DOI QR코드

DOI QR Code

Power Consumption Prediction Scheme Based on Deep Learning for Powerline Communication Systems

전력선통신 시스템을 위한 딥 러닝 기반 전력량 예측 기법

  • Lee, Dong Gu (Dept. of Wireless Communications Engineering, KwangWoon University) ;
  • Kim, Soo Hyun (Dept. of Wireless Communications Engineering, KwangWoon University) ;
  • Jung, Ho Chul (Dept. of Wireless Communications Engineering, KwangWoon University) ;
  • Sun, Young Ghyu (Dept. of Wireless Communications Engineering, KwangWoon University) ;
  • Sim, Issac (Dept. of Wireless Communications Engineering, KwangWoon University) ;
  • Hwang, Yu Min (Dept. of Wireless Communications Engineering, KwangWoon University) ;
  • Kim, Jin Young (Dept. of Wireless Communications Engineering, KwangWoon University)
  • Received : 2018.09.10
  • Accepted : 2018.09.19
  • Published : 2018.09.30

Abstract

Recently, energy issues such as massive blackout due to increase in power consumption have been emerged, and it is necessary to improve the accuracy of prediction of power consumption as a solution for these problems. In this study, we investigate the difference between the actual power consumption and the predicted power consumption through the deep learning- based power consumption forecasting experiment, and the possibility of adjusting the power reserve ratio. In this paper, the prediction of the power consumption based on the deep learning can be used as a basis to reduce the power reserve ratio so as not to excessively produce extra power. The deep learning method used in this paper uses a learning model of long-short-term-memory (LSTM) structure that processes time series data. In the computer simulation, the generated power consumption data was learned, and the power consumption was predicted based on the learned model. We calculate the error between the actual and predicted power consumption amount, resulting in an error rate of 21.37%. Considering the recent power reserve ratio of 45.9%, it is possible to reduce the reserve ratio by 20% when applying the power consumption prediction algorithm proposed in this study.

최근 전력 사용량의 증가로 인한 대규모 블랙아웃 등 에너지 문제가 대두되고 있으며, 이 문제들로 인해 전력 소비량 예측에 대한 정확도를 개선할 필요성이 부각되었다. 본 연구에서는 딥 러닝 기반의 전력 사용량 예측 실험을 통해서 실제 전력 소비량과 예측된 전력 소비량의 차이를 계산하고, 이를 통해서 전력 예비율을 기존 대비 하향 조정할 수 있는 가능성에 대해서 살펴본다. 예비 전력은 사용하지 않으면 손실되는 전력으로, 본 논문에서의 딥 러닝 기반 전력 소비량 예측을 통해서 여분의 전력을 과도하게 생산하지 않도록 오차범위 내에서 전력 예비율을 감소시킬 수 있는 기반을 마련할 수 있다. 본 논문에서 사용하는 딥 러닝 기법은 시계열 데이터를 처리하는 Long-Short-Term-Memory(LSTM) 구조의 학습 모델을 이용한다. 컴퓨터 시뮬레이션에서는 임의 생성한 전력 소비 데이터를 토대로 모델을 학습시키고, 학습된 모델을 토대로 전력 사용 예측값을 구하고 실제 전력 소비량 간에 오차를 계산한 결과 오차율 21.37%를 얻을 수 있었다. 이는 최근의 전력 예비율 45.9%를 고려할 때, 본 연구에서 제안한 전력 소비량 예측 알고리즘을 적용하는 경우 20% 포인트 정도의 예비율 감축이 가능하다.

Keywords

References

  1. S. Kalyanaraman, "Back to the future: lessons for internet of energy networks," IEEE Internet Computing, vol.20, no.1, pp. 60-65, 2016. DOI:10.1109/MIC.2016.19
  2. Y. J. Lin, H. A. Latchman, M. Lee and S. Katar, "A power line communication network infrastructure for the smart home," IEEE Wireless Communications, vol.9, no.6, pp. 104-111, 2002. DOI:10.1109/MWC.2002.1160088
  3. K. B. Song, "Development of short-term load forecasting algorithm using hourly temperature," Journal of KIEE, vol.63, no.4, pp. 451-454, 2014. DOI:10.5370/KIEE.2014.63.4.451
  4. K. B. Song and S. K. Ha, "An algorithm of short-term load forecasting," Journal of KIEE, vol.53A, no.10, pp. 529-535, 2004.
  5. S. I. Kong, K. B. Song and Y. S. Baek, "The daily peak load forecasting in summer with the sensitivity of temperature," Journal of KIEE, vol.53A, no.10, pp. 358-363, 2004.
  6. S. B. Kong, M. S. Jang and S. K. Joo, "Short-term load forecasting for weekdays using a big data analysis," in Proc. of 2015 Spring Conference of KIEE, Seoul, pp. 144-145, 2015.
  7. H. N. Lee, J. H. Han and M. H. Lee, "Research of power peak estimation and prediction," Korea Energy Economic Review, vol.9, no.2, pp. 83-99, 2010.
  8. I. Goodfellow, Deep Learning, MIT Press, 2016.
  9. S. P. Kim, Deep Learning First Step, Hanbit media, 2016.
  10. K. Greff, R. K. Srivastava, J. Koutnik, B. R. Steunebrink and J. Schmidhuber, "LSTM: a search space odyssey," IEEE Trans. Neural Networks and Learning Systems, vol.28, no.10, pp. 2222-2232, 2017. DOI:10.1109/TNNLS.2016.2582924
  11. H. Salehinejad, S. Sankar, J. Barfett, E. Colak and S. Valaee, "Recent advances in recurrent neural networks," arXiv preprint arXiv:1801.01078, 2018.
  12. A. Krizhevsky, I. Sutskever, and G. E. Hinton, "Imagenet classification with deep convolutional neural networks," in Proc. of the 25th International Conference on Neural Information Processing Systems, Nevada, U.S.A., pp. 1097-1105, 2012.
  13. P. Sibi, S. A. Jones and P. Siddarth, "Analysis of different a activation function using back propagation neural networks," Journal of Theoretical and Applied Information Technology (JATIT), vol.47, no.3, pp. 1264-1268, 2013. DOI:10.1155/2012/490647
  14. Z. Wang, "Mean squared error: love it or leave it? a new look at signal fidelity Measures," IEEE Signal Processing Magazine, vol.26, no.1, pp. 98-117, 2009. DOI:10.1109/MSP.2008.930649