A Novelty Detection Approach for Tendons of Prestressed Concrete Bridges Based on a Convolutional Autoencoder and Acceleration Data

Lee, Kanghyeok; Jeong, Seunghoo; Sim, Sung‐Han; Shin, Do Hyoung

doi:10.3390/s19071633

Cited by 23 publications

(20 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Tables 4 and 5 display examples of confusion matrices for the DNN and CNN models when the accuracy is 68%, which is similar to the average accuracy of both models. In particular, the false negative rate (FNR) is more important than the false positive rate (FPR) from the perspective of practical and usable damage prediction models [56]. Therefore, the FNRs and FPRs of both the DNN and CNN models were compared using confusion matrices.…”

Section: Resultsmentioning

confidence: 99%

Prediction of Heavy Rain Damage Using Deep Learning

Lee

Choi

Shin

et al. 2020

Water

Self Cite

View full text Add to dashboard Cite

Heavy rain damage prediction models were developed with a deep learning technique for predicting the damage to a region before heavy rain damage occurs. As a dependent variable, a damage scale comprising three categories (minor, significant, severe) was used, and meteorological data 7 days before the damage were used as independent variables. A deep neural network (DNN), convolutional neural network (CNN), and recurrent neural network (RNN), which are representative deep learning techniques, were employed for the model development. Each model was trained and tested 30 times to evaluate the predictive performance. As a result of evaluating the predicted performance, the DNN-based model and the CNN-based model showed good performance, and the RNN-based model was analyzed to have relatively low performance. For the DNN-based model, the convergence epoch of the training showed a relatively wide distribution, which may lead to difficulties in selecting an epoch suitable for practical use. Therefore, the CNN-based model would be acceptable for the heavy rain damage prediction in terms of the accuracy and robustness. These results demonstrated the applicability of deep learning in the development of the damage prediction model. The proposed prediction model can be used for disaster management as the basic data for decision making.

show abstract

Section: Resultsmentioning

confidence: 99%

Prediction of Heavy Rain Damage Using Deep Learning

Lee

Choi

Shin

et al. 2020

Water

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, recently developed recursive canonical correlation analysis algorithm do not require a reference for accurate assessment [17]. Another approach to ND also relies on a distance metric, but this time the distance is not measured from reference points stored in memory but from a sample's reconstructed version [5,[18][19][20][21]. In the last quoted approach, the boundary is defined by a trained model.…”

Section: Novelty Detection In Condition Monitoringmentioning

confidence: 99%

Comparison of Novelty Detection Methods for Detection of Various Rotary Machinery Faults

Górski

Jabłoński

Heesch

et al. 2021

Sensors

View full text Add to dashboard Cite

Condition monitoring is an indispensable element related to the operation of rotating machinery. In this article, the monitoring system for the parallel gearbox was proposed. The novelty detection approach is used to develop the condition assessment support system, which requires data collection for a healthy structure. The measured signals were processed to extract quantitative indicators sensitive to the type of damage occurring in this type of structure. The indicator’s values were used for the development of four different novelty detection algorithms. Presented novelty detection models operate on three principles: feature space distance, probability distribution, and input reconstruction. One of the distance-based models is adaptive, adjusting to new data flowing in the form of a stream. The authors test the developed algorithms on experimental and simulation data with a similar distribution, using the training set consisting mainly of samples generated by the simulator. Presented in the article results demonstrate the effectiveness of the trained models on both data sets.

show abstract

“…The BSSM of tall structures for the most part utilizes vibration information [3]. The danger represents the progressions of auxiliary parameters, for example, the firmness and damping coefficients [4].…”

Section: Introductionmentioning

confidence: 99%

An Efficient Data Mining Technique for Structural Strength Monitoring System

Rao¹,

Chaparala²

2021

ISI

View full text Add to dashboard Cite

A fundamental target of strength monitoring frameworks for different structures is to analyze the condition of the structure and to assess its conceivable danger and furthermore to investigation, identification, and characterization of danger in complex structures is a critical part of auxiliary strength checking. The capacities are browsed as lexicon of time-recurrence movement and scaled variants of a basic Gaussian hypothesis work. This word reference is likewise adjusted to utilize genuine estimated information. Characterization is then accomplished by coordinating the removed damage includes in the time-frequency. In this paper, we utilize our model to assess our information mining approach for the fault checking. The balanced scratch-off and high-pass sifting strategies are consolidated adequately to take care of basic issues in numerical reconciliation signs gathered from sensors are disintegrated into direct blends of very confined Gaussian capacities utilizing the coordinating significance decay calculation. The combination exactness is enhanced and contrasted with former numerical integrators. Rough set analysis uses only internal knowledge and does not rely on prior model assumption as fuzzy set methods or probabilistic models do. In this manuscript a novel hybrid algorithm combining the features of Rough set Support vector machine (Rs-SVM) classified structures and Rough set Artificial Neural Network (Rs-ANN) classified structures are used. At long last the vertices of the structure of different types are connected and analysed by the Hybrid algorithm and furthermore to additionally enhance order execution, the data gathered from numerous sensors is incorporated utilizing a Bayesian sensor combination approach.

show abstract

A Novelty Detection Approach for Tendons of Prestressed Concrete Bridges Based on a Convolutional Autoencoder and Acceleration Data

Cited by 23 publications

References 34 publications

Prediction of Heavy Rain Damage Using Deep Learning

Prediction of Heavy Rain Damage Using Deep Learning

Comparison of Novelty Detection Methods for Detection of Various Rotary Machinery Faults

An Efficient Data Mining Technique for Structural Strength Monitoring System

Contact Info

Product

Resources

About