Signal processing of stochastic forecast: seasonal electric power consumption in the scope of efficiency connected to renewable energy

Abstract

The electric power system's seasonality would mitigate the problem of the energy volatility crisis. Seasonal identification is important in energy usage analysis to make a better an organization's energy performance. Identification of the seasonal component in a time series data is essential since it defines a fluctuation within an interval of time. This seasonal component can be identified as a periodic phenomenon. Signal processing models can be used in analyzing periodic data in the representation of sinusoidal function. Seasonal variations are an important factor influencing the behavior of an electric power and energy load. This paper introduces the application of deterministic functions in the Ornstein\textendash Uhlenbeck process as a sinusoidal signal processing model in identifying seasonal components represented in periodic terms. We conducted extensive experiments to validate the model on three datasets in electric power and energy systems namely electricity load, household electric power consumption, and power consumption of the three source stations. Further, it can be obtained that the periodic continuous-time-inhomogeneous signal of Ornstein\textendash Uhlenbeck model can be used to identify the seasonal term of the very short to medium horizon forecast of the electric power and energy system. Seasonal changes in energy consumption can be used in energy management in the scope of energy efficiency connected to renewable energy plans.