一种基于小波变换和ARIMA的短期电价混合预测模型

为在实时电价情况下预测未来24小时电价, 提出一种基于小波变换和差分自回归移动平均(ARIMA)的短期电价混合预测模型。该模型分别根据是否受到需求量影响使用ARIMA模型对多尺度小波变换分解后的时间序列进行预测。同时提出一种电价突变点发现和处理算法。使用澳大利亚新南威尔士州2012年真实数据验证表明, 相对ARIMA预测, 改进后的混合模型在不考虑需求量影响时预测精度更高; 电价突变点发现和处理算法能够准确处理电价异常点, 提高预测精度。

Hybrid model based on wavelet transform and ARIMA for short-term electricity price forecasting

In order to forecast the next 24 hour's electricity price, this paper proposed a hybrid model based on wavelet transform and autoregressive integrated moving average(ARIMA) for short-term electricity price forecasting. According to whether considering the influence of demands, it used ARIMA to forecast the time series depcomposed by wavelet transform. It proposed an electricity price anomalies detect and process algorithm to handle the condition where price changed drastically. The numerial example based on the historical data of the Australian national electricity market, New South Wales, in the year 2012, shows that the hybrid model, not considering the influence of demands, got a more precise result than the ARIMA mo-del; electricity price anomalies detect and process algorithm can process the electricity price anomalies precisely and improve the predict accuracy.