基于时间序列的电能路由器能流预测模型

Energy Flow Forecasting for Energy Router Based on Time Series

为解决以电能路由器为基本节点的能源互联网中的能流预测问题,分析了居民用电、大型商业用电、小型商业用电、光伏出力四类能流的主要影响因素,利用时间序列中的ARIMA模型构建了不同性质的能流预测模型,在此基础上综合考虑不同性质的能流和能源局域网的路由策略,进而构建了电能路由器能流预测模型。将该模型应用于New Hampshire Electric Co-op（NHEC）2013年3月的能流预测时,发现在路由策略下分布式电源充分条件下储能设备充裕时,对电能路由器能流EEF无影响;储能设备有限时,对EEF影响较大,EEF频繁波动;无储能设备时,对EEF影响较大,10：00~19：00时波动大;在分布式电源有限条件下,三种储能设备情况下,对电能路由器能流EEF均影响较大。

In order to solve the problem of energy flow forecasting in energy Internet consisting of energy router as its basic node,analyzing the main factors of four types of energy flow including the residential electricity,large commercial electricity,small commercial electricity and photovoltaic output,ARIMA model in time series was constructed to predict the different kinds of energy flow.And then considering the different nature of energy flow and energy LAN routing strategy,energy flow forecasting model for energy router was established.The model was applied in New Hampshire Electric Co-op（NHEC）energy flow forecasting in March of 2013.It proves that when the energy storage device is abundant in the routing policy,it has no effect on the energy flowEEEF.When energy storage device is finite,there is a greater impact on EEEF,which has frequent fluctuations.Without energy storage equipment,EEEFhas great fluctuations between10：00to 19：00pm.Under the distributed power limited conditions and within three kinds of energy storage device,it has greater impact on EEEFof energy router.