基于小波神经网络方法的电力需求预测

Electric Power Demand Forecasting Based on Wavelet Neural Network Method

当前,诸多研究人员被电力负载预测所吸引,由于其是精确计划、调度及运维电力系统的先决条件。众多因素均影响着电力负载预测,因此提出一个混合模型来提升预测的准确性是有必要的。文中提出一种采用2种方法的新的混合负载估计方案:小波变换(avelet transform,WT)和人工神经网络(artificial neural network,ANN)。为了将大型非对称时变电力原始数据集合考虑到其中,根据时间和频率采用小波技术来分解数据,众多小波函数可以采用,但选择一种合适的小波函数在设计此模型中扮演着关键作用。文中采用了以下几种类型的小波函数,即Haar小波函数、Deubechies小波函数、Symlet小波函数以及Coiflet小波函数,将电力负载数据分解成不同的段。随后,使用ANN来预测负载的非线性数据。由AEMO获取一周每天24 h的数据验证了文中所设计模型的有效性。

At present, many researchers are attracted to power load forecasting, because it is a prerequisite for accurate planning, scheduling, operation and maintenance of power systems. As many factors affect the load forecasting, it is necessary to propose a hybrid model to improve the accuracy of the prediction. This paper presents a new hybrid load estimation scheme using two methods- Wavelet Transform （WT）and Artificial Neural Network（ANN）. In order to take large unsymmetrical time-varying electric power raw data set into account, the wavelet technology is used to decompose the data according to time and frequency. There are many wavelet functions that can be used, but the choice of a suitable wavelet function plays a key role in the design of this model. In this paper, we use the following types of wavelet functions, namely Haar wavelet function, Deubechies wavelet function, Symlet wavelet function and Coiflet wavelet function to decompose the electric load data into different segments. Subsequently, the ANN is used to predict the nonlinear data of the load. The validity of the model designed in this paper is verified by AEMO＇ s data of 24 hours a day.