Daily Load Forecasting Based on a Combination of Classification and Regression Tree and Deep Belief Network
Open Access
- 10 November 2021
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 152226-152242
Abstract
The next-day load forecasting is complex due to the load pattern variations driven by external factors, such as weather and time. This study proposes a hybrid model that incorporates the Classification and Regression Tree (CART) with pruning conditions and a Deep Belief Network (DBN) to improve forecasting accuracy. The CART can recognize the load patterns by classifying similar groups with low variance, thus reducing the complexity of the forecasting model. The actual 48-period load data from the Electricity Generating Authority of Thailand (EGAT) is used. The proposed model is compared with six widely used standalone forecasting benchmark models and provides better at the minimum 0.46% mean absolute percentage error. Moreover, the forecasting performance of DBN and the other four benchmark models are improved by using our hybrid approach.Keywords
Funding Information
- Center of Excellence in Logistics and Supply Chain Systems Engineering and Technology (COE LogEn), Sirindhorn International Institute of Technology (SIIT), Thammasat University (TU), Thailand
This publication has 42 references indexed in Scilit:
- A new method for pattern recognition in load profiles to support decision-making in the management of the electric sectorInternational Journal of Electrical Power & Energy Systems, 2013
- A new cascade NN based method to short-term load forecast in deregulated electricity marketEnergy Conversion and Management, 2013
- Classification and regression treesWIREs Data Mining and Knowledge Discovery, 2011
- Functional clustering and linear regression for peak load forecastingInternational Journal of Forecasting, 2010
- An Optimal Constrained Pruning Strategy for Decision TreesINFORMS Journal on Computing, 2009
- Electricity customer classification using frequency–domain load pattern dataInternational Journal of Electrical Power & Energy Systems, 2006
- Model Selection for CART Regression TreesIEEE Transactions on Information Theory, 2005
- New algorithms for learning and pruning oblique decision treesIEEE Transactions on Systems, Man and Cybernetics, Part C (Applications and Reviews), 1999
- Long Short-Term MemoryNeural Computation, 1997
- A performance comparison of trained multilayer perceptrons and trained classification treesProceedings of the IEEE, 1990