A5026885241- Oil and Gas Production Techniques
- Fault Detection and Control Systems
- Machine Fault Diagnosis Techniques
- Nuclear Engineering Thermal-Hydraulics
- Water Systems and Optimization
- Hydraulic and Pneumatic Systems
- Anomaly Detection Techniques and Applications
- Nuclear reactor physics and engineering
- Heat transfer and supercritical fluids
- Engineering Diagnostics and Reliability
- Heat Transfer and Boiling Studies
- Solar Thermal and Photovoltaic Systems
- Advanced Algorithms and Applications
- Adsorption and Cooling Systems
- Navier-Stokes equation solutions
- Non-Destructive Testing Techniques
- Industrial Technology and Control Systems
- Air Quality Monitoring and Forecasting
- Advanced Data Processing Techniques
- Quality and Safety in Healthcare
- Structural Health Monitoring Techniques
- Drilling and Well Engineering
- Aquatic and Environmental Studies
- Cyclone Separators and Fluid Dynamics
- Heat Transfer and Optimization
China General Nuclear Power Corporation (China)
2022-2025
Domain adaptation can transfer cross-domain diagnosis knowledge by minimizing the divergence of labeled source and unlabeled target data. However, model neglects to maximize physics prior during feature extraction distribution alignment, resulting in a noninterpretable even negative transfer. Hence, physics-informed domain network, termed adaptive fault attention residual network (AFARN), is proposed. First, an mechanism designed refine features guided bearing characteristics, suited...
Journal bearings are the key components of nuclear circulating water pump (NCWP), and accurate remaining useful life (RUL) prediction is great significance for improving reliability, safety, maintenance planning NCWP. However, it difficult to quantify uncertainty bearing RUL based on current deep learning (DL) model, resulting in a lack credibility effective convincing predicted by model. Meanwhile, all existing hybrid models basically simple combinations, they cannot solve quantification...
Due to coupled nonlinearities and complex measurement noise, assess the condition of rotor system remains a challenge, particularly in cases where historical run-to-failure data is lacking. To this end, we proposed hierarchical physics-informed neural network (HPINN) identify/discover ordinary differential equations (ODEs) healthy/faulty from noise measurements then based on discovered ODEs. Specifically, ODEs healthy are first stably identified noisy through HPINN guided by dynamics. Based...
As a critical component of nuclear power units, the direct cooling water system plays key role in overall performance. To maintain economic efficiency, it is necessary to adjust circulating flow rate as conditions change. Understanding how this responds dynamically varying environmental factors—such seawater temperature and tidal levels—is essential for precise control. While previous studies have explored methods such variable frequency control, predictive maintenance, digital twin...
Existing domain adaptation methods strive to align all domains equally under a single shift dimension, which poses two problems. On the one hand, multiaspect transferring factors and homogenous alignment may lead suboptimal results in more distant domains. other such global ignores local discriminatory information, making class boundary samples susceptible misclassification. Hence, three-types-of-graph-relational guided (TGGDA) is proposed. First, <italic...
Deep learning (DL) has demonstrated splendid performance in fault diagnosis with sufficient samples and ideal operating environments. However, practice, it is hard to acquire adequate from high-reliability equipment, thus inducing challenges applying DL for intelligent diagnosis. Moreover, data distribution shift occurs due working load variation or environment noise interference, severely degrading models' performance. Aiming at the above problems, this article proposes a method called Q...
To diagnose common failures in vertical Essential Service Water Pumps (SEC), a method combining the wavelet packet transform (WPT) and support vector machine (SVM) was adopted. This allowed us to construct diagnostic model capable of classifying multiple states, including six types faults normal conditions SEC pumps. The utilized coefficients capture sub-bands with higher energy share reconstruct signals. inputs 12 frequency features into analyze vibration signals gathered from pump...
Essential service water pumps are necessary safety devices responsible for discharging waste heat from containments through seawater; their condition monitoring is critical the safe and stable operation of seaside nuclear power plants. However, it difficult to directly apply existing intelligent methods these pumps. Therefore, an framework designed, including parallel implementation unsupervised anomaly detection fault diagnosis. A model preselection algorithm based on highest validation...
This paper proposes a data-driven prediction scheme for the remaining life of centrifugal pump bearings based on KPCA–LSTM network. A bearing fault experiment bench is built to collect data, and performance time domain, frequency time-frequency domain characteristics under different working conditions analyzed. Time characteristics, wavelet packet decomposition energy CEEMDAN features are found be able capture information conditions. Therefore, 43 sensitive determined from domain. Through...
We introduce a new hyperbolic approximation to the incompressible Navier-Stokes equations by incorporating first-order relaxation and using artificial compressibility method. With two parameters in model, we rigorously prove asymptotic limit of system towards as both tend zero. Notably, convergence approximate pressure variable is achieved help linear `auxiliary' energy-type error estimates its differences with two-parameter model equations.
Natural draft hybrid cooling (NDHC) for thermal power generating units is proposed to achieve a balance of energy and water consumption arid areas. This study examines the two main design forms with airside in serial parallel heat exchange based on same tower shell transfer Taking full consideration cycle unit, simplified simulation models different systems are established show influences ambient conditions marketing factors. Results that both designs have better efficiency than either dry...
In small modular reactors,the extremely short distance between the SG and RCP results in a non-uniform inflow at pump inlet.To investigate effect of on SMRCPs,the internal vortex structure SMRCP was investigated.Firstly,the uniform models were constructed for comparison,and CFD method used to obtain hydraulic characteristic curve flow field information.The show that efficiency head model is both higher than model.Secondly,by analyzing velocity streamline distribution under steady calculation...