Multi-GRU Prediction System for Electricity Generation's Planning and Operation

Weixian Li, Thillainathan Logenthiran, Wai Lok Woo

Research output: Contribution to journalArticlepeer-review

13 Citations (Scopus)

Abstract

Electricity generation's planning and operation have been key factors for any economic development in the power industries but it can only be achieved if the generation was accurately forecasted. This made forecasting systems essential to planning and operation in the electricity market. In this study, a novel system called multi-GRU (gated recurrent unit) prediction system was developed based on GRU models. It has four level of prediction process which consists of data collection and pre-processed module, multi-features input model, multi-GRU forecast model and mean absolute percentage error. The data collection and pre-processed module collect and reorganise the real-time data using the window method. Multi-features input model uses single input feeding method, double input feeding method, and multiple feeding method for features input to the multi-GRU forecast model. Multi-GRU forecast model integrates GRU variation such as regression model, regression with time steps model, memory between batches model, and stacked model to predict the future electricity generation and uses mean absolute percentage error to evaluate the prediction accuracy. The proposed systems achieved high accuracy prediction results for electricity generation.
Original languageEnglish
Pages (from-to)1630-1637
Number of pages8
JournalIET Generation, Transmission & Distribution
Volume13
Issue number9
Early online date15 Nov 2018
DOIs
Publication statusPublished - 7 May 2019

Fingerprint

Dive into the research topics of 'Multi-GRU Prediction System for Electricity Generation's Planning and Operation'. Together they form a unique fingerprint.

Cite this