Forecasting of daily load curve on monthly peak day using load research data and harmonics model

This work applies a harmonics model to forecast a daily load curve on a monthly peak day. The harmonics model is applied to capture periodic pattern of daily load and to avoid relying on weather-sensitive parameters. The harmonics model is a function of a base load, an hourly load, and a Fourier series. The dataset was obtained from load research data over the year 2008-2012 from the Provincial Electricity Authority of Thailand. It was found that the fifth harmonic model is proper to forecast the load curves on the monthly peak day. The forecasts of different tariff schedules are performed to compare the load patterns of residential, commercial, and industrial customers.