Short-Term Power Load Forecasting of a Large Vessel using Deep Stacking Network Architecture

Chang Woo Hong, Min Seung Ko, Hong Ryeol Kim, Soyeon Kim, Kyeon Hur

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)


The power load prediction in vessel is an important factor in determining the capacity and number of generators, and in particular the consumption of fuel oil which determines the number of days that can be sailed. In addition, short-term load forecasting is important for the capacity and scheduling of the ESS that will be applied in the future vessel. In this paper, we present a deep stack neural network for short-term load prediction in large vessels. The network is constructed using Convolutional Neural Network (CNN), Bidirectional Long-Short Term Memory (Bi-LSTM), and Long-Short Term Memory (LSTM). CNN is used for spatial feature extraction and Bi-LSTM is used to utilize information at both pre and post stages. Finally, LSTM is used to extract temporal characteristics. The voyage data of the Mokpo National Maritime University training ship was used for the short-term load prediction, and the predicted results are verified by the Mean Squared Error (MSE) and Mean Absolute Error (MAE).

Original languageEnglish
Pages (from-to)534-541
Number of pages8
JournalTransactions of the Korean Institute of Electrical Engineers
Issue number4
Publication statusPublished - 2020 Apr

Bibliographical note

Funding Information:
This research was supported by Korea Electric Power Corporation (Grant number: R17XA05-4). This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (Grant number: 2018R1D1A1A09083054).

Publisher Copyright:
Copyright © The Korean Institute of Electrical Engineers

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering


Dive into the research topics of 'Short-Term Power Load Forecasting of a Large Vessel using Deep Stacking Network Architecture'. Together they form a unique fingerprint.

Cite this