This paper presents a new framework that aims to improve the efficiency of time response analysis for nonlinear dynamical systems by combining conventional time integration methods with Proper Generalized Decomposition (PGD). The PGD approach utilizes low-dimensional subspaces of the time response to approximate the solution as a low-order separated representation of spatial and temporal components, with the Galerkin projection employed to formulate subproblems for each component. The subproblem for spatial basis is viewed as computing a reduced-order criterion, and the temporal problem projected to a subspace spanning this criterion uses time integration to obtain time coefficients. During the time integration, the spatial modes obtained from the calculation of the previous step are used as a reduced basis, and additional spatial modes are added until the residual of equations of motion satisfy the target tolerance. Numerical examples demonstrate that the proposed method allows significant computational savings compared to conventional time integration methods while accurately reflecting the nonlinear behavior.
Tire maintenance plays a crucial role in vehicle performance, with the tire being identified as the most important factor. In this study, we introduce an intelligent tire system equipped with composite sensors to enhance driving safety and vehicle management. Analysis of actual traffic accident data reveals that approximately 10% of accidents are attributed to tire-related issues, emphasizing the significance of tire maintenance. However, our investigation suggests that while conventional time and frequency domain techniques are available for fault detection in intelligent tires, they tend to exhibit slightly lower performance compared to those utilizing artificial intelligence. To address this limitation, we propose a deep learning-based diagnosis method. By attaching a 3-axis accelerometer sensor to the tire tread and simulating various failure modes, including Belt/ Bead separation, comprehensive data for analysis were collected. We develop a novel approach using multi-scale feature fusion with adaptive weight calculation using 1-D convolution principles, which significantly improves fault detection accuracy. Experimental results demonstrate the effectiveness of our proposed method, achieving a 100% F1 Score in the classification of Tire Separation faults. Visualization using Uniform Manifold Approximation and Projection (UMAP) further confirms distinct clustering for each fault state. Overall, our study offers valuable insights into tire fault diagnosis and management, contributing to enhanced vehicle safety and performance.
In most engineering systems, the acquisition of faulty data is difficult or sometimes not feasible, while normal data are secured. To solve these problems, this paper proposes an fault diagnosis method for electric motor using only normal data with self-labeling based on stacked time-series imaging method. Since only normal data are used for fault diagnosis, a self-labeling method is used to generate a new labeled dataset based on pretext task. To emphasize faulty features from non-stationary faulty data, stacked time-series imaging method is developed. The overall procedure includes the following steps: (1) transformation of a one-dimensional current signal to a two-dimensional image in time-domain, (2) adding sparse features with sparse dictionary learning, (3) stacked images through every window size, and (4) fault classification based on Convolutional Neural Network (CNN) and Mahalanobis distance. Transformation of the time-series signal is based on Recurrence Plots (RP). The proposed RP method develops from sparse dictionary learning that provides the dominant fault feature representations in a robust way. To verify the proposed method, data from real-field manufacturing line is used.
In this paper, a fast and robust infrared remote target detection network is proposed based on deep learning. Furthermore, we construct our own IR image database imitating humans in remote maritime rescue situations using FLIR M232 IR camera. First, IR image is preprocessed with contrast enhancement for data augmentation and to increase Signal-to-Noise Ratio (SNR). Second, multi-scale feature extraction is performed combined with fixed weighted kernels and convolutional neural network layers. Lastly, the feature map is mapped into a likelihood map indicating the potential locations of the targets. Experimental results reveal that the proposed method can detect remote targets even under complex backgrounds surpassing the previous methods by a significant margin of +0.62 in terms of mIOU.
Nowadays, vibration monitoring system (VMS) using machine learning has been increasingly used to predict rotor faults. However, a sufficient amount of fault data is harder to collect practically than the normal data, as a result, the imbalanced training data set can significantly affect the accuracy of the trained classifier. In this paper, we proposed a data augmentation approach that uses physics-based high-fidelity dynamics simulation as an alternative to acquiring practical fault data. The overall procedure includes: (1) The high-fidelity numerical simulation model reflecting the behavior of rotor to obtain the vibration signature of fault data; and, (2) data augmentation with the simulation fault data in conjunction with experimental normal and fault datasets. A fully connected Neural Network (FCNN) is applied to build the classification model that identifies rotor faults focusing on mass unbalance. The rotor system considered in this study consists of rigid discs, shaft, eccentric mass, and bearing housings. The numerical simulation model in this work considers high-fidelity physical behaviors such flexible multibody dynamics having centrifugal force and gyroscopic effects. Time domain data of the vertical and horizontal vibration responses of bearing housings are obtained from simulation and then FFT is applied to extract the main feature in frequency domain, which is the amplitude of the 1X harmonics of the vibration responses. The data augmentation is accomplished with frequency domain data both from simulation results and experimental acquisition. This approach can tackle data imbalance problem which is one of the most critical hurdles in neural net-based. fault diagnosis. From experimental verification, high accuracies more than 90 % of rotor fault diagnosis, which demonstrates the effectiveness of the proposed framework compared to the model with insufficient fault data.
Time-series signal collected from rotating machinery is subjected to different environmental and operational conditions. The vibration signal is sensitively affected by external noises and load conditions. To solve these problems, this paper presents a diagnostic method for rotating machinery using the proposed robust time-series imaging method. The overall procedure includes the following three key steps: (1) transformation of a one-dimensional current signal to a twodimensional image in time-domain, (2) extracting features using convolutional neural networks, and (3) calculating a health indicator using Mahalanobis distance. Transformation of the time-series signal is based on recurrence plots (RP). The original RP method provides a binary image that makes it insensitive to detecting faulty signal. The proposed RP method develops from sparse dictionary learning that provides the dominant fault feature representations in a robust way. The proposed RP method can detect the weak difference between normal and fault signal, while enhancing robustness to external noise. The dataset acquired from KAIST rotor testbed is used to examine the proposed method’s capability to monitor the condition of rotating machinery. The results show that the proposed method outperforms vibration signalbased condition monitoring methods.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.