Predição de séries temporais no contexto de Smart Grids

Abstract

Smart grids emerge as the next technological breakthrough to be achieved for systems of generation, transmission and distribution of energy in the context of new sensing components, control, supervision and operation applicable to the Electric Power System (EPS). In the current context of regulated energy market and also in the likely long-term free market scenario, it is imperative that quality and operating costs are kept under control for greater competitiveness. This work turns to the study of forecasting methods in the scope of smart grids and the applicability of these methods in their intelligent behavior. As smart grids are in constant evolution within electric power systems, it is imperative to have a consistent framework for short-term prediction meeting operating characteristics. Traditional methods of time series prediction has been applied in EPS, such as SARIMA and SARFIMA. This latter faces seasonal time series that follow long memory processes. To improve the accuracy of these methods, hybrid methods will be developed integrating fuzzy logic to SARIMA and SARFIMA models. The proposed models are based on the technique of fuzzy time series (FTS). The model parameters are estimated by means of an evolutionary algorithm. The proposedmodels meet the need for methods that rely less on strong stationarity assumptions and are parsimonious in its parameters, even though it deals with stochastic long memory processes. In this work, algorithmic solutions will be proposed and analyzed, mainly in the form of hybridization SARFIMA methods and Fuzzy Time Series (SARFIMAFTS).To validate the proposed models it will be explored four databases to validate the proposed methodology. The first database consists of electric power demands of four profiles of customers of a large Brazilian distributor. Data were collected between 01/01/2003 and 01/12/2011. The second database refers to data demands collected in four European countries with hourly demands. Another base consists of nationalsystems hemi-time data from France and England collected in 2005. Finally, understanding that short-term forecasting plays a key role in energy management in a smart grid, the last base consists of information demand of four micro networks. The presented methodology is extensive as to be applied in various kinds of problems. The applicability of the proposed methodology in smart grids is very interesting because itwill allow performing calculations quickly on a reduced set of parameters.