Privacy protection as a major concern of the industrial big data enabling entities makes massive safety-critical operation wind turbine unable to exert its great value because threat privacy leakage. How improve diagnostic accuracy decentralized machines without transfer remains an open issue; especially these are almost accompanied by skewed class distribution in real industries. In this study, class-imbalanced privacy-preserving federated learning framework for fault diagnosis is proposed....

In this paper, a robust fault estimation approach is proposed for multi-input and multioutput nonlinear dynamic systems on the basis of back propagation neural networks. The augmented system approach, input-to-state stability theory, linear matrix inequality optimization, network training/learning are integrated so that simultaneous estimate states actuator faults achieved. approaches finally applied to 4.8 MW wind turbine benchmark system, effectiveness well demonstrated.

In response to the high demand of operation reliability and predictive maintenance, health monitoring fault diagnosis classification have been paramount for complex industrial systems (e.g., wind turbine energy systems). this study, data-driven strategies are addressed under various faulty scenarios. A novel algorithm is by integrating fast Fourier transform uncorrelated multi-linear principal component analysis techniques in order achieve effective three-dimensional space visualization a...

Diagnosing potential faults is of great importance to ensure reliability battery management systems. This because a current or voltage sensor fault often results in an inaccurate state-of-charge estimate. A temperature will cause abnormal thermal management. internal resistance (BIR) can lead increase energy and power losses, capacity fading, further degradation health. In addition, frequent data transmission diagnosis unit waste communication resources. To this end, combined model-based...

Brain-Computer Interfaces (BCI) use electroencephalography (EEG) signals recorded from the scalp to create a new communication channel between brain and an output device by bypassing conventional motor pathways of nerves muscles. One most important components BCI is feature extraction EEG signals. How rapidly reliably extract features for expressing states different mental tasks crucial element exact classification. This paper presents approach that performs during imagined right left hand...

Rolling bearings are the key components of rotating machinery. Incipient fault diagnosis bearing plays an increasingly important role in guaranteeing normal and safe operation However, because high complexity feature extraction, incipient faults rolling difficult to diagnose. To solve this problem, paper presents a new intelligent identification method based on variational mode decomposition (VMD), principal component analysis (PCA), support vector machines (SVM). In proposed method,...

Controlled flight into terrain (CFIT) can result in significant aircraft damage and human casualties. Analyzing incident factors their evolutionary relationships aviation safety reports helps explore the inherent mechanisms of CFIT, thereby potentially reducing occurrence. This study proposes a methodology combining named entity recognition (NER) Bayesian network (BN) to address challenges efficiently extracting from textual crew’s perspective analyzing overall evolution process CFIT...

In fault-diagnosis classification, a pressing issue is the lack of target-fault samples. Obtaining fault data requires great amount time, energy and financial resources. These factors affect accuracy diagnosis. To address this problem, novel fault-diagnosis-classification optimization method, namely TLSCA-SVM, which combines sine cosine algorithm support vector machine (SCA-SVM) with transfer learning, proposed here. Considering availability data, thesis uses generated by analog circuits...

Rapid developments of offshore wind industry offer a strong demand opportunity for turbine remote diagnosis. As turbines are often located in harsh and communication-constrained environments, the collection transmission data is severely restricted, which poses serious challenge to conventional centralized diagnostic paradigm that relies on aggregation. To address this challenge, we propose novel event-triggered federated learning framework decentralized fault diagnosis turbines....

A finite-time attitude compensation control scheme is developed for an over-activated rigid spacecraft subject to actuator faults, misalignment, external disturbances and uncertain inertia parameters. The controller synthesised based on the sliding mode theory, guarantees reachability of system states. sufficient condition accommodate misalignment faults presented. An optimised allocation algorithm Karush–Kuhn–Tucker then applied distribute commanded each actuator, optimise consumption...

Unexpected faults in actuators and sensors may degrade the reliability safety of aero engineering systems. Therefore, there is motivation to develop integrated fault tolerant control techniques with applications In this paper, discrete-time dynamic systems, presence simultaneous actuator/sensor faults, partially decoupled unknown input disturbances, sensor noises, are investigated. A jointly state/fault estimator formulated by integrating an observer, augmented system approach, optimization...

Clustering is able to find out implicit data distribution and especially useful in driven machine learning. Density based clustering has an attractive property of detecting clusters arbitrary structures. The density peak algorithm makes use two assumptions detect cluster centers then groups the other data. This approach simple implement shown be promising many experiments. However, we its results are dependent on kernel types parameters, difference across also influences significantly. In...

A rigid satellite fault diagnosis strategy, subject to faults of external disturbances and thruster faults, is developed. In this design, an equivalent idea introduced design a sliding mode observer that can detect identify the failures indicated previously. Considering measurability parameters satellite, such as angular velocity attitude, implemented. Next, amplitudes be detected, identified, estimated via observer; these tasks are accomplished with zero estimation error within finite...

The mechanical–electrical coupling properties of piezoelectric semiconductors endow these materials with novel device applications in microelectromechanical systems, sensors, human–computer interfaces, etc. When an applied strain is exerted on a semiconductor, charges are generated at the surface or interface which can be utilized to control electronic transport characteristics. This fundamental working mechanism piezotronic devices, called effect. In present report, series transistors...