Enhancing Neural Network Electricity Load Forecast with Wavelet Techniques  

Oleh:

Young Fei Low ; Dong ZhaoYang

Jenis: Article from Article
Dalam koleksi: Final Program and Book of Abstracts: The 4th Asian Control Conference, September 25-27, 2002 (Sep. 2002), page 203-208.
Topik: Neural Network; Wavelet; Time Series; Forecast; Short Term Load Forecast; Electricity Market
Fulltext: AC021748.PDF (550.15KB)

Isi artikel

This paper investigates an approach to enhance the capability of neural network based time series electricity demand forecast method. Wavelet decomposition is employed to explore the characteristics of demand time series, to extract more patterns from the series, and therefore enhance the forecast capability of neural network based models. The enhancement is carried out through a neural-wavelet model, which performs forecast on wavelet coefficients after decomposition, and reconstructs the overall forecast for higher accuracy. As typical time series, electricity load signals are studied in this paper. Real electricity load signals from NSW electricity market, as part of the Australian National Electricity Market (NEM), are tested to show the effectiveness of this forecast model. The model isimplemented in a Java platform in order to enable easy access and visualization through internet.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali