MATLAB-Based Forecasting Source Code Using ARIMA Models
January 2022
Introduction Aims Applications
Making informed planning and ESPO aims to develop a software The software package is designed as a operation decisions in energy systems package that allows the practitioners source code in the MATLAB needs an accurate uncertainty model of and researchers in energy sector to environment, and can be used to forecast future realizations of parameters. The provide forecasts using the ARIMA uncertain data such as energy uncertainties can be forecasted from models. consumption, energy price, renewable short-term (e.g., hours to days) to long- energy, and climatic indicators. Its To this end, the main features of the term (e.g., months to years) periods applications can be different depending on software package are: depending on the problem under study. the time horizon (i.e., from short-term to Among different forecasting tools, To receive time series data and long-term) and the uncertainty model autoregressive integrated moving ARIMA model (p,d,q)(P,D,Q)S as chosen by the user. The single-point average (ARIMA) is a univariate model input arguments. forecasts can be incorporated into a and recognized as a practical method for deterministic model to make informed To implement the mathematical forecasting uncertain parameters in decisions. Additionally, the scenarios and structure of ARIMA model. energy sector. Especially, practitioners confidence bounds can be used in and researchers in energy sector needs a To provide forecasts under different stochastic programming and robust software package that can easily uncertainty models: i) single-point optimization models, respectively. The implement the ARIMA models and forecasts, ii) scenarios, and iii) software package also allows the user to provide appropriate forecasts as well. confidence bounds. adjust the number of scenarios and the confidence level of forecasts bounds.