Deep learning on Wavelets Analysis with Adam to predict Electricity load

Authors

  • Vuong Minh Chau Hanoi University of Science, Vietnam National University, Hanoi, Vietnam
  • Do Viet Bach Hanoi University of Science, Vietnam National University, Hanoi, Vietnam
  • Nguyen Quang Dat Hanoi University of Science, Vietnam National University, Hanoi, Vietnam
  • Vu Duc Hoan The Air Force officer’s College
  • Duong Xuan Bien Le Quy Don Technical University, Hanoi, Vietnam

Keywords:

Time series forecasting, GRU, LSTM, Wavelets, Adam

Abstract

Hanoi is the biggest economiccentrein Northern Vietnam with a population of nearly 10 million people. Its power supply system is supported and developed by the government, as part of the country's electrical distribution system. Because of this,forecasting Hanoi's electricity load is vital to improving the citizens' lives, especially when the power supply of Northern Vietnam is barely enough. In this research, we are proposing a new model that uses Wavelets along with Deep learning, with Adam as optimization in order to replace outdated manual statistical methods. Our model shows higher accuracy (best case is higher than 1.22%) when compared to the traditional methods of GRU and LSTM (without and with Adam optimization).

Downloads

Published

2023-04-28

How to Cite

Chau, V. M., Bach, D. V., Dat, N. Q., Hoan, V. D., & Bien, D. X. (2023). Deep learning on Wavelets Analysis with Adam to predict Electricity load. Proceedings of International Conference on Intelligent Systems and New Applications, 1, 14–18. Retrieved from https://proceedings.icisna.org/conf/index.php/ICISNA/article/view/85

Issue

Vol. 1 (2023): International Conference on Intelligent Systems and New Applications (ICISNA'23)

Section

Proceedings Paper