In this paper, the procedure for the strength evaluation of the piping system installed on liquefied natural gas (LNG) carriers is discussed. A procedure that accounts for the ship’s wave load and hull motion acceleration (as well as the deformation due to the thermal expansion and contraction experienced by the hull during seafaring operations) is presented. The load due to the temperature and self-weight of the piping installed on the deck is also considered. Various operating and load conditions of the LNG piping system are analyzed. Stress analysis is performed by combining various conditions of sustained, occasional, and expansion loads. Stress is assessed using finite element analysis based on beam elements that represent the behavior of the piping. The attributes of the piping system components (such as valves, expansion joints, and supports) are represented in the finite element model while CAESAR-II, a commercial software is used for finite element analysis. Component modeling, load assignment, and load combinations are presented to evaluate pipe stresses under various load conditions. An evaluation model is selected for the piping arrangement of LNG and the evaluated stress is compared with the allowable stress defined by the American Society of Mechanical Engineers (ASME).

The influence of environmental noise is generally excluded during research on machine fault diagnosis using acoustic signals. This study proposes a fault diagnosis method using a variational autoencoder (VAE) and domain adaptation neural network (DANN), both of which are based on unsupervised learning, to address this problem. The proposed method minimizes the impact of environmental noise and maintains the fault diagnosis performance in altered environments. The fault diagnosis algorithm was implemented using acoustic signals containing noise, present in the malfunctioning industrial machine investigation and inspection open dataset, and the fault prediction performance in noisy environments was examined based on forklift acoustic data using the VAE and DANN. The VAE primarily learns from normal state acoustic data and determines the occurrence of faults based on reconstruction error. To achieve this, statistical features of Mel frequency cepstral coefficients were extracted, generating features applicable regardless of signal length. Additionally, features were enhanced by applying noise reduction techniques via magnitude spectral subtraction and feature optimization, reflecting the characteristics of rotating equipment. Furthermore, data were augmented using generative adversarial networks to prevent overfitting. Given that the forklift acoustic data possess time-series characteristics, the exponentially weighted moving average was determined to quantitatively track time-series changes and identify early signs of faults. The VAE defined the reconstruction error as the fault index, diagnosing the fault states and demonstrating excellent performance using time-series data. However, the fault diagnosis performance of the VAE tended to decrease in noisy environments. Moreover, applying DANN for fault diagnosis significantly improved diagnostic performance in noisy environments by overcoming environmental differences between the source and target domains. In particular, by adapting the model learned in the source domain to the target domain and considering the domain differences based on signal-to-noise ratio, high diagnostic accuracy was maintained regardless of the noise levels. The DANN evaluated interdomain similarity using cosine similarity, enabling the accurate classification of fault states in the target domain. Ultimately, the combination of the VAE and DANN techniques enabled effective fault diagnosis even in noisy environments.

In order to utilize wave energy, various wave power systems are being actively researched and developed and interest in them is increasing. To maximize the operational efficiency, it is very important to monitor and maintain the fault of components of the system. In recent years, interest in the management cost, high reliability and facility utilization of such systems has increased. In this regard, fault diagnosis technology including fault factor analysis and fault reproduction is drawing attention as an important main technology. Therefore, in this study, to reproduce and monitor the faults of a wave power system, firstly, the failure mode of the system was analyzed using FMEA analysis. Secondly, according to the derived failure mode and effect, the thrust bearing was selected as a target for fault reproduction and a test equipment bench was constructed. Finally, with the vibration data obtained by conducting the tests, the vibration spectrum was analyzed to extract the features of the data for each operating status; the data was classified by applying the three machine learning algorithms: naïve Bayes (NB), k-nearest neighbor (k-NN), and multi-layer perceptron (MLP). The criteria for determining the fault were derived. It is estimated that a more efficient fault diagnosis is possible by using the standard and fault monitoring method of this study.