Hybrid evolutionary-adaptive approach to predict electricity prices and wind power in the short-term

Abstract

Nowadays, with the new paradigm shift in the energy sector and the advent of the smart grid, or even with the mandatory imposition for a gradual reduction of greenhouse gas emissions, the renewable producers, namely the wind power producers are faced with the competitiveness and deregulated structure that characterizes the liberalized electricity market. In a liberalized electricity market, the most important signal for all market players corresponds to the electricity prices. In this sense, accurate approaches for short-term electricity prices prediction are needed, and also for short-term wind power prediction due to the increasing share of wind generation. Hence, this paper presents a new hybrid evolutionary-adaptive approach for wind power and electricity market prices prediction, in the short-term, based on mutual information, wavelet transform, evolutionary particle swarm optimization and adaptive neuro-fuzzy inference system, tested on real case studies, proving its superiority in a comprehensive comparison with other approaches previously published in the scientific literature.