基于相空间重构ARIMA和SVR的母线净负荷预测

Forecasting of bus.bar net load based on PSR.ARIMA and SVR

中文关键词: 净负荷预测  ARIMA  CC法相空间重构  SVR

英文关键词: net load forecasting  ARIMA  CC method  SVR

中文摘要:

      实现对含分布式电源母线净负荷的实时跟踪和预测，以分布式光伏并入电网的母线负荷情况为基础，根据净负荷本身固有的线性和非线性属性，提出了基于相空间重构自回归滑动平均( autoregressive moving arerage, ARIMA)和支持向量机( support vector machine, SVR)的母线净负荷预测方法。首先基于历史净负荷数据使用 ARIMA建立拟合模型，对净负荷线性成分预测分析，之后用 CC法对非线性成分进行相空间重构，利用 SVR模型对非线性部分进行预测。数据建模的结果表明，提出的 ARIMA.CC_SVR预测模型对含有分布式光伏成分的母线净负荷适用性较强。

英文摘要:

      In order to realize the real.time tracking and fore.casting of net load with distributed power bus, in the light of the busload with distributed photovoltaic power grid, according to the linearand nonlinear properties of net load inherent, the autoregressivemoving average with the phase space reconstruction and support vec.tor machine bus.bar net load forecasting method is proposed.Firstly,based on the historical net load data, the ARIMA model is estab.lished to forecast the linear components of the net load.Then, thephase space of the nonlinear components is reconstructed by CCmethod.Finally, the nonlinear part is forecasted by the SVR model.The results of data modeling show that the proposed ARIMA .CC_SVR forecasting model is suitable for the net load of the bus.barwith distributed PV components.