A5069590205- Risk and Safety Analysis
- Fault Detection and Control Systems
- Engineering Diagnostics and Reliability
- Oil and Gas Production Techniques
- Machine Fault Diagnosis Techniques
- Reliability and Maintenance Optimization
- Advanced Machining and Optimization Techniques
- Advanced machining processes and optimization
- Occupational Health and Safety Research
- Advanced Surface Polishing Techniques
- Offshore Engineering and Technologies
- Structural Integrity and Reliability Analysis
- Software Reliability and Analysis Research
- Water Systems and Optimization
- Infrastructure Resilience and Vulnerability Analysis
- Anomaly Detection Techniques and Applications
- Electrodeposition and Electroless Coatings
- Electrohydrodynamics and Fluid Dynamics
- Drilling and Well Engineering
- Metal and Thin Film Mechanics
- Corrosion Behavior and Inhibition
- Erosion and Abrasive Machining
- Oil Spill Detection and Mitigation
- Hydrogen embrittlement and corrosion behaviors in metals
- Power System Reliability and Maintenance
China University of Petroleum, East China
2016-2025
National Marine Environmental Forecasting Center
2024
China University of Petroleum, Beijing
2010-2021
Southwest Jiaotong University
2019
City University of Hong Kong
2015-2018
ORCID
2018
Norwegian University of Science and Technology
2016
Permanent magnet synchronous motor and power electronics-based three-phase inverter are the major components in modern industrial electric drive system, such as electrical actuators an all-electric subsea Christmas tree. Inverters weakest switches most vulnerable inverters. Fault detection diagnosis of inverters extremely necessary for improving system reliability. Motivated by solving uncertainty problem fault inverters, which is caused various reasons, bias noise sensors, this paper...
Fault diagnosis is useful in helping technicians detect, isolate, and identify faults, troubleshoot. Bayesian network (BN) a probabilistic graphical model that effectively deals with various uncertainty problems. This increasingly utilized fault diagnosis. paper presents bibliographical review on use of BNs the last decades focus engineering systems. work also general procedure modeling BNs; processes include BN structure modeling, parameter inference, identification, validation,...
Transient fault (TF) and intermittent (IF) of complex electronic systems are difficult to diagnose. As the performance products degrades over time, results diagnosis could be different at times for given identical symptoms. A dynamic Bayesian network (DBN)-based methodology in presence TF IF is proposed. DBNs used model degradation process products, Markov chains transition relationships four states, i.e., no fault, TF, IF, permanent fault. Our can identify faulty components distinguish...
The Bayesian network (BN) is a powerful model for probabilistic knowledge representation and inference increasingly used in the field of reliability evaluation. This paper presents bibliographic review BNs that have been proposed evaluation last decades. Studies are classified from perspective objects evaluation, i.e., hardware, structures, software, humans. For each classification, construction validation BN-based emphasized. general procedural steps including BN structure modeling,...
In dynamic complex environments, the degradation of structure systems is generally caused not by a single factor but multiple ones, and process subject to high level uncertainty. This article contributes hybrid physics-model-based data-driven remaining useful life (RUL) estimation methodology considering influence causes using Bayesian networks (DBNs). The model parameter DBNs for are established on basis theoretical or empirical physical models, thereby solving problem insufficient data. An...
End-to-end intelligent diagnosis of rotating machinery under speed fluctuation and limited samples is challenging in industrial practice. The existing methods usually focus on the data distribution or learning strategy with particularity. Generative adversarial network (GAN) provides a generation solution portability fault samples. However, GAN has problems gradient vanishing, weak extraction global features, redundant training. This article proposes dual-threshold attention-guided (DTAGAN)...
The existing fault diagnosis methods of rotating machinery constructed with both shallow learning and deep models are mostly based on vibration analysis under steady speed. However, the speed frequently changes to meet practical engineering needs. largely depend domain experience feature extraction, training a model requires large samples long time. In addition, monitoring has shortcomings contact measurement, small coverage, noise interference. To address these problems, this article...
The application of convolutional neural network (CNN) has greatly promoted the scope and scenario intelligent fault diagnosis brought about a significant improvement model performance. Solving feature extraction machinery with heavy noise is beneficial for stable industrial production. However, local properties CNN prevent it from obtaining global features to collect sufficient information, leading degradation performance under noise. In this article, novel framework named Convformer-NSE...
This article proposes a methodology for the application of Bayesian networks in conducting quantitative risk assessment operations offshore oil and gas industry. The method involves translating flow chart into network directly. proposed consists five steps. First, is translated network. Second, influencing factors nodes are classified. Third, each factor established. Fourth, entire model Lastly, analyzed. Subsequently, categories factors, namely, human, hardware, software, mechanical,...
Vibration signals and infrared images have different advantages characteristics. Although a few recent researches explored their information fusion in rotating machinery fault diagnosis, they show limited performance when facing strong interference imbalanced cases. Therefore, framework based on confidence weight support matrix machine (CWSMM) is proposed. In this framework, CWSMM can not only fully leverage the structure of thermography vibration time–frequency images, but also has...