ABSTRACT
This chapter focuses on electricity price forecast in the uniform pricing structure, which is the most commonly accepted structure of electricity markets around the world. Importance and complexity of electricity price forecast motivate many research works in the area. In general, electricity price forecast methods can be divided into two main categories. Methods of the first category try to directly predict electricity price by analyzing the electricity market dynamics and effective parameters on the market price, such as production costs and strategic behavior of market participants. Electricity price is a nonlinear function of many candidate inputs including its past values as well as past and forecast values of the exogenous variables. The electricity price signal has a short-run trend characteristic as well as daily periodicity behavior and weekly periodicity behavior. The hybrid data models contain the information content of both time- and frequency-domain candidate features for the electricity price forecast process.
