小波包变换在冲击负荷地区短期负荷预测中的应用

Application of Wavelet Packet Transform in Short Term Load Forecasting of Impact Load Area

针对大量冲击负荷接入系统给负荷预测造成困难,提出一种基于小波包变换的电力系统短期负荷预测新方法:通过小波包变换将负荷序列分解成不同频率的负荷子序列,根据负荷分量不同的变化特点,分别采用相匹配的模型进行预测,最后叠加各子序列的预测值,得出实际预测结果。选择广东某电网实测负荷数据作为算例,仿真结果表明该方法能有效提高冲击负荷地区的短期负荷预测精度。

Facing the difficulties caused by a large number of random impact loads access systems for load forecasting, this paper proposes a new method based on wavelet packet transform for short term load prediction ： decomposing the load sequence into load sub-sequences under different band by wavelet packet transform ; adopts matched models to predict according to different features of load subsequences, superpose all sequences forecasting outputs, obtain the final forecasted resuhs. At the end, a certain power system in Guangdong is chosen to be the case, the results show that this method can improve the impact load area for short term load forecasting accuracy.