اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدOn wind Turbine failure detection from measurements of phase currents: a permutation entropy approach
100
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
This article presents the applicability of Permutation Entropy based complexity measure of a time series for detection of fault in wind turbines. A set of electrical data from one faulty and one healthy wind turbine were analysed using traditional FastFourier analysis in addition to Permutation Entropy analysis to compare the complexity index of phase currents of the two turbines over time. The 4 seconds length data set did not reveal any low frequency in the spectra of currents, neither did they show any meaningful differences of spectrum between the two turbine currents. Permutation Entropy analysis of the current waveforms of same phases for the two turbines are found to have different complexity values over time, one of them being clearly higher than the other. The work of Yan et. al. in has found that higher entropy values related to thepresence of failure in rotary machines in his study. Following this track, further efforts will be put into relating the entropy difference found in our study to possible presence of failure in one of the wind energy conversion systems.
قيم البحث
اقرأ أيضاً
Wind farm design primarily depends on the variability of the wind turbine wake flows to the atmospheric wind conditions, and the interaction between wakes. Physics-based models that capture the wake flow-field with high-fidelity are computationally v
ery expensive to perform layout optimization of wind farms, and, thus, data-driven reduced order models can represent an efficient alternative for simulating wind farms. In this work, we use real-world light detection and ranging (LiDAR) measurements of wind-turbine wakes to construct predictive surrogate models using machine learning. Specifically, we first demonstrate the use of deep autoencoders to find a low-dimensional emph{latent} space that gives a computationally tractable approximation of the wake LiDAR measurements. Then, we learn the mapping between the parameter space and the (latent space) wake flow-fields using a deep neural network. Additionally, we also demonstrate the use of a probabilistic machine learning technique, namely, Gaussian process modeling, to learn the parameter-space-latent-space mapping in addition to the epistemic and aleatoric uncertainty in the data. Finally, to cope with training large datasets, we demonstrate the use of variational Gaussian process models that provide a tractable alternative to the conventional Gaussian process models for large datasets. Furthermore, we introduce the use of active learning to adaptively build and improve a conventional Gaussian process model predictive capability. Overall, we find that our approach provides accurate approximations of the wind-turbine wake flow field that can be queried at an orders-of-magnitude cheaper cost than those generated with high-fidelity physics-based simulations.
Measuring the size of permanent electric dipole moments (EDM) of a particle or system provides a powerful tool to test Beyond-the-Standard-Model physics. The diamagnetic $^{129}$Xe atom is one of the promising candidates for EDM experiments due to it
s obtainable high nuclear polarization and its long spin-coherence time in a homogeneous magnetic field. By measuring the spin precession frequencies of polarized $^{129}$Xe and $^{3}$He, a new upper limit on the $^{129}$Xe atomic EDM $d_mathrm{A}(^{129}mathrm{Xe})$ was reported in Phys. Rev. Lett. 123, 143003 (2019). This writeup proposes a new evaluation method based on global phase fitting (GPF) for analyzing the continuous phase development of the $^{3}$He-$^{129}$Xe comagnetometer signal. The Cramer-Rao Lower Bound on the $^{129}$Xe EDM for the GPF method is theoretically derived and shows the benefit of achieving high statistical sensitivity without bringing new systematic uncertainties. The robustness of the GPF method is verified with Monte-Carlo studies. By optimizing the analysis parameters and adding few more data that could not be analyzed with the former method, a result of [ { d_mathrm{A} (^{129}mathrm{Xe})=(1.1 pm 3.6_mathrm{(stat)} pm 2.0_mathrm{(syst)})times 10 ^{-28} e~mathrm{cm}}, ] is obtained and is used to derive the upper limit of $^{129}$Xe permanent EDM at 95% C.L. [ {|d_text{A}(^{129}text{Xe})| < 8.3 times 10^{-28}~e~mathrm{cm}}. ] This limit is a factor of 1.7 smaller as compared to the previous result.
The maximum entropy principle can be used to assign utility values when only partial information is available about the decision makers preferences. In order to obtain such utility values it is necessary to establish an analogy between probability an
d utility through the notion of a utility density function. According to some authors [Soofi (1990), Abbas (2006a) (2006b), Sandow et al. (2006), Friedman and Sandow (2006), Darooneh (2006)] the maximum entropy utility solution embeds a large family of utility functions. In this paper we explore the maximum entropy principle to estimate the utility function of a risk averse decision maker.
Velocity measurements of wind blowing near the North Sea border of Northern Germany and velocity measurements under local isotropic conditions of a turbulent wake behind a cylinder are compared. It is shown that wind gusts - measured by means of velo
city increments - do show similar statistics to the laboratory data, if they are conditioned on an averaged wind speed value. Clear differences between the laboratory data and the atmospheric wind velocity measurement are found for the waiting time statistics between successive gusts above a certain threshold of interest.
A promising method for real time early warning of gravity driven rupture that considers both the heterogeneity of natural media and characteristics of acoustic emissions attenuation is proposed. The method capitalizes on co-detection of elastic waves
emanating from micro-cracks by multiple and spatially separated sensors. Event co-detection is considered as surrogate for large event size with more frequent co-detected events marking imminence of catastrophic failure. Using a spatially explicit fiber bundle numerical model with spatially correlated mechanical strength and two load redistribution rules, we constructed a range of mechanical failure scenarios and associated failure events (mapped into AE) in space and time. Analysis considering hypothetical arrays of sensors and consideration of signal attenuation demonstrate the potential of the co-detection principles even for insensitive sensors to provide early warning for imminent global failure.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
