In the contemporary big data era, data-driven prognostic and health management (PHM) methodologies have emerged as indispensable tools for ensuring the secure and reliable operation of complex equipment systems. Central to these methodologies is the accurate prediction of remaining useful life (RUL), which serves as a pivotal cornerstone for effective maintenance and operational decision-making. While significant advancements in computer hardware and artificial intelligence (AI) algorithms have catalyzed substantial progress in AI-based RUL prediction, extant research frequently exhibits a narrow focus on specific algorithms, neglecting a comprehensive and comparative analysis of AI techniques across diverse equipment types and operational scenarios. This study endeavors to bridge this gap through the following contributions: (1) A rigorous analysis and systematic categorization of application scenarios for equipment RUL prediction, elucidating their distinct characteristics and requirements. (2) A comprehensive summary and comparative evaluation of several AI algorithms deemed suitable for RUL prediction, delineating their respective strengths and limitations. (3) An in-depth comparative analysis of the applicability of AI algorithms across varying application contexts, informed by a nuanced understanding of different application scenarios and AI algorithm research. (4) An insightful discussion on the current challenges confronting AI-based RUL prediction technology, coupled with a forward-looking examination of its future prospects. By furnishing a meticulous and holistic understanding of the traits of various AI algorithms and their contextual applicability, this study aspires to facilitate the attainment of optimal application outcomes in the realm of equipment RUL prediction.

Traditional reliability allocation methods face three inherent limitations: (1) the insufficient consideration of influencing factors, (2) the equal treatment of impact factors, and (3) the presence of fuzzy information. To overcome these issues, we propose an integrated reliability allocation method that combines a two-layer integrated factors framework with a cloud model. First, a two-layer framework of influencing factors was constructed, and a mathematical model for global evaluation based on these factors was developed, along with an indicator allocation model derived from the evaluation model. The cloud model was employed to quantitatively express experts’ linguistic evaluations, eliminating the influence of randomness and fuzziness in the assessment process. Both theoretical analysis and practical examples show that the proposed method outperforms three other reliability allocation methods.

Traditional in-service inspection (ISI) methods for pipelines have certain limitations in identifying pipeline leakages. When these methods are directly applied to the ISI of Hua-long pressurized reactor (HPR1000) nuclear power plants, where the system complexity has significantly increased, they may lead to insufficient inspection efficiency and an extremely heavy workload. In this study, based on the framework of typical risk-informed analysis methods for nuclear power plants in the industry and integrating domestic engineering practical experience, an optimized ISI model for pipelines in HPR1000 nuclear power plants was constructed, and a pilot application was carried out on the chemical and volume control system (RCV) of the primary circuit. The inspection strategy was optimized through a series of steps, including determining the analysis scope, conducting pipe segment failure analysis, constructing a risk matrix, selecting inspection elements, and assessing risk impacts. Case studies showed that the risk-informed in-service inspection (RI-ISI) method successfully classified over 3000 welds in the RCV system based on risk levels (high, medium, low). After optimization, 16 low-risk welds (risk level 7) and one of the two medium-risk welds (risk level 4) that originally required volumetric inspection were exempted from inspection. Quantitative risk analysis confirmed that the increments in core damage frequency (CDF) and large early-release frequency (LERF) caused by this optimization were far below the regulatory limits. This method significantly reduces the inspection burden of medium- and low-risk pipelines while ensuring that high-risk areas receive priority attention, providing important technical support for the safe and efficient operation and maintenance of HPR1000 and subsequent third-generation nuclear power units.

In the context of intelligent manufacturing, deep learning-based remaining useful life (RUL) prediction has become a research hotspot in the field of Prognostics and Health Management (PHM). The traditional approaches often require strong programming skills and repeated model building, posing a high entry barrier. To address this, in this study, we propose and implement a visualization tool that supports multiple model selections and result visualization and eliminates the need for complex coding and mathematical derivations, helping users to efficiently conduct RUL prediction with lower technical requirements. This study introduces and summarizes various novel neural network models for DL-based RUL prediction. The models are validated using the NASA and HNEI datasets, and among the validated models, the LSTM model best met the requirements for remaining useful life (RUL) prediction. In order to achieve the low-code usage of deep learning for RUL prediction, the following tasks were performed: (1) multiple models were developed using the Python (3.9.18) language and were implemented on the PyTorch (1.12.1) framework, providing users with the freedom to choose their desired model; (2) a user-friendly and low-code RUL prediction interface was built using Streamlit, enabling users to easily make predictions; (3) the visualization of prediction results was implemented using Matplotlib (3.8.2), allowing users to better understand and analyze the results. In addition, the tool offers functionalities such as automatic hyperparameter tuning to optimize the performance of the prediction model and reduce the complexity of operations.