Probabilistic functional forecasting of Residual Demand Curves in electricity markets

Abstract

Residual Demand Curves (RDCs) in electricity markets represent the market-clearing price as a function of the energy traded. Being able to generate accurate scenarios of these curves is critical for market agents' strategic planning.For every hourly auction, one of these curves is observed, hence, forming a functional time series (FTS), i.e. a sequence of functions observed over time. Functional Data Analysis (FDA) has emerged as a methodology for exploring and modeling these kinds of data, known as functional data. [...]