摘要
一般采用小波分解的电价预测方法是将历史电价分解后分别预测,预测过程中没有引入其他电价影响因素,或者是直接引入未经小波分解的影响因素。提出一种小波分析与神经网络相结合的预测方法,将历史电价和历史负荷都进行小波多分辨率单尺度分解,分解成概貌电价、细节电价和概貌负荷、细节负荷。在此基础上,用历史概貌电价和概貌负荷序列训练BP神经网络,预测出未来的概貌电价;用历史细节电价和细节负荷序列训练BP神经网络,预测出未来的细节电价。将概貌电价和细节电价进行重构,得到最终的预测电价。对美国PJM电力市场的实际电价(LMP)进行预测,验证了该方法的有效性和可行性。
The common electricity price forecast method based on wavelet analysis decomposes the historic prices and forecasts respectively,without introducing the influencing factors of price to the forecast model,or introducing the influencing factors of price undecomposed by wavelet analysis. A method combining wavelet analysis and neural network is proposed,which decomposes both historic prices and loads into approximate price series,detailed price series,approximate load series,and detailed load series by multi- resolution wavelet analysis. The BP neural network is trained by the approximate price series and approximate load series to forecast the future approximate price and by the detailed price series and detailed load series to forecast the future detailed price. The final forecasted price is the reconstruction of the forecasted approximate price and the forecasted detailed price. The simulation of PJM power market＇s LMP verifies the effectiveness and feasibility of the proposed method.
出处
《电力自动化设备》
EI
CSCD
北大核心
2007年第11期26-29,33,共5页
Electric Power Automation Equipment
关键词
电价预测
BP神经网络
小波分析
电力市场
electrity price forecast
BP neural network
wavelet analysis
power market
