THE APPLICATION OF STATISTICAL PARAMETERS OF PHASE RESOLVED PD DISTRIBUTION TO AGING EXTENT ASSESSMENT OF LARGE GENERATOR INSULATION

Objective To investigate the characteristic parameters employed to describe the aging extent of stator insulation of large generator and study the aging laws. Methods Multi stress aging tests of model generator stator bar specimens were performed and PD measurements were conducted using digital PD detector with frequency range from 40?kHz to 400?kHz at different aging stage. Results From the test results of model specimens it was found that the skewness of phase resolved PD distribution might be taken as the characterization parameters for aging extent assessment of generator insulation. Furthermore, the measurement results of actual generator stator bars showed that the method based on statistical parameters of PD distributions are prospective for aging extent assessment and residual lifetime estimation of large generator insulation. Conclusion Statistical parameters of phase resolved PD distribution was proposed for aging extent assessment of large generator insulation.

On the surface roughness characteristics of the land fast sea-ice in the Bohai Sea

The surface roughness characteristics （e.g., height and slope） of sea ice are critical for determining the parameters of an electromagnetic scattering, a surface emission and a surface drag coefficients. It is also important in identifying various ice types, retrieval ice thickness, surface temperature and drag coefficients from remote sensing data. The point clouds （a set of points which are usually defined by X, Y, and Z coordinates that represents the external surface of an object on earth） of land fast ice in five in situ sites in the eastern coast Bohai Sea were measured using a laser scanner-Trimble GX during 2011-2012 winter season. Two hundred and fifty profiles selected from the point clouds of different samples have been used to calcu- late the height root mean square, height skewness, height kurtosis, slope root mean square, slope skewness and slope kurtosis of them. The root mean square of the height, the root mean square of the slope and the correlation length are about 0.090, 0.075 and 11.74 m, respectively. The heights of 150 profiles in three sites manifest the Gaussian distribution and the slopes of total 250 profiles distributed exponentially. In addition, the fractal dimension and power spectral density profiles were calculated. The results show that the fractal dimension of land fast ice in the Bohai Sea is about 1.132. The power spectral densities of 250 profiles can be expressed through an exponential autocorrelation function.

Data-driven estimation of joint roughness coefficient

Joint roughness is one of the most important issues in the hydromechanical behavior of rock mass.Therefore,the joint roughness coefficient(JRC)estimation is of paramount importance in geomechanics engineering applications.Studies show that the application of statistical parameters alone may not produce a sufficiently reliable estimation of the JRC values.Therefore,alternative data-driven methods are proposed to assess the JRC values.In this study,Gaussian process(GP),K-star,random forest(RF),and extreme gradient boosting(XGBoost)models are employed,and their performance and accuracy are compared with those of benchmark regression formula(i.e.Z2,Rp,and SDi)for the JRC estimation.To analyze the models’performance,112 rock joint profile datasets having eight common statistical parameters(R_(ave),R_(max),SD_(h),iave,SD_(i),Z_(2),R_(p),and SF)and one output variable(JRC)are utilized,of which 89 and 23 datasets are used for training and validation of models,respectively.The interpretability of the developed XGBoost model is presented in terms of feature importance ranking,partial dependence plots(PDPs),feature interaction,and local interpretable model-agnostic explanations(LIME)techniques.Analyses of results show that machine learning models demonstrate higher accuracy and precision for estimating JRC values compared with the benchmark empirical equations,indicating the generalization ability of the data-driven models in better estimation accuracy.

The improved model of estimating global whitecap coverage based on satellite data

The pro and con of whitecap parameterizations and a statistical wave breaking model are discussed. An improved model is derived by combining satellite-based parameterization and the wave breaking model. The appropriate constants for the general wave state are obtained by considering the breaking condition of the wave slope and fitting with the satellite-based parameterization. The result is close to the constants based on the whitecap data from Monahan. Comparing with satellite-based data and the original model＇s results, the improved model＇s results are consistent with satellite-based data and previous studies. The global seasonal distributions of the whitecap coverage averaged from 1998 to 2008 are presented. Spatial and seasonal features of the whitecap coverage are analyzed.

Spectral classification of stars based on LAMOST spectra

In this work, we select spectra of stars with high signal-to-noise ratio from LAMOST data and map their MK classes to the spectral features. The equivalent widths of prominent spectral lines, which play a similar role as multi-color photometry, form a clean stellar locus well ordered by MK classes. The advantage of the stellar locus in line indices is that it gives a natural and continuous classification of stars consistent with either broadly used MK classes or stellar astrophysical parameters. We also employ an SVM-based classification algorithm to assign MK classes to LAMOST stellar spectra. We find that the completenesses of the classifications are up to 90% for A and G type stars, but they are down to about 50% for OB and K type stars. About 40% of the OB and K type stars are mis-classified as A and G type stars,respectively. This is likely due to the difference in the spectral features between late B type and early A type stars or between late G and early K type stars being very weak. The relatively poor performance of the automatic MK classification with SVM suggests that the direct use of line indices to classify stars is likely a more preferable choice.

