In this paper, a novel method is proposed based on a windowed one-dimensional convolutional neural network (1D CNN) for multiclass damage identification using vibration responses of a full-scale bridge. The measured data is first augmented by extracting samples of windows of raw acceleration time-series to alleviate the problem of a limited training dataset. 1D CNN is developed to classify the windowed time-series into multiple damage classes. The damage is quantified using the predicted class probabilities, and the damage is localized if the predicted class is equal to the assigned damage class, exceeding a threshold associated with majority voting. The proposed network is optimally tuned with respect to various hyper-parameters such as window size, random initialization of weights, etc., to achieve the best classification performance using a global 1D CNN model. The proposed method is validated using the Z24 bridge benchmark data for multiclass classification for two different damage scenarios, namely, pier settlement and rupture of tendons, 1Sony, September 22, 2021 under the various extent of the damage. The damage identification is carried out on various bridge components to collectively identify the structural component with a damaged signature. The resultsshow that the proposed windowed 1D CNN method achieves an accuracy of 97%, and performs well with different types of damage.
Multivariate time-series classification problems are found in many industrial settings; for example, fault detection in a manufacturing process by monitoring sensors signals. It is difficult to obtain large labeled datasets in these settings, for reasons such as limitations in the automatic recording, the need for expert root-cause analysis, and the very limited access to human experts. Therefore, methods that perform classification in a label efficient manner are useful for building and deploying machine learning models in the industrial setting. In this work, we apply a self-supervised learning method called Contrastive Predictive Coding (CPC) to classification tasks on three industrial multivariate time-series datasets. First, the CPC neural network (CPC base) is trained with a large number of unlabeled time-series data instances. Then, a standard supervised classifier such as a multi-layer perception (MLP) is trained on available labeled data using the output embeddings from the pre-trained CPC base. On all three classification datasets, we see increased label efficiency (ability to reach a goal accuracy level with less labeled examples). In the low data regime (10's or few 100's of labeled examples), the CPC pre-trained model achieves high accuracy with up to 15x less labels than a model trained only on labeled data. We also conduct experiments to evaluate the usefulness of CPC pre-trained classifiers as base models to start an active learning loop, and find that uncertainty sampling does not perform significantly better than random sampling during the initial queries.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.