Short-Term Demand Forecast Using Fourier Series

This paper presents a model to forecast load in short-term (24 h) aiming to support decision-making during contingencies in power distribution systems. The methodology is based on the calculation of both, coefficients and angles of the Fourier series in order to characterize the load curve. These parameters are fitted using historical load reports. The proposed methodology was performed using the data of two power transformers to show the behaviour of the demand curve for weekdays as well as for weekends. The quadratic mean errors obtained using the proposed algorithm and according to the computational efficiency in the fitting of the coefficients, it is possible implement the algorithm in the power systems control and operation planning.