Journal of IKEEE (전기전자학회논문지)

Institute of Korean Electrical and Electronics Engineers (한국전기전자학회)
권호 표지
DOI QR코드

DOI QR Code

Mid-Term Energy Demand Forecasting Using Conditional Restricted Boltzmann Machine

조건적 제한된 볼츠만머신을 이용한 중기 전력 수요 예측

  • Kim, Soo-Hyun (Dept. of Electronic Convergence Engineering, Kwangwoon University) ;
  • Sun, Young-Ghyu (Dept. of Electronic Convergence Engineering, Kwangwoon University) ;
  • Lee, Dong-gu (Dept. of Electronic Convergence Engineering, Kwangwoon University) ;
  • Sim, Is-sac (Dept. of Electronic Convergence Engineering, Kwangwoon University) ;
  • Hwang, Yu-Min (Dept. of Electronic Convergence Engineering, Kwangwoon University) ;
  • Kim, Hyun-Soo (Co. Gridwiz) ;
  • Kim, Hyung-suk (Dept. of Electronic Convergence Engineering, Kwangwoon University) ;
  • Kim, Jin-Young (Dept. of Electronic Convergence Engineering, Kwangwoon University)
  • Received : 2019.03.08
  • Accepted : 2019.03.25
  • Published : 2019.03.31
Download PDF
⟨ Previous Next ⟩

Abstract

Electric power demand forecasting is one of the important research areas for future smart grid introduction. However, It is difficult to predict because it is affected by many external factors. Traditional methods of forecasting power demand have been limited in making accurate prediction because they use raw power data. In this paper, a probability-based CRBM is proposed to solve the problem of electric power demand prediction using raw power data. The stochastic model is suitable to capture the probabilistic characteristics of electric power data. In order to compare the mid-term power demand forecasting performance of the proposed model, we compared the performance with Recurrent Neural Network(RNN). Performance comparison using electric power data provided by the University of Massachusetts showed that the proposed algorithm results in better performance in mid-term energy demand forecasting.

미래에 스마트 그리드 도입을 위해 전력수요예측은 중요한 연구 분야 중 하나이다. 하지만 전력데이터는 많은 외부적 요소들에 영향을 받기 때문에 예측하기 어렵다. 기존의 전력수요예측 방법들은 가공되지 않은 전력데이터를 그대로 이용하기 때문에 정확도 높은 예측을 하는데 한계가 있어왔다. 본 논문에서는 가공되지 않은 전력데이터를 이용하는 전력수요예측의 문제를 해결하기 위해 확률기반 학습알고리즘을 제안한다. 확률 모델은 전력데이터의 확률적 특성을 분석하기에 적합하다. 제안한 모델의 중기 전력수요예측 성능을 비교하기 위해 신경망 네트워크 중 하나인 순환신경망과 성능 비교를 해보았다. 매사추세츠 대학에서 제공한 전력데이터를 이용하여 성능 비교를 한 결과 본 논문에서 제안한 확률기반 학습알고리즘이 중기 수요예측에 더 좋은 성능을 나타냄을 확인하였다.

Keywords

JGGJB@_2019_v23n1_127_f0001.png 이미지

Fig. 1. Structure of Boltzmann Machine and Restricted Boltzmann Machine. 그림 1. 볼츠만머신과 제한된 볼츠만머신 구조

JGGJB@_2019_v23n1_127_f0002.png 이미지

Fig. 2. Structure of CRBM. 그림 2. 조건적 제한된 볼츠만머신 구조

JGGJB@_2019_v23n1_127_f0003.png 이미지

Fig. 3. Training data for learning model. 그림 3. 모델 학습을 위한 훈련 데이터

JGGJB@_2019_v23n1_127_f0004.png 이미지

Fig. 4. Validation data for performance evaluation. 그림 4. 성능 평가를 확인하기 위한 검증 데이터

JGGJB@_2019_v23n1_127_f0005.png 이미지

Fig. 5. Simulation results using Recurrent Neural Network. 그림 5. RNN을 이용한 시뮬레이션 결과

JGGJB@_2019_v23n1_127_f0006.png 이미지

Fig. 6. Simulation results using Conditional Restricted Boltzmann Machine. 그림 6. CRBM을 이용한 시뮬레이션 결과

Table 1. Parameters of experiments. 표 1. 실험 파라미터

JGGJB@_2019_v23n1_127_t0001.png 이미지

Table 2. Performance indicator of models. 표 2. 모델 성능 지표

JGGJB@_2019_v23n1_127_t0002.png 이미지

References

  1. Y. Fu, D. Sun, Y. Wang, L. Feng and W. Zhao, "Multi-level load forecasting system based on power grid planning platform with integrated information," in Proc. of 2017 Chinese Automation Congress(CAC), IEEE, Jinan, China, pp. 933-938, 2017. DOI: 10.1109/CAC.2017.8242900
  2. D. Zhang, Y. Yan, X. Li, X. Ren, J. Zhang and F. Zhang, "Mid-long term electricity demand forecasting based on markov chain screening combination forecasting models," Power System Protection and Control, vol. 44, no. 12, pp. 63-67, 2016. DOI: 10.1109/CICED.2016.7576282
  3. I. E. Kafazi, R. Bannari and A. Abouabdellah, "Modeling and forecasting energy demand," in Proc. of 2016 International Renewable and Sustainable Energy Conference (IRSEC), IEEE, Marrakech, Morocco, pp. 746-750, 2016. DOI: 10.1109/IRSEC.2016.7983974
  4. L. Wei and S. Yumin, "Prediction of energy production and energy consumption based on BP neural networks," in Proc. of 2008 IEEE International Symposium on Knowledge Acquisition and Modeling Workshop, Wuhan, China, pp. 15-19, 2008. DOI: 10.1109/KAMW.2008.4810454
  5. M. de Oliveira, "The influence of ARIMAGARCH parameters in feed forward neural networks prediction," Neural computing & applications, vol. 20, no. 5, pp. 687-701, 2011. DOI: 10.1007/s00521-010-0410-8
  6. UMass website; Available at http://traces.cs.umass.edu/
  7. A. Fischer and C. Igel, An introduction to restricted boltzmann machines. Springer, 2012.
  8. G. W. Taylor, G. E. Hinton and S. T. Roweis, "Modeling human motion using binary latent variables," in Proc. of Twentieth Annual Conference on Neural Information Processing Systems, Vancouver, Canada, pp. 1345-1352, 2006.
  9. G. E. Hinton, A practical guide to training restricted boltzmann machines. Springer, 2012.
  10. G. E. Hinton, "Training products of experts by minimizing contrastive divergence," Neural Computation, vol. 14, no. 8, pp. 1771-1800, 2002. DOI: 10.1162/089976602760128018
  11. S. Wang, Y. Liu and X. Zhang, "A differentiated DBN model based on CRBM for time series forecasting," in Proc. of 2017 IEEE 17th International Conference on Communication Technology (ICCT), pp. 1926-1931, 2017. DOI: 10.1109/ICCT.2017.8359965
  12. X. Cai and X. Lin, "Forecasting high dimensional volatility using conditional restricted boltzmann machine on GPU," in Proc. of Parallel and Distributed Processing Symposium Workshops & PhD Forum (IPDPSW), pp. 1979-1986, 2012. DOI: 10.1109/IPDPSW.2012.258