Unsupervised learning framework for temperature compensated damage identification and localization in ultrasonic guided wave SHM with transfer learning

Optimal principal component and measurement interval selection for PCA reconstruction-based anomaly detection in uncontrolled structural health monitoring

PCA reconstruction-based techniques are widely used in guided wave structural health monitoring to facilitate unsupervised damage detection. The measurement interval of collecting evaluation data significantly influences the correlation among the data points, impacting principal component values and, consequently, the accuracy of damage detection. Despite its importance, there has been limited research on the selection of suitable components and measurement intervals to reduce false alarms. This paper seeks to develop strategies for identifying the optimal number of principal components and measurement intervals for PCA reconstruction-based damage detection methods. Our results indicate that the patterns of change in reconstruction coefficients, based on the number of components used in PCA reconstruction and the measurement interval for collecting evaluation data, are effective indicators for determining the optimal principal components and measurement intervals for damage detection, without using any damage information. The effectiveness of the indicators for determining optimal components and measurement intervals is validated using evaluation sets collected under uncontrolled and dynamic monitoring conditions, with measurement intervals ranging from 86 to 8600 s per measurement.

Domain-separated capsule network for damage detection in aluminum plates under varying vibration conditions

The 2024 aluminum alloy, known for its high strength and resistance to fatigue, is widely used in critical parts of aircraft such as wings and fuselages. Techniques that use ultrasonic guided waves for structural health monitoring are commonly applied to detect damage in metal plates. However, changes in environmental vibrations can alter the signals collected, greatly affecting the accuracy of damage identification in aluminum alloy plates. To tackle this challenge, a domain-separated capsule network (DS-CapsNet) has been developed to reduce the impact of environmental vibrations on the accuracy of damage detection. DS-CapsNet integrates a Capsule Network with an attention mechanism to extract and reconstruct damage-related features while minimizing vibration-induced interference. Additionally, a dynamic adversarial factor is introduced to optimize feature alignment between different domains, enhancing the robustness of the model. Moreover, a multi-head self-attention mechanism improves classification performance by effectively capturing complex damage features. Experimental results demonstrate that the proposed DS-CapsNet consistently outperforms a broad range of baseline models, including traditional classifiers, deep learning networks, and domain adaptation approaches, confirming its robustness under varying vibration conditions.

Guided wave multi-frequency damage localization method in variable-thickness structures by one pair of sensors based on frequency-dependent velocity anisotropy

Variable thickness structures are prevalent in aircraft, ships, and other machines, necessitating numerous sensors for health monitoring to reduce safety hazards. This paper presents a guided wave multi-frequency localization method based on frequency-dependent velocity anisotropy. This method achieves damage localization in variable-thickness structures with a pair of sensors and can effectively reduce the number of sensors used for monitoring. Variations in structural thickness cause a gradient in guided wave velocity that bends the propagation path. Different thickness variations with different directions cause wave velocity anisotropy. As a result, variations in thickness cause possible damage loci determined by echo time to deviate from an elliptical shape. Because the velocity anisotropy is frequency-dependent, damage loci at different frequencies are close but do not overlap and intersect only at the damage location. So, the multi-frequency method can increase the damage information acquired by a single pair of sensors, enabling damage localization. Experimental validation was conducted on a steel plate with linearly varying thicknesses. The feasibility of the multi-frequency localization method was verified by successfully locating the damage at three different locations using a pair of receiver-excitation sensors. In addition, the experiments demonstrated the capability of this multi-frequency method in improving the localization accuracy of sensor networks. The method has potential applications in monitoring systems lightweight, phased arrays, and imaging enhancement.

Mechanics-informed autoencoder enables automated detection and localization of unforeseen structural damage

Structural health monitoring ensures the safety and longevity of structures like buildings and bridges. As the volume and scale of structures and the impact of their failure continue to grow, there is a dire need for SHM techniques that are scalable, inexpensive, can operate passively without human intervention, and are customized for each mechanical structure without the need for complex baseline models. We present a novel "deploy-and-forget" approach for automated detection and localization of damage in structures. It is a synergistic integration of entirely passive measurements from inexpensive sensors, data compression, and a mechanics-informed autoencoder. Once deployed, the model continuously learns and adapts a bespoke baseline model for each structure, learning from its undamaged state's response characteristics. After learning from just 3 hours of data, it can autonomously detect and localize different types of unforeseen damage. Results from numerical simulations and experiments indicate that incorporating the mechanical characteristics into the autoencoder allows for up to a 35% improvement in the detection and localization of minor damage over a standard autoencoder. Our approach holds significant promise for reducing human intervention and inspection costs while enabling proactive and preventive maintenance strategies. This will extend the lifespan, reliability, and sustainability of civil infrastructures.

