Open AccessDiscrete cosine transform‐based predictive model extended in the least‐squares sense for hourly load forecasting
Author(s) -
Yang ZongChang
Publication year - 2016
Publication title -
iet generation, transmission and distribution
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.92
H-Index - 110
eISSN - 1751-8695
pISSN - 1751-8687
DOI - 10.1049/iet-gtd.2016.0689
Subject(s) - discrete cosine transform , decorrelation , computer science , algorithm , least squares function approximation , sine and cosine transforms , discrete fourier transform (general) , mathematics , control theory (sociology) , fourier transform , statistics , artificial intelligence , fractional fourier transform , image (mathematics) , fourier analysis , estimator , mathematical analysis , control (management)
Electric load forecasting is essential to economic operation and power system control in the electric power industry. Modelling and forecasts for different time horizons are of increasingly importance for different operations. Discrete cosine transform (DCT) is one Fourier‐related computational technique. It is widely used in signal and image processing for its remarkable characteristics on optimal decorrelation and energy compaction. From time‐series analysis perspective, this study proposes a DCT‐based predictive model for forecasting hourly load movement. Based on finite hourly load observations, the proposed DCT‐based predictive model combines with the least‐squares approach to get the optimum DCT coefficients, which are approximated in the least‐squares sense. Then the obtained least‐squares‐optimum DCT coefficients are employed for forecast modelling to predict the load future movement. Experimental results and analysis show workability of the proposed forecast modelling. It is indicated that the DCT‐based least‐squares predictive model predicts hourly load movement most fitting with about 12‐term DCT coefficients.