Recurrent inception convolution neural network for multi short-term load forecasting

Junhong Kim, Jihoon Moon, Een Jun Hwang, Pilsung Kang

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Smart grid and microgrid technology based on energy storage systems (ESS) and renewable energy are attracting significant attention in addressing the challenges associated with climate change and energy crises. In particular, building an accurate short-term load forecasting (STLF) model for energy management systems (EMS) is a key factor in the successful formulation of an appropriate energy management strategy. Recent recurrent neural network (RNN)-based models have demonstrated favorable performance in electric load forecasting. However, when forecasting electric load at a specific time, existing RNN-based forecasting models neither use a predicted future hidden state vector nor the fully available past information. Therefore, once a hidden state vector has been incorrectly generated at a specific prediction time, it cannot be corrected for enhanced forecasting of the following prediction times. To address these problems, we propose a recurrent inception convolution neural network (RICNN) that combines RNN and 1-dimensional CNN (1-D CNN). We use the 1-D convolution inception module to calibrate the prediction time and the hidden state vector values calculated from nearby time steps. By doing so, the inception module generates an optimized network via the prediction time generated in the RNN and the nearby hidden state vectors. The proposed RICNN model has been verified in terms of the power usage data of three large distribution complexes in South Korea. Experimental results demonstrate that the RICNN model outperforms the benchmarked multi-layer perception, RNN, and 1-D CNN in daily electric load forecasting (48-time steps with an interval of 30 min).

Original languageEnglish
Pages (from-to)328-341
Number of pages14
JournalEnergy and Buildings
Volume194
DOIs
Publication statusPublished - 2019 Jul 1

Fingerprint

Convolution
Recurrent neural networks
Electric load forecasting
Neural networks
Energy management systems
Energy management
Climate change
Energy storage

Keywords

  • Convolution neural network
  • Deep learning
  • Load forecasting
  • Recurrent inception convolution neural network
  • Recurrent neural network

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Building and Construction
  • Mechanical Engineering
  • Electrical and Electronic Engineering

Cite this

Recurrent inception convolution neural network for multi short-term load forecasting. / Kim, Junhong; Moon, Jihoon; Hwang, Een Jun; Kang, Pilsung.

In: Energy and Buildings, Vol. 194, 01.07.2019, p. 328-341.

Research output: Contribution to journalArticle

@article{99de3f7cb4a64ffaaaf7d7b7859d2431,
title = "Recurrent inception convolution neural network for multi short-term load forecasting",
abstract = "Smart grid and microgrid technology based on energy storage systems (ESS) and renewable energy are attracting significant attention in addressing the challenges associated with climate change and energy crises. In particular, building an accurate short-term load forecasting (STLF) model for energy management systems (EMS) is a key factor in the successful formulation of an appropriate energy management strategy. Recent recurrent neural network (RNN)-based models have demonstrated favorable performance in electric load forecasting. However, when forecasting electric load at a specific time, existing RNN-based forecasting models neither use a predicted future hidden state vector nor the fully available past information. Therefore, once a hidden state vector has been incorrectly generated at a specific prediction time, it cannot be corrected for enhanced forecasting of the following prediction times. To address these problems, we propose a recurrent inception convolution neural network (RICNN) that combines RNN and 1-dimensional CNN (1-D CNN). We use the 1-D convolution inception module to calibrate the prediction time and the hidden state vector values calculated from nearby time steps. By doing so, the inception module generates an optimized network via the prediction time generated in the RNN and the nearby hidden state vectors. The proposed RICNN model has been verified in terms of the power usage data of three large distribution complexes in South Korea. Experimental results demonstrate that the RICNN model outperforms the benchmarked multi-layer perception, RNN, and 1-D CNN in daily electric load forecasting (48-time steps with an interval of 30 min).",
keywords = "Convolution neural network, Deep learning, Load forecasting, Recurrent inception convolution neural network, Recurrent neural network",
author = "Junhong Kim and Jihoon Moon and Hwang, {Een Jun} and Pilsung Kang",
year = "2019",
month = "7",
day = "1",
doi = "10.1016/j.enbuild.2019.04.034",
language = "English",
volume = "194",
pages = "328--341",
journal = "Energy and Buildings",
issn = "0378-7788",
publisher = "Elsevier BV",

}

TY - JOUR

T1 - Recurrent inception convolution neural network for multi short-term load forecasting

AU - Kim, Junhong

AU - Moon, Jihoon

AU - Hwang, Een Jun

AU - Kang, Pilsung

PY - 2019/7/1

Y1 - 2019/7/1

N2 - Smart grid and microgrid technology based on energy storage systems (ESS) and renewable energy are attracting significant attention in addressing the challenges associated with climate change and energy crises. In particular, building an accurate short-term load forecasting (STLF) model for energy management systems (EMS) is a key factor in the successful formulation of an appropriate energy management strategy. Recent recurrent neural network (RNN)-based models have demonstrated favorable performance in electric load forecasting. However, when forecasting electric load at a specific time, existing RNN-based forecasting models neither use a predicted future hidden state vector nor the fully available past information. Therefore, once a hidden state vector has been incorrectly generated at a specific prediction time, it cannot be corrected for enhanced forecasting of the following prediction times. To address these problems, we propose a recurrent inception convolution neural network (RICNN) that combines RNN and 1-dimensional CNN (1-D CNN). We use the 1-D convolution inception module to calibrate the prediction time and the hidden state vector values calculated from nearby time steps. By doing so, the inception module generates an optimized network via the prediction time generated in the RNN and the nearby hidden state vectors. The proposed RICNN model has been verified in terms of the power usage data of three large distribution complexes in South Korea. Experimental results demonstrate that the RICNN model outperforms the benchmarked multi-layer perception, RNN, and 1-D CNN in daily electric load forecasting (48-time steps with an interval of 30 min).

AB - Smart grid and microgrid technology based on energy storage systems (ESS) and renewable energy are attracting significant attention in addressing the challenges associated with climate change and energy crises. In particular, building an accurate short-term load forecasting (STLF) model for energy management systems (EMS) is a key factor in the successful formulation of an appropriate energy management strategy. Recent recurrent neural network (RNN)-based models have demonstrated favorable performance in electric load forecasting. However, when forecasting electric load at a specific time, existing RNN-based forecasting models neither use a predicted future hidden state vector nor the fully available past information. Therefore, once a hidden state vector has been incorrectly generated at a specific prediction time, it cannot be corrected for enhanced forecasting of the following prediction times. To address these problems, we propose a recurrent inception convolution neural network (RICNN) that combines RNN and 1-dimensional CNN (1-D CNN). We use the 1-D convolution inception module to calibrate the prediction time and the hidden state vector values calculated from nearby time steps. By doing so, the inception module generates an optimized network via the prediction time generated in the RNN and the nearby hidden state vectors. The proposed RICNN model has been verified in terms of the power usage data of three large distribution complexes in South Korea. Experimental results demonstrate that the RICNN model outperforms the benchmarked multi-layer perception, RNN, and 1-D CNN in daily electric load forecasting (48-time steps with an interval of 30 min).

KW - Convolution neural network

KW - Deep learning

KW - Load forecasting

KW - Recurrent inception convolution neural network

KW - Recurrent neural network

UR - http://www.scopus.com/inward/record.url?scp=85064953044&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85064953044&partnerID=8YFLogxK

U2 - 10.1016/j.enbuild.2019.04.034

DO - 10.1016/j.enbuild.2019.04.034

M3 - Article

AN - SCOPUS:85064953044

VL - 194

SP - 328

EP - 341

JO - Energy and Buildings

JF - Energy and Buildings

SN - 0378-7788

ER -