基于多层小波分解的日前太阳辐照度预测研究

Day-ahead Irradiance ForecastingBased Multi-level Wavelet Decomposition

由于太阳辐照度的随机波动特性,大型光伏发电并网会给电力系统的运行带来极大困难,光伏发电功率的预测是解决此问题的关键措施之一。提出了一种基于多层小波分解的太阳辐照度预测方法,首先,根据天气状态将每日的辐照度曲线划分为不同的波动模式;然后针对不同天气下的波动模式分别建立预测模型,使用多层小波分解后的数据预测第二天连续24小时的辐照度值;最后建立基于数据驱动的融合模型,将不同天气模式下的辐照度多层小波分解预测值进行融合,以获得最终的辐照度预测结果。仿真结果表明辐照度预测结果精度与小波分解层数和天气模式高度相关,且所提算法能够有效提高短期辐照度预测精度。

Due to the stochastic fluctuant characteristic of solar irradiance,large-scalegrid-connected photovoltaic(PV)power plant can bring great difficulties to the operation of power system.One feasible way to solve this problem is PV power forecasting.In this paper,a multi-level wavelet decomposition based day-ahead solar irradiance forecasting method is proposed in this paper.First,the daily solar irradiance series are classified into different patterns according to weather conditions.Then for each weather pattern,the solar irradiance of the next day 24 hours is forecasted using decomposed data series at different WD level.Then a data-Zdriven fusion model corresponding to the weather pattern is applied to fuse the predictions into the final forecasting output.Simulations show that the forecasting accuracyusing different WD level data is highly relevant to the weather conditions,and the effectiveness of the proposed method is proved by the simulation results.