Yiwei Yuan

Variable selection based on frequent pattern tree for spectroscopy quantitative analysis

Spectroscopy is an important component analysis method and full spectrum prediction method may be complicated and inaccurate. In order to find out the irrelevant variables, a variable selection method based on the frequent pattern tree(FP-tree) is proposed in this paper. The proposed method firstly formulates an orthogonal array to generate wavelength selection plans, which makes the wavelength selection more reasonable. Then, partial least square(PLS) is performed on the spectral data which is selected by the orthogonal array, and the root-mean-square error of cross validation is adopted as the evaluation criteria of the performances of each prediction model. Based on the results of the evaluation, A set of data which contains the wavelength selection of these model whose performance are evaluated as good can be got. The support count of each wavelength is determined according to the data set and the infrequent wavelengths are discarded while frequent wavelengths are sorted in decreasing support counts. After this, an FP-tree is built based on processed data set. The final selection result is a branch in FP-tree which has the maximum number of sum frequent count. The prediction model is built based on the selection result by PLS. The full spectrum PLS, the uninformative variable elimination with the PLS method and the proposed method are conducted on a real spectral data set of flue gas, the effectiveness of these methods are compared and discussed. The experimental results show that the proposed method is more accurate and has presentable compression performance.

Statistical operators calculation of partial discharge on floating electrode defect in GIS

This paper considers the features extraction of Gas Insulated metal-enclosed switchgear(GIS) which suffer from partial discharge (PD) by floating electrode defect. The 220 kV GIS bus bar cavity is taken as an object to design physical model of floating electrode defect. The experiment PD data are obtained from a 220 kV GIS equipment using Ultra-high frequency(UHF) partial discharge measurement. The data is stored as phase resolved pulse sequence (PRPS) data format. This data format allows the actual test voltage (rms) of every cycle, the number of measuring cycles, phase information of all PD pulses, together with their peak values to be obtained, so the discussion of statistical analysis of the floating electrode defect can be performed. Firstly, the phase resolved partial discharge (PRPD) pattern is calculated from the PRPS data sets. Then, the statistical operators at different voltage levels are extracted from the PRPD patterns of the floating electrode defect. On the basis of the present experimental research and statistical analysis, statistical operators such as skewness, kurtosis can be used to classify different PD defects.

Measurement and analysis of partial discharge on floating electrode defect in GIS

In this paper, a 220kV GIS bus bar model is established to study partial discharge characteristics of floating electrode defect in Gas Insulated Switchgear (GIS), which is one of the common sources of PD. The 220kV GIS bus bar cavity is taken as an object to design physical model of floating electrode defect. The PD development process is researched by applying gradually increased voltage. The UHF method and the acoustic method are used to detect the PD. The PRPD patterns at different voltage levels of PD are accessed through the PD data of UHF signal. The discharge patterns of different voltage levels at continuous mode and phase mode are obtained through ultrasonic method. These measurements can provide test basis for the assessment of partial discharge provoked by floating electrode defect in GIS. It is shown by experiments that the PD characteristics change as the applied voltage raised. Their specific performances are the discharge number at positive and negative half cycle getting more, and the phase interval gradually widened with the increase of applied voltage. But the discharge amplitude has no dramatic change with the increase of the applied voltage. By comparison of the UHF method and the acoustic method, it is concluded that UHF method has advantages of interference immunity, high sensitivity. The acoustic method is electromagnetic noise immunity. The acoustic sensor do not require intrusive installation and do not destroy the GIS structure itself. Compared to the UHF sensor, it may be more cost effective. But the acoustic signal is attenuated and the measurement sensitivity is under the influence of the temperature. The two methods both can make PD location using the time difference of the different sensors. The two kinds of measurements can provide test basis for the assessment of partial discharge provoked by floating electrode defect in GIS.

Study on the Partial Discharge Characteristics and Development Process in Use of the Multiple Discharge Patterns for the Typical Defects in Gas-Insulated Switchgear

This paper presents an experiment to study the partial discharge (PD) characteristics and development process of the physical models of general insulation defects in gas-insulated switchgear (GIS). Four types of typical artificial insulation defects are designed to imitate the insulation faults in an actual GIS. The experiment PD data are obtained from a 220-kV GIS equipment using ultrahigh-frequency (UHF) PD measurement. The PD data obtained by UHF detection are stored as phase-resolved pulse sequence (PRPS) data. Then, the multiple PD patterns, such as PRPS analysis pattern, phase-resolved PD pattern, and Δu pattern, can be calculated from the PRPS data sets. On the basis of the present experimental research, the multiple PD patterns can be used to classify different PD defects and evaluate the stage of PD development process comprehensively of each typical defect.

New local density definition based on minimum hyper sphere for outlier mining algorithm using in industrial databases

Outlier detection is an important procedure in industrial dataset preprocess to guarantee the industrial process operating normally. This paper proposed a new local density definition in the basis of the minimum hyper sphere for outlier mining algorithm. First, the novel local k-density definition of an object is proposed by using the minimum enclosing hyper sphere algorithm. After this, the new k-density definition is adopt in local outlier factor (LOF) algorithm, INFLuenced Outlierness (INFLO) algorithm, and the density-similarity-neighbor-based outlier mining (DSNOF) algorithm constructing ndLOF algorithm, ndINFLO algorithm, and ndDSNOF algorithm. Finally, we evaluate the performance of ndLOF algorithm, ndINFLO algorithm, and ndDSNOF algorithm with LOF algorithm, INFLO algorithm, and DSNOF algorithm on synthetic datasets. The experiments results confirm that the presented definition is meaningful and the outlier mining algorithms improved by the new definition have higher quality of outlier mining.