1-D CNNs for structural damage detection: Verification on a structural health monitoring benchmark data

Abdeljaber, Osama; Avcı, Onur; Kıranyaz, Serkan; Boashash, Boualem; Sodano, Henry A.; Inman, Daniel J.

doi:10.1016/j.neucom.2017.09.069

Cited by 385 publications

(196 citation statements)

References 33 publications

Supporting

Mentioning

161

Contrasting

Unclassified

Order By: Relevance

“…This is why they are often referred to as, "2D CNNs". As an alternative, a modified version of 2D CNNs called 1D Convolutional Neural Networks (1D CNNs) has recently been developed [47][48][49][50][51][52][53][54][55][56]. These studies have shown that for certain applications 1D CNNs are advantageous and thus preferable to their 2D counterparts in dealing with 1D signals due to the following reasons:…”

Section: D Convolutional Neural Networkmentioning

confidence: 99%

“…It was noticed that the process of generating the data required to train the 1D CNNs in [50,52,71] requires a large number of measurement sessions especially for a large civil structure. Therefore, Avci et al in [53] and then Abdeljaber et al in [54] developed a novel approach based on 1D CNNs, which require significantly less effort and labeled data for training. This approach was successfully tested over the data provided under the Experimental Phase II of the SHM Benchmark Problem [72].…”

Section: Figure 11mentioning

confidence: 99%

“…To incorporate such drawbacks, in 2015 Kiranyaz et al [47] proposed the first compact and adaptive 1D CNNs to operate directly on patient-specific ECG data. In a relatively short time, 1D CNNs have become popular with a stateof-the-art performance in various signal processing applications such as early arrhythmia detection in electrocardiogram (ECG) beats [47][48][49], structural health monitoring and structural damage detection [50][51][52][53][54], high power engine fault monitoring [55] and real-time monitoring of high-power circuitry [56]. Furthermore, two recent studies have utilized 1D CNNs for damage detection in bearings [57][58][59][60].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

1-D Convolutional Neural Networks for Signal Processing Applications

Kıranyaz

İnce

Abdeljaber

et al. 2019

ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

252

View full text Add to dashboard Cite

During the last decade, Convolutional Neural Networks (CNNs) have become the de facto standard for various Computer Vision and Machine Learning operations. CNNs are feed-forward Artificial Neural Networks (ANNs) with alternating convolutional and subsampling layers. Deep 2D CNNs with many hidden layers and millions of parameters have the ability to learn complex objects and patterns providing that they can be trained on a massive size visual database with ground-truth labels. With a proper training, this unique ability makes them the primary tool for various engineering applications for 2D signals such as images and video frames. Yet, this may not be a viable option in numerous applications over 1D signals especially when the training data is scarce or application-specific. To address this issue, 1D CNNs have recently been proposed and immediately achieved the state-of-the-art performance levels in several applications such as personalized biomedical data classification and early diagnosis, structural health monitoring, anomaly detection and identification in power electronics and motor-fault detection. Another major advantage is that a real-time and low-cost hardware implementation is feasible due to the simple and compact configuration of 1D CNNs that perform only 1D convolutions (scalar multiplications and additions). This paper presents a comprehensive review of the general architecture and principals of 1D CNNs along with their major engineering applications, especially focused on the recent progress in this field. Their state-of-the-art performance is highlighted concluding with their unique properties. The benchmark datasets and the principal 1D CNN software used in those applications are also publically shared in a dedicated website.

show abstract

Section: D Convolutional Neural Networkmentioning

confidence: 99%

Section: Figure 11mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

1-D Convolutional Neural Networks for Signal Processing Applications

Kıranyaz

İnce

Abdeljaber

et al. 2019

ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

252

View full text Add to dashboard Cite

show abstract

“…Recently, the use of CNNs for damage detection and identification in civil structural systems based on vibration response data has been recognized as a promising approach. [26][27][28] Although the roots of CNN go back to 1960s, 29,30 the CNN framework is based on LeNet-5 that was developed later by LeCun et al 31 LeNet-5 uses three main ideas, (a) local receptive fields, (b) shared weights, and (c) spatial or temporal subsampling to ensure that the features would be invariant to shift, scale, and distortion. 32 In LeNet-5, the input image is convolved with trainable filters.…”

Section: Introductionmentioning

confidence: 99%

Vibration‐based structural condition assessment using convolution neural networks

Khodabandehlou

Pekcan

Fadali

2018

Struct Control Health Monit

103

View full text Add to dashboard Cite

Summary A novel vibration‐based structural health monitoring (SHM) approach that uses two‐dimensional deep convolution neural networks (CNN) is introduced. The CNN extracts the features from acceleration response histories and drastically reduces the dimension of response history to make damage state classification possible with limited number of acceleration measurements. The proposed method was validated, and its applicability and efficiency were demonstrated using vibration response data recorded during the shake‐table testing of a one‐fourth–scale model of a reinforced concrete highway bridge. The proposed method predicted predefined damage states with 100% accuracy using recorded (acceleration) vibration response data. The method was shown to be robust and sensitive to very small changes in structural condition. It is also noted that the CNN‐based SHM method is scalable to any large number of damage states (including extent and location) with suitable network training. The required training data may be generated analytically using a nonlinear finite element model.

show abstract

“…Early investigations using wavelet examination were led to neighborhood damage recognition in machinery [2,3]. For damage identification, nonparametric worldwide damage identification techniques utilize measurable intends to dissect the vibration reaction of the structure [4]. For all of a sudden happened structural failure, there is an oscillatory process starting from damage happening and proceeding to the minute when the irregular displacement amplitude achieves steady state.…”

Section: Introductionmentioning

confidence: 99%

Optimal Method for Identification of Cracks in Different Beams using Fuzzy with Elephant Based Neural Network

Pitchaiah*,

Rao

2019

IJITEE

View full text Add to dashboard Cite

Failure of Structures i.e., beams can be avoided by identifying the damage in the structure at its beginning and proper retrofitting. Recently, the researchers created a structure to recognize crack damage using a cracked beam component model that originates from the fracture mechanics and local flexibility rules. The present work exhibits the analysis of cracked beam with a machine learning model to assess the stiffness of the structure. Here Fuzzy Optimal Neural Network (FONN) is considered, in addition, the stiffness reduction technique, especially concerning thick beams, is featured with a survey of other crack models. The extricated model data are utilized to conversely recognize the cracks with the cracked beam component model through a model updating technique. The optimal Neural Network based stiffness computation utilizes a global searching procedure using Adaptive Elephant Herding Optimization (AEHO) to identify the number of cracks in various beams. From the proposed model, the attained results are compared with the existing research work, and other optimization and machine learning models.

show abstract

1-D CNNs for structural damage detection: Verification on a structural health monitoring benchmark data

Cited by 385 publications

References 33 publications

1-D Convolutional Neural Networks for Signal Processing Applications

1-D Convolutional Neural Networks for Signal Processing Applications

Vibration‐based structural condition assessment using convolution neural networks

Optimal Method for Identification of Cracks in Different Beams using Fuzzy with Elephant Based Neural Network

Contact Info

Product

Resources

About