小波变换与PAR方法在负荷短期预测中的应用

摘要

小波变换具有很多良好的性质,它能任意提取负荷序列的细节.针对负荷序列以天、周、年为周期的变化特点,引用小波变换算法对负荷进行了不同频段上的分解,然后利用能够准确描述数据周期性变化的PAR模型对负荷进行短期预测,建立了基于小波变换的PAR模型.实验证明,这种预测方法非常有效.%Wavelet transform has many advantages, and it can extract any details of load sequence. In according to the characteristics of Load sequences changes by the period of days, weeks, years, using wavelet transform algorithm on the decomposition of different frequency ranges, and then established the PAR model based on the wavelet transform which can describe the changes of data periodicity accurately, to forecast the load in a short time. Experiments show that this prediction method is very effective.