%0 Journal Article
%T 新建电气化铁路牵引负荷预测<br>Prediction of Traction Load for New Electrified Railway
%A 张丽艳
%A 李群湛
%A 朱毅&lt
%A br&gt
%A ZHANG Liyan
%A LI Qunzhan
%A ZHU Yi
%J 西南交通大学学报
%D 2016
%R 10.3969/j.issn.0258-2724.2016.04.020
%X 为了评估新建电气化铁路对电网电能质量的影响,提出了一种基于实测数据的牵引负荷统计预测方法.该方法基于大量的牵引负荷实测数据,在统计分析其分布特征的基础上,选择带电有效系数、最大值、方差和偏度系数作为描述牵引负荷概率分布的主要特征量;应用模糊C均值聚类法,将42组牵引负荷实测数据分成10类,根据铁路设计部门提供的牵引负荷特征值,判断新建电气化铁路牵引负荷归属10类概率模型特征库中的某一类,进而可知其概率分布,采用蒙特卡洛抽样,即可获得新建电气化铁路牵引变电所馈线电流的预测数据;用均方差指标对拟合曲线进行误差校验,误差均在0.1以内,证实了方法的有效性.<br>： A statistical forecasting method based on the measured data is proposed to evaluate the influence of a new electrified railway on the power quality of power grid. Through analysis of statistical distribution characteristics of a large number of traction load test data, parameters such as the charged effective coefficient, maximum value, variance, and skewness coefficient are selected to describe the main characteristics of the traction load in probability distribution. Using the fuzzy C-means clustering algorithm, 42 groups of measured data are divided into 10 types. Based on the characteristic values of traction load provided by railway design departments, the probability model of the new electrified railway can be derived from the above 10 types, and its probability distribution function (PDF) is accordingly obtained. Finally, the feeder currents in traction substation of the new electrified railway are predicted by Monte Carlo sampling method. The validity of the method is verified by the fitting curves of PDF using the predicted data and the mean square error of PDF is less than 0.1
%K 牵引负荷预测
%K 模糊C均值聚类法
%K 概率密度函数
%K &lt
%K br&gt
%K traction load prediction
%K fuzzy C-means clustering algorithm
%K probability density function
%U http://manu19.magtech.com.cn/Jweb_xnjd/CN/abstract/abstract12301.shtml