The worldwide electric power industry has been moving into a scenario of market deregulation and the same has been happening in the Iberian Peninsula since 2007 when MIBEL - Iberian Electricity Market started and was opened to all consumers from Portugal and Spain in a free competitive market. Forecasting demand is essential for all market players because the market works one day-ahead and all sale and purchase orders of electricity have to be made until 11a.m. of the previous day of dispatch. However forecasting techniques applied in Electricity Markets are still in their early stages of maturity. The empirical analysis of electricity demand time series in a nonlinear framework is important to provide understanding of the process governing the underlying system, helping to distinguish between stochastic and chaotic processes and allowing an adequate choice of the forecasting techniques. The purpose of this paper is to characterize the chaotic behavior of the electricity demand time series using correlation dimension, entropy and maximal Lyapunov exponent tests and to predict the next 24h in the Iberian Electricity Market.

--

electricity market, nonlinear time series