Abstract:The development of structural health monitoring (SHM) on civil infrastructures has resulted in enormous amount of acquired data along with the pressure of data processing and data mining. Abnormalities in data can lead to serious analytical error in later assessment. Such anomalous data patterns generally account for a relatively small portion of the overall dataset, which can be easily misclassified as normal data by regular classifiers. In this paper, a novel and robust data anomaly detection framework was proposed. The core novelty in this framework is the utilization of learning active learning (LAL) and AdaBoost algorithm aiming to reduce the costly manual work of labeling and improve the classification of anomaly patterns. Furthermore, the problem of biased classification brought by imbalanced datasets has also been solved by the LAL. Wavelet packet transform was also utilized to extract features from the acceleration data. The methodologies were firstly introduced precisely in this paper followed by two study cases to verify the feasibility of the proposed framework for data anomaly detection. The first case was a dataset with the anomalies synthetically added to the acceleration time history data measured in dynamic tests of a grid structure, including five kinds of data abnormalities. Both the balanced and imbalanced datasets were studied and analyzed, where a comparative study was carried out between the LAL-AdaBoost and uncertainty sampling-based AdaBoost with the same training and testing sets. The results showed that LAL-AdaBoost outperformed in both scenarios with higher accuracies and faster convergence speed. Then, a further study was carried out using acceleration data collected from a long-span bridge. By querying only limited amount of the training set, the proposed framework could accurately detect and classify 97.95% anomaly patterns of the testing set, showing great potential for further and broader application in the field of SHM data processing.

On statistic alignment for domain adaptation in structural health monitoring

Attention-based Cross-Layer Domain Alignment for Unsupervised Domain Adaptation

HOMDA: High-Order Moment-Based Domain Alignment for Unsupervised Domain Adaptation

Trust-aware Conditional Adversarial Domain Adaptation with Feature Norm Alignment.

Domain-adapted Gaussian mixture models for population-based structural health monitoring

A fuzzy-set-based joint distribution adaptation method for regression and its application to online damage quantification for structural digital twin

On risk-based active learning for structural health monitoring

Unsupervised Domain Adaptation Approach for Vision-Based Semantic Understanding of Bridge Inspection Scenes Without Manual Annotations

Discriminative subspace embedded dynamic geometrical and statistical alignment based on pseudo-label correction for cross-domain bearing fault diagnosis

Distribution Sub-Domain Adaptation Deep Transfer Learning Method for Bridge Structure Damage Diagnosis Using Unlabeled Data

Deep Joint Distribution Alignment: A Novel Enhanced-Domain Adaptation Mechanism for Fault Transfer Diagnosis

Multi-level joint distributed alignment-based domain adaptation for cross-scenario strip defect recognition

Missing-Class-Robust Domain Adaptation by Unilateral Alignment for Fault Diagnosis

Aligning Discriminative And Representative Features: An Unsupervised Domain Adaptation Method For Building Damage Assessment

A novel and robust data anomaly detection framework using LAL-AdaBoost for structural health monitoring

A novel joint distinct subspace learning and dynamic distribution adaptation method for fault transfer diagnosis

Joint Discriminative Adversarial Domain Adaptation for Cross-Domain Fault Diagnosis

A statistical distribution recalibration method of soft labels to improve domain adaptation for cross-location and cross-machine fault diagnosis

Adaptive graph-guided joint soft clustering and distribution alignment for cross-load and cross-device rotating machinery fault transfer diagnosis

Intelligent fault diagnosis of double-aligned domain adaptation based on multi-structure fusion and multi-attention mechanism

Deep joint subdomain alignment for unsupervised domain adaptation