Wave Height Prediction Suitable for Maritime Transportation Based on Green Ocean of Things

Document Type


Publication Title

IEEE Transactions on Artificial Intelligence


Nowadays, the application fields of the Internet of Things (IoT) involve all aspects. This article combines ocean research with the IoT, in order to investigate the wave height prediction to assist ships to improve the economy and safety of maritime transportation and proposes an ocean IoT Green Ocean of Things (GOoT) with a green and low-carbon concept. In the wave height prediction, we apply a hybrid model (EMD-TCN) combining the temporal convolutional network (TCN) and the empirical mode decomposition (EMD) to the buoy observation data. We then compare it with TCN, long short-term memory (LSTM), and hybrid model EMD-LSTM. By testing the data of eight selected NDBC buoys distributed in different sea areas, the effectiveness of the EMD-TCN hybrid model in wave height prediction is verified. The hysteresis problem in previous wave height prediction research is eliminated, while improving the accuracy of the wave height prediction. In the 24 h, 36 h, and 48 h wave height prediction, the minimum mean absolute errors are 0.1265, 0.1689, and 0.1963, respectively; the maximum coefficient of determination are 0.9388, 0.9019, and 0.8712, respectively. In addition, in the short-term prediction, the EMD-TCN hybrid model also performs well, and has strong versatility.

First Page


Last Page




Publication Date



Empirical mode decomposition (EMD), Internet of Things (IoT), maritime transportation, temporal convolutional network (TCN), wave height prediction


IR Deposit conditions:

OA version (pathway a) Accepted version

No embargo

When accepted for publication, set statement to accompany deposit (see policy)

Must link to publisher version with DOI

Publisher copyright and source must be acknowledged