Multi-feature Fusion and Damage Identification of Large Generator Stator Insulation Based on Lamb Wave Detection and SVM Method

Li, Ruihua; Gu, Haojie; Hu, Bo; She, Zhifeng

doi:10.3390/s19173733

Cited by 17 publications

(8 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…ANNs are a mathematical model which simulates a biological neural system, and hence has the ability to deal with non-linear problems. Compared to normal ANNs, Li et al proposed a multi-feature fusion method to detect damage based on a support vector machine (SVM) algorithm, which was proved to overperform ANNs [ 28 ]. The performance of ANN and SVM for impact detection were compared by Yue et al [ 29 ] to confirm their improved performance.…”

Section: Machine Learning Methodologiesmentioning

confidence: 99%

Structural Health Monitoring Impact Classification Method Based on Bayesian Neural Network

Seno

Khodaei

2022

Polymers

View full text Add to dashboard Cite

This paper proposes a novel method for multi-class classification and uncertainty quantification of impact events on a flat composite plate with a structural health monitoring (SHM) system by using a Bayesian neural network (BNN). Most of the existing research in passive sensing has focused on deterministic approaches for impact detection and characterization. However, there are variability in impact location, angle and energy in real operational conditions which results in uncertainty in the diagnosis. Therefore, this paper proposes a reliability-based impact characterization method based on BNN for the first time. Impact data are acquired by a passive sensing system of piezoelectric (PZT) sensors. Features extracted from the sensor signals, such as their transferred energy, frequency at maximum amplitude and time interval of the largest peak, are used to develop a BNN for impact classification (i.e., energy level). To test the robustness and reliability of the proposed model to impact variability, it is trained with perpendicular impacts and tested by variable angle impacts. The same dataset is further applied in a method called multi-artificial neural network (multi-ANN) to compare its ability in uncertainty quantification and its computational efficiency against the BNN for validation of the developed meta-model. It is demonstrated that both the BNN and multi-ANN can measure the uncertainty and confidence of the diagnosis from the prediction results. Both have very high performance in classifying impact energies when the networks are trained and tested with perpendicular impacts of different energy and location, with 94% and 98% reliable predictions for BNN and multi-ANN, respectively. However, both metamodels struggled to detect new impact scenarios (angled impacts) when the data set was not used in the development stage and only used for testing. Including additional features improved the performance of the networks in regularization; however, not to the acceptable accuracy. The BNN significantly outperforms the multi-ANN in computational time and resources. For perpendicular impacts, both methods can reach a reliable accuracy, while for angled impacts, the accuracy decreases but the uncertainty provides additional information that can be further used to improve the classification.

show abstract

Section: Machine Learning Methodologiesmentioning

confidence: 99%

Structural Health Monitoring Impact Classification Method Based on Bayesian Neural Network

Seno

Khodaei

2022

Polymers

View full text Add to dashboard Cite

show abstract

“…The support vectors are the data points that are closest to the hyperplane and have the greatest impact on its position. The margin is the space between the hyperplane and the support vectors [ 16 , 17 ].…”

Section: Analysis Processmentioning

confidence: 99%

“…The vectors corresponding to these data points on the margin boundaries are the support vectors, these vectors support the determination of the model by limiting the margin width d . They give the algorithm its name of support vector machine [ 17 , 21 ]. Consider the data point vector

on the positive margin boundary and data point vector

on the negative margin boundary, as follows:

…”

Section: Analysis Processmentioning

confidence: 99%

Suitability Analysis of Machine Learning Algorithms for Crack Growth Prediction Based on Dynamic Response Data

Omar

Khan

Starr

2023

Sensors

View full text Add to dashboard Cite

Machine learning has the potential to enhance damage detection and prediction in materials science. Machine learning also has the ability to produce highly reliable and accurate representations, which can improve the detection and prediction of damage compared to the traditional knowledge-based approaches. These approaches can be used for a wide range of applications, including material design; predicting material properties; identifying hidden relationships; and classifying microstructures, defects, and damage. However, researchers must carefully consider the appropriateness of various machine learning algorithms, based on the available data, material being studied, and desired knowledge outcomes. In addition, the interpretability of certain machine learning models can be a limitation in materials science, as it may be difficult to understand the reasoning behind predictions. This paper aims to make novel contributions to the field of material engineering by analyzing the compatibility of dynamic response data from various material structures with prominent machine learning approaches. The purpose of this is to help researchers choose models that are both effective and understandable, while also enhancing their understanding of the model’s predictions. To achieve this, this paper analyzed the requirements and characteristics of commonly used machine learning algorithms for crack propagation in materials. This analysis assisted the authors in selecting machine learning algorithms (K nearest neighbor, Ridge, and Lasso regression) to evaluate the dynamic response of aluminum and ABS materials, using experimental data from previous studies to train the models. The results showed that natural frequency was the most significant predictor for ABS material, while temperature, natural frequency, and amplitude were the most important predictors for aluminum. Crack location along samples had no significant impact on either material. Future work could involve applying the discussed techniques to a wider range of materials under dynamic loading conditions.

show abstract

“…It is randomly divided into training set and test set with a ratio of 8:2. Because the dimension of the samples is too high, the Principal Component Analysis (PCA) method [ 30 , 31 ] is utilized for pre-processing. Samples of the training set are used as input of the SVM classifier, and the label set is formed by the defect classes of the corresponding samples (0 and 1 represent crack and delamination, respectively).…”

Section: Identification Modelmentioning

confidence: 99%

An Improved RAPID Imaging Method of Defects in Composite Plate Based on Feature Identification by Machine Learning

Deng

Zhang

et al. 2022

Sensors

View full text Add to dashboard Cite

The RAPID (reconstruction algorithm for probabilistic inspection of defect) method based on Lamb wave detection is an effective method to give the position information of a defect in composite plate. In this paper, an improved RAPID imaging method based on machine learning (ML) is proposed to precisely visualize the location and features of defects in composite plate. First, the specific feature information of the defect, such as type, size and direction, can be identified by analyzing the detection signals through multiple machine learning models. Then, according to the obtained defect features, the scaling parameter β of the RAPID method which controls the size of the elliptical area is revised, and weights are set to the important detection paths which are related to defect features to realize precise defect imaging. The simulation results show that the proposed method can intuitively characterize the location and related feature information of the defect, and effectively improve the accuracy of defect imaging.

show abstract

Multi-feature Fusion and Damage Identification of Large Generator Stator Insulation Based on Lamb Wave Detection and SVM Method

Cited by 17 publications

References 30 publications

Structural Health Monitoring Impact Classification Method Based on Bayesian Neural Network

Structural Health Monitoring Impact Classification Method Based on Bayesian Neural Network

Suitability Analysis of Machine Learning Algorithms for Crack Growth Prediction Based on Dynamic Response Data

An Improved RAPID Imaging Method of Defects in Composite Plate Based on Feature Identification by Machine Learning

Contact Info

Product

Resources

About