摘要
基于华东某市配电网故障报修数据,开展配电网故障数量短期预测研究。提出基于气温的季节判定方法,综合采用多元回归和时间序列分析手段,构建分季节的气象影响故障量预测模型,确定温度、天气等气象因素与故障量的量化关系,并针对剔除气象因素影响的剩余故障量,构建自回归移动平均时间序列预测模型,捕捉故障量的时间序列变化趋势。通过上述模型的综合应用,实现了配电网故障数量较高精度的短期预测。
Based on the failure data from a distribution network in east China, the paper proposes an air temperature based season judgment method. Through multiple regression and time series analysis, the paper establishes the forecasting model for distribution network faults in different seasons considering meteorological factors such as temperature and weather, and determines a quantitative relationship between the meteorological factors and the faults. For the failures which are not explained by the meteorological factors, the paper builds autoregressive integrated moving average（ARIMA） time series model for further prediction, in order to get the high short-period prediction precision of the distribution network failures by the comprehensive application of the regression model and time series model.
出处
《陕西电力》
2016年第1期68-72,共5页
Shanxi Electric Power
关键词
配电网
故障
预测
线性回归
时间序列
ARIMA
dislribulion network
fault
predietion
multiple regression
time series
ARIMA
