为解决因排查效率低、数据更新不及时等因素导致低压配电网户变关系连接形式与实际不符的问题,提出一种基于角度分段线性近似(anglepiecewiselinearrepresentation,APLR)和改进密度峰值聚类(improved clustering by fast search find of...为解决因排查效率低、数据更新不及时等因素导致低压配电网户变关系连接形式与实际不符的问题,提出一种基于角度分段线性近似(anglepiecewiselinearrepresentation,APLR)和改进密度峰值聚类(improved clustering by fast search find of density peaks,ICFSFDP)相结合的户变关系识别方法。首先,根据电压曲线中相邻线段的角度变化量提取曲线的转折点,利用APLR对曲线进行自适应降维重构;随后,使用ICFSFDP算法对降维数据组展开聚类分析,在决策图中由拟合函数与坐标轴围成面积的最小值得到最优类簇数目,进而得到聚类和非聚类中心用户;最后,使用动态时间弯曲(dynamic time warping,DTW)距离计算聚类和非聚类中心用户之间的距离相似度,进而得到户变关系。将所提方法应用于模拟和真实数据中,均可证实所提方法的有效性。算例分析结果表明:该方法能够对时间间隔不同、不等维的序列进行分析,且不需要人为设定聚类算法的参数,户变关系识别准确率高。展开更多
To manage the wall thinning of carbon steel piping in nuclear power plants, the utility of Korea has performed thickness inspection for some quantity of pipe components during every refueling outage and determined whe...To manage the wall thinning of carbon steel piping in nuclear power plants, the utility of Korea has performed thickness inspection for some quantity of pipe components during every refueling outage and determined whether repair or replacement after evaluating UT data. Generally used UT thickness data evaluation methods are Band, Blanket, and PTP (Point to Point) methods. Those may not desirable to identify wall thinning on local area caused by erosion. This is because the space between inspecting points of those methods are wide for covering full surface being inspected components. When the evaluation methods are applied to a certain pipe component, unnecessary re-inspection may also be generated even though wall thinning of components does not progress. In those cases, economical loss caused by repeated inspection and problems of maintaining the pipe integrity followed by decreasing the number of newly inspected components may be generated. EPRI (Electric Power Research Institute in USA) has suggested several statistical methods such as FRIEDMAN test method, ANOVA (Analysis of Variance) method, Monte Carlo method, and TPM (Total Point Method) to distinguish whether multiple inspecting components have been thinned or not. This paper presents the NAM (Near Area of Minimum) method developed by KEPCO-E & C for distinguishing whether multiple inspecting components have been thinned or not. In addition, this paper presents the analysis results for multiple inspecting ones over three times based on the NAM method compared with the other methods suggested by EPRI.展开更多
Based on tropical cyclone (TC) data for the period 1949 to 2008 and following the Gumbe-Imethod, Pearson-Ⅲ method and determinacy method, this article estimates the possible minimum centralpressure of TCs affecting s...Based on tropical cyclone (TC) data for the period 1949 to 2008 and following the Gumbe-Imethod, Pearson-Ⅲ method and determinacy method, this article estimates the possible minimum centralpressure of TCs affecting southern Fujian where a nuclear power will be located. Results show that theobserved minimum central pressure of TCs agrees well with the results determined with the methods aboveand there is little difference between them (the minimum central pressure is 867.4 hPa and 868.1 hPa,respectively, in a 1,000-yr return period). Established with the theory of atmospheric dynamics, thedeterminacy method yields a result of 867.28 hPa/1000 years, only a little smaller than the result of theprobability method. Because of randomicity in parameter adjustment with the Pearson-Ⅲ method whereasthe determinacy method is theoretically solid and its estimates are the smallest of the three methods, it istherefore reasonable, for security and conservative concerns, to adopt the result determined with thedeterminacy method as the possible maximum intensity of TC (with the central pressure being 867.28 hPa ina 1,000-yr return period).展开更多
文摘为解决因排查效率低、数据更新不及时等因素导致低压配电网户变关系连接形式与实际不符的问题,提出一种基于角度分段线性近似(anglepiecewiselinearrepresentation,APLR)和改进密度峰值聚类(improved clustering by fast search find of density peaks,ICFSFDP)相结合的户变关系识别方法。首先,根据电压曲线中相邻线段的角度变化量提取曲线的转折点,利用APLR对曲线进行自适应降维重构;随后,使用ICFSFDP算法对降维数据组展开聚类分析,在决策图中由拟合函数与坐标轴围成面积的最小值得到最优类簇数目,进而得到聚类和非聚类中心用户;最后,使用动态时间弯曲(dynamic time warping,DTW)距离计算聚类和非聚类中心用户之间的距离相似度,进而得到户变关系。将所提方法应用于模拟和真实数据中,均可证实所提方法的有效性。算例分析结果表明:该方法能够对时间间隔不同、不等维的序列进行分析,且不需要人为设定聚类算法的参数,户变关系识别准确率高。
文摘To manage the wall thinning of carbon steel piping in nuclear power plants, the utility of Korea has performed thickness inspection for some quantity of pipe components during every refueling outage and determined whether repair or replacement after evaluating UT data. Generally used UT thickness data evaluation methods are Band, Blanket, and PTP (Point to Point) methods. Those may not desirable to identify wall thinning on local area caused by erosion. This is because the space between inspecting points of those methods are wide for covering full surface being inspected components. When the evaluation methods are applied to a certain pipe component, unnecessary re-inspection may also be generated even though wall thinning of components does not progress. In those cases, economical loss caused by repeated inspection and problems of maintaining the pipe integrity followed by decreasing the number of newly inspected components may be generated. EPRI (Electric Power Research Institute in USA) has suggested several statistical methods such as FRIEDMAN test method, ANOVA (Analysis of Variance) method, Monte Carlo method, and TPM (Total Point Method) to distinguish whether multiple inspecting components have been thinned or not. This paper presents the NAM (Near Area of Minimum) method developed by KEPCO-E & C for distinguishing whether multiple inspecting components have been thinned or not. In addition, this paper presents the analysis results for multiple inspecting ones over three times based on the NAM method compared with the other methods suggested by EPRI.
基金Natural Science Foundation of Fujian Province(2007J0122)Natural Science Foundation of China(90915002)
文摘Based on tropical cyclone (TC) data for the period 1949 to 2008 and following the Gumbe-Imethod, Pearson-Ⅲ method and determinacy method, this article estimates the possible minimum centralpressure of TCs affecting southern Fujian where a nuclear power will be located. Results show that theobserved minimum central pressure of TCs agrees well with the results determined with the methods aboveand there is little difference between them (the minimum central pressure is 867.4 hPa and 868.1 hPa,respectively, in a 1,000-yr return period). Established with the theory of atmospheric dynamics, thedeterminacy method yields a result of 867.28 hPa/1000 years, only a little smaller than the result of theprobability method. Because of randomicity in parameter adjustment with the Pearson-Ⅲ method whereasthe determinacy method is theoretically solid and its estimates are the smallest of the three methods, it istherefore reasonable, for security and conservative concerns, to adopt the result determined with thedeterminacy method as the possible maximum intensity of TC (with the central pressure being 867.28 hPa ina 1,000-yr return period).