A tool for the morphological classification of galaxies: the concentration index

Using galaxy data from the Sloan Digital Sky Survey Data Release 8, I ex- plore whether the concentration index is a good morphological classification tool and find that a reasonable sample of pure late-type galaxies can be constructed with the choice of the r-band concentration index ci=2.85. The opposite is not true, however, due to the fairly high contamination of an early-type sample by late-type galaxies. In such an analysis, the influence of selection effects is less important. To disentangle correlations of the morphology and concentration index with stellar mass, star forma- tion rate （SFR）, specific star formation rate （SSFR） and active galactic nucleus （AGN） activity, I investigate correlations of the concentration index with these properties at a fixed morphology and correlations of the morphology with these properties at a fixed concentration index. It is found that at a fixed morphology, high-concentration galaxies are preferentially more massive and have a lower SFR and SSFR than low- concentration galaxies, whereas at a fixed concentration index, elliptical galaxies are preferentially more massive and have a lower SFR and SSFR than spiral galaxies. This result shows that the stellar mass, SFR and SSFR of a galaxy are correlated with its concentration index as well as its morphology. In addition, I note that AGNs are pref- erentially found in more concentrated galaxies only in the sample of spiral galaxies.

Degeneracy and discreteness in cosmological model fitting

We explore the problems of degeneracy and discreteness in the standard cosmological model（ΛCDM）. We use the Observational Hubble Data（OHD） and the type Ia supernovae（SNe Ia） data to study this issue. In order to describe the discreteness in fitting of data, we define a factor G to test the influence from each single data point and analyze the goodness of G. Our results indicate that a higher absolute value of G shows a better capability of distinguishing models, which means the parameters are restricted into smaller confidence intervals with a larger figure of merit evaluation. Consequently, we claim that the factor G is an effective way of model differentiation when using different models to fit the observational data.

DWT Power Spectrum of the Two-Degree Field Galaxy Redshift Survey

The power spectrum of the two-degree Field Galaxy Redshift Survey （2dFGRS） sample is estimated with the discrete wavelet transform （DWT） method. The DWT power spectra within 0.035 〈 k 〈 2.2 h Mpc^-1 are measured for three volume-limited samples defined in consecutive absolute magnitude bins - 19 - - 18, - 20 - - 19 and - 21 - - 20. We show that the DWT power spectrum can effectively distinguish ACDM models of σ8 = 0.84 and σ8 = 0.74. We adopt maximum likelihood method to perform three-parameter fitting of the bias parameter b, pairwise velocity dispersion σpv and redshift distortion parameterβ = Ωm^0.6/b to the measured DWT power spectrum. The fitting results state that in a σ8 = 0.84 universe the best-fit values of Ωm given by the three samples are mutually consistent within the range 0.28 - 0.36, and the best fitted values of Opv are 398-27^＋35, 475-29^37 and 550 ± 20 km s^-1 for the three samples, respectively. In the model of σ8 = 0.74, our three samples give very different values of Ωm. We repeated the fitting using the empirical formula of redshift distortion. The result of the model of low σ8 is still poor, especially, one of the best-fit values of σpv is as large as 10^3 km s^-1. We also repeated our fitting by incorporating a scale-dependent galaxy bias. This gave a slightly lower value of Ωm. Differences between the models of σ8 = 0.84 and σ8 = 0.74 still exist in the fitting results. The power spectrum of 2dFGRS seems to disfavor models with low amplitude of density fluctuations if the bias parameter is assumed to be scale independent. For the fitting value of Ωm to be consistent with that given by WMAP3, strong scale dependence of the bias parameters is needed.

Stator Fault Diagnosis of Induction Motor Based on Discrete Wavelet Analysis and Neural Network Technique

A novel approach by introducing a statistical parameter to estimate the severity of incipient stator inter-turn short circuit(ITSC)faults in induction motors(IMs)is proposed.Determining the incipient ITSC fault and its severity is challenging for several reasons.The stator currents in the healthy and faulty cases are highly similar during the primary stage of the fault.Moreover,the conventional statistical parameters resulting from the analysis of fault signals do not consistently show a systematic variation with respect to the increase in fault intensity.The objective of this study is the early detection of incipient ITSC faults.Furthermore,it aims to determine the percentage of shorted turns in the faulty phase,which acts as an indicator for severe damage to the stator winding.Modeling of the motor in healthy and defective cases is performed using the Clarke Concordia transform.A discrete wavelet transform is applied to the motor currents using a Daubechies-8 wavelet.The statistical parameters L1 and L2 norms are computed for the detailed coefficients.These parameters are obtained under a variety of loads and defects to acquire the most accurate and generalized features related to the fault.Combining L1 and L2 norms creates a novel statistical parameter with notable characteristics to achieve the research aim.An artificial neural network-based back propagation algorithm is employed as a classifier to implement the classification process.The classifier output defines the percentage of defective turns with a high level of accuracy.The competency of the adopted methodology is validated via simulations and experiments.The results confirm the merits of the proposed method,with a classification test correctness of 95.29%.

Probabilistic model for vessel-bridge collisions in the Three Gorges Reservoir

Based on a field observation on vessel transit path of three bridges over the Yangtze River in the Three Gorges Reservoir,and an analysis of the geometric probabilistic model of transiting vessels in collision probability calculation,the aberrancy angle and vessel velocity probabilistic model related with impact force,a probabilistic model is established and also verified by goodness-of-fit test.The vessel transit path distribution can be expressed by the normal distribution model.For the Three Gorges Reservoir,the mean and standard deviation adopt 0.2w and 0.1w,respectively(w is the channel width).The aberrancy angle distribution of vessels accepts maximum I distribution model,and its distribution parameters can be taken as 0.314 and 4.354.The velocity distribution of up-bound and down-bound vessels can also be expressed by the normal distribution model.