“21st PSCC 2020 papers submission and review platform

Full Program »

A Deep Learning Method for Forecasting Residual Market Curves

View File
PDF
0.9MB

Forecasts of residual demand curves (RDCs) are valuable information for price-maker market agents since it enables an assessment of their bidding strategy in the marketclearing price. This paper describes the application of deep learning techniques, namely long short-term memory (LSTM) network that combines past RDCs and exogenous variables (e.g., renewable energy forecasts). The main contribution is to build up on the idea of transforming the temporal sequence of RDCs into a sequence of images, avoiding any feature reduction and exploiting the capability of LSTM in handling image data. The proposed method was tested with data from the Iberian dayahead electricity market and outperformed machine learning models with an improvement of above 35% in both root mean square error and Fr`echet distance.

Author(s):

Alex Coronati    
INESC TEC
Portugal

José Ricardo Andrade    
INESC TEC
Portugal

Ricardo Bessa    
INESC TEC
Portugal

 

Powered by OpenConf®
Copyright ©2002-2014 Zakon Group LLC