is applied to a guided wave-based SHM dataset for varying temperatures and damage locations, which is freely available on the Open Guided Waves platform. A classification rate of 96.2% is achieved, demonstrating reliable and automated damage detection. Moreover, the ability of the ML model to identify a damaged structure at untrained damage locations and temperatures is demonstrated.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Towards Interpretable Machine Learning for Automated Damage Detection Based on Ultrasonic Guided Waves

Schnur

Goodarzi

Lugovtsova

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…This approach demonstrates that the achievable spatial accuracy is on the order of the wavelength that corresponds to the main A 0 received mode with frequency content well below 100 kHz. Another paper following the work of Hesser was published by Mariani et al [ 160 ], where the autonomous defects classification is explored with a CNN approach that overcomes the limitations of extensive baseline data archives.…”

Section: Artificial Intelligence and Machine Learningmentioning

confidence: 99%

Ultrasonic Guided-Waves Sensors and Integrated Structural Health Monitoring Systems for Impact Detection and Localization: A Review

Capineri

Bulletti

2021

Sensors

View full text Add to dashboard Cite

This review article is focused on the analysis of the state of the art of sensors for guided ultrasonic waves for the detection and localization of impacts for structural health monitoring (SHM). The recent developments in sensor technologies are then reported and discussed through the many references in recent scientific literature. The physical phenomena that are related to impact event and the related main physical quantities are then introduced to discuss their importance in the development of the hardware and software components for SHM systems. An important aspect of the article is the description of the different ultrasonic sensor technologies that are currently present in the literature and what advantages and disadvantages they could bring in relation to the various phenomena investigated. In this context, the analysis of the front-end electronics is deepened, the type of data transmission both in terms of wired and wireless technology and of online and offline signal processing. The integration aspects of sensors for the creation of networks with autonomous nodes with the possibility of powering through energy harvesting devices and the embedded processing capacity is also studied. Finally, the emerging sector of processing techniques using deep learning and artificial intelligence concludes the review by indicating the potential for the detection and autonomous characterization of the impacts.

show abstract

“…Intuitively, the reason why max pooling can work effectively is that after a feature is found, its precise position is far less important than its relative position with other features [22]. The pooling layer will continuously reduce the spatial size of the data, so the number of parameters and the amount of calculation will also decrease, which also controls the over fitting to a certain extent [23,24].…”

Section: Poolingmentioning

confidence: 99%

Alcoholism via 6-layer customized deep convolution neural network

Zhu

2021

STEMedicine

View full text Add to dashboard Cite

Background: Alcoholism is caused by excessive alcohol into the human body. Alcohol primarily damages the central nervous system of the human body and causes the nervous system function disorder and inhibition. Severe addiction can lead to respiratory circulation center inhibition, paralysis and even death. So far, the diagnosis of alcoholism is done by radiologist's manual CT examination. However, the diagnosis process is time-consuming, subjective and boring for doctors. External factors, such as extreme fatigue, lack of sleep and mental concentration, can easily affect the diagnosis process.Methods: In order to solve this problem, this paper proposed a new neural network based on computer vision, which used deep convolution neural network to diagnose alcoholism automatically. A total of 216 brain images were collected. In the 6-layer customized deep convolution neural network structure, there were four convolution layers and two fully connected layers, and each convolution layer was connected with a pooling layer.Results: The results showed that the accuracy, sensitivity, specificity, precision, F1, MCC and FMI were 95.96%±1.44%, 95.96%±1.66%, 95.95%±1.67%, 95.73%±1.72%, 95.84%±1.48%, 91.92%±2.87% and 95.84%±1.48% respectively.Conclusion: It can be concluded from comparison results that the proposed neural network structure is more effective than four state-of-the-art approaches. The proposed method has high accuracy and can be used as a diagnostic method for alcoholism.

show abstract

Causal dilated convolutional neural networks for automatic inspection of ultrasonic signals in non-destructive evaluation and structural health monitoring

Cited by 45 publications

References 43 publications

Towards Interpretable Machine Learning for Automated Damage Detection Based on Ultrasonic Guided Waves

Towards Interpretable Machine Learning for Automated Damage Detection Based on Ultrasonic Guided Waves

Ultrasonic Guided-Waves Sensors and Integrated Structural Health Monitoring Systems for Impact Detection and Localization: A Review

Alcoholism via 6-layer customized deep convolution neural network

Contact Info

Product

Resources

About