Convolutional neural network-based data anomaly detection method using multiple information for structural health monitoring

Abstract

Structural health monitoring (SHM) is used worldwide for managing and maintaining civil infrastructures. SHM systems have produced huge amounts of data, but the effective monitoring, mining, and utilization of this data still need in-depth study. SHM data generally includes multiple types of anomalies caused by sensor faults or system malfunctions that can disturb structural analysis and assessment. In the routine data pre-processing, multiple signal processing techniques are required to detect the anomalies, respectively, which is inefficient. The large variations of extracted features from massive SHM data make the data anomaly detection techniques prone to be over-processed or under-processed. Even with expert intervention, the parameter tuning, associated with multiple data preprocessing methods, is still a challenge, making the procedure expensive and inefficient. In addition, one data anomaly detection technique frequently mis-detects other types of anomaly. In this work, we focus on the anomaly detection in the stage of data pre-processing that little work has been done based on the real-world continuous SHM data with multiclass anomalies. We proposed a novel data anomaly detection method based on a convolutional neural network (CNN) that imitates human vision and decision making. First, we split raw time series data into sections, and visualized the data in time and frequency domain, respectively. Then each section's images were stacked as a single dual-channel image and labeled according to graphical features (multi-2D image space expression). Second, a CNN was designed and trained for data anomaly classification, during which the descriptions and representations of the anomalies' features were learned by convolution. To validate our work, we considered the effects of balanced and imbalanced training sets and training ratios on actual acceleration data from an SHM system for a long-span cable-stayed bridge. The results show that our method could detect the multipattern anomalies of SHM data efficiently with high accuracy. The proposed dual-information CNN-based design makes this detection process readily scalable, faster, and more accurate, thereby providing a novel perspective with strong potential for SHM data preprocessing. KEYWORDScomputer vision, convolutional neural network (CNN), data anomaly detection, long-span bridge, structural health monitoring (SHM) | INTRODUCTIONStructural health monitoring (SHM) is used worldwide to manage and maintain civil infrastructure systems by evaluating their structural loads, responses, and real-time performance, and by predicting the future behavior of structures of all types. 1-4 SHM has produced huge amounts of data. For example, the SHM system used for the Sutong Bridge in China, which has 785 sensors, produces 2.5 TB of data each year. The effective mining and utilization of SHM data is an important topic that needs in-depth study. However, researchers have faced challenges because the SHM data generally include multiple types of anomalies caused by sensor fa...

Deep learning for data anomaly detection and data compression of a long‐span suspension bridge

“…Compared with the traditional existing image‐based methods, a sophisticated DL model trained on a large‐scale dataset can overcome the image complexity and become a generalized tool that performs well on arbitrary field data. In the field of SHM data anomaly detection, Tang, Chen, Bao, and Li () proposed computer vision and deep learning‐based methods that can detect various types of anomaly with high accuracy.…”

Data-Driven Structural Health Monitoring and Damage Detection through Deep Learning: State-of-the-Art Review

“…A parallel configuration of CNN ( Figure 3) can be used for a robust damage localization and damage intensity estimation, where the time-domain data are processed using the one-dimensional (1D) CNN (upper branch) and the time-frequency-domain (WPT) or vision data are processed using 2D CNN (lower branch), and the feature maps are concatenated in the end for classification or regression. Details of the CNN configuration in Figure 3 are available from Azimi and Pekcan [57].…”

“…The location of sensors with the highest novelty rates was considered to be the approximate location of loose-bolt damage. Wavelet packet transform (WPT) of vibration signals, as well as vision data, are efficient in damage localization, according to the studies that were conducted by Pan et al [20,21] and Pan and Yang [57]. A parallel configuration of CNN ( Figure. 3) can be used for a robust damage localization and damage intensity estimation, where the time-domain data are processed using the one-dimensional (1D) CNN (upper branch) and the time-frequency-domain (WPT) or vision data are processed using 2D CNN (lower branch), and the feature maps are concatenated in the end for classification or regression.…”

Hierarchical outlier detection approach for online distributed structural identification

“…Outliers will be detected if they do not belong to any cluster or if the cluster size is significantly smaller than the others. Classification‐based approaches 45–48 use artificial intelligence algorithms to train a classification model and new data points can be classified using this trained model. The performance of most of the aforementioned methods can be guaranteed only when an appropriate threshold for the outlier criterion is selected.…”