Condition Monitoring Using a Latent Space of Variational Autoencoder Trained Only on a Healthy Machine

Machine learning generative models have opened up a new perspective for automated machine diagnostics. These methods improve decision-making by extracting features, classifying, and creating new observations using deep neural networks. Generative modeling aims to determine the joint distribution of input data. This contrasts traditional methods used in diagnostics based on discriminative models and the conditional probability distribution of the target variable at known feature values. In the variational autoencoder (VAE) algorithms trained by the authors, the parameters of diagnostic features are random variables, the distributions of which can be approximated based on data, and the identification of probability distributions is based on variational inference. Variational inference is a tool that deals with difficult statistical problems and is usually faster than classical methods. VAEs can detect anomalies, predict failures, and optimize processes. This paper proposes an unsupervised approach to fault diagnosis using only healthy data with automatic feature extraction from the continuous probabilistic latent subspace of the VAE encoder and reduction in PCA or t-SNE. The solution, verified in the example of simulation data, is a response to a common problem related to the lack or difficulty of obtaining marked data in defected states of devices and mechanical structures.

Large-area inspection of defects in metal plates using multi-mode guided acoustic waves and sparse sensor networks

Various types of defects can be induced during the manufacturing or operation of engineering structures. For effective detection and characterization of the defects in large engineering structures, this paper proposes a large-area inspection technique that combines multi-mode guided acoustic waves with sparse sensor networks. The basic sparse sensor network employed in this study is composed of one transmitter and three receivers, distributed in a square lattice on the test plates. Multi-mode guided waves were excited and acquired by means of commercial single-element sensors of the network. To experimentally demonstrate the proposed technique, four different types of defects were simulated in aluminum test plates, including aluminum tape-based material addition, drilled material loss, indented deformation, and thermal embrittlement. For the evaluation of defects, acoustic response of each defect was analyzed based on the combination of linear vs. nonlinear acoustic characteristics, dependence on the type of the guided acoustic mode, and the directionality of the acoustic response on the network. Results indicate that each of the four representative defects can be uniquely identified (classified) and quantified using the proposed technique.

Unsupervised deep learning framework for temperature-compensated damage assessment using ultrasonic guided waves on edge device

Fueled by the rapid development of machine learning (ML) and greater access to cloud computing and graphics processing units, various deep learning based models have been proposed for improving performance of ultrasonic guided wave structural health monitoring (GW-SHM) systems, especially to counter complexity and heterogeneity in data due to varying environmental factors (e.g., temperature) and types of damages. Such models typically comprise of millions of trainable parameters, and therefore add to cost of deployment due to requirements of cloud connectivity and processing, thus limiting the scale of deployment of GW-SHM. In this work, we propose an alternative solution that leverages TinyML framework for development of light-weight ML models that could be directly deployed on embedded edge devices. The utility of our solution is illustrated by presenting an unsupervised learning framework for damage detection in honeycomb composite sandwich structure with disbond and delamination type of damages, validated using data generated by finite element simulations and experiments performed at various temperatures in the range 0-90 °C. We demonstrate a fully-integrated solution using a Xilinx Artix-7 FPGA for data acquisition and control, and edge-inference of damage. Despite the limited number of features, the lightweight model shows reasonably high accuracy, thereby enabling detection of small size defects with improved sensitivity on an edge device for online GW-SHM.

Quantitative Detection of Pipeline Cracks Based on Ultrasonic Guided Waves and Convolutional Neural Network

In this study, a quantitative detection method of pipeline cracks based on a one-dimensional convolutional neural network (1D-CNN) was developed using the time-domain signal of ultrasonic guided waves and the crack size of the pipeline as the input and output, respectively. Pipeline ultrasonic guided wave detection signals under different crack defect conditions were obtained via numerical simulations and experiments, and these signals were input as features into a multi-layer perceptron and one-dimensional convolutional neural network (1D-CNN) for training. The results revealed that the 1D-CNN performed better in the quantitative analysis of pipeline crack defects, with an error of less than 2% in the simulated and experimental data, and it could effectively evaluate the size of crack defects from the echo signals under different frequency excitations. Thus, by combining the ultrasonic guided wave detection technology and CNN, a quantitative analysis of pipeline crack defects can be effectively realized.

Deep Learning for Structural Health Monitoring: Data, Algorithms, Applications, Challenges, and Trends

Environmental effects may lead to cracking, stiffness loss, brace damage, and other damages in bridges, frame structures, buildings, etc. Structural Health Monitoring (SHM) technology could prevent catastrophic events by detecting damage early. In recent years, Deep Learning (DL) has developed rapidly and has been applied to SHM to detect, localize, and evaluate diverse damages through efficient feature extraction. This paper analyzes 337 articles through a systematic literature review to investigate the application of DL for SHM in the operation and maintenance phase of facilities from three perspectives: data, DL algorithms, and applications. Firstly, the data types in SHM and the corresponding collection methods are summarized and analyzed. The most common data types are vibration signals and images, accounting for 80% of the literature studied. Secondly, the popular DL algorithm types and application areas are reviewed, of which CNN accounts for 60%. Then, this article carefully analyzes the specific functions of DL application for SHM based on the facility's characteristics. The most scrutinized study focused on cracks, accounting for 30 percent of research papers. Finally, challenges and trends in applying DL for SHM are discussed. Among the trends, the Structural Health Monitoring Digital Twin (SHMDT) model framework is suggested in response to the trend of strong coupling between SHM technology and Digital Twin (DT), which can advance the digitalization, visualization, and intelligent management of SHM.