Iranian Researchers Find Innovative AI Model for Precise Prediction of Sea Wave Height

The model is designed in the form of an intelligent graphical user interface based on artificial intelligence (AI) so that engineers and planners in the field of renewable energy can easily use it to predict sea conditions. The main goal of this achievement is to support the operation and planning of wave energy converters and facilitate decision-making in projects related to marine energy.

“One of the most important obstacles on this path is the strong need of conventional machine learning models for a huge amount of historical data; data that is not only costly and time-consuming to collect in marine environments, but is also practically impossible in many coastal areas of the world, specially in developing countries,” said Amir Hossein Shahbazbegian, an undergraduate student in marine engineering.

“Our research was a direct response to this challenge. In this project, an optimized machine learning model was developed; an advanced and intelligent model based on marine artificial intelligence that has a special ability to learn from small and sparse data,” he added.

“We trained the model with data from four marine buoys located in different parts of the Australian coast, different areas that were deliberately chosen to cover a wide range of climatic conditions and wave patterns,” Shahbazbegian explained, adding, “The model could predict waves with an accuracy of between 93% and 97%, outperforming many conventional machine learning algorithms.”

In relevant remarks in 2024, the first directional wave recorder was built in the National Research Institute of Oceanography and Atmospheric Sciences of Iran.

“Knowing various phenomena, including the weather, the wave, the current regime and other natural phenomena requires field data collection with access to up-to-date and accurate equipment; therefore, for this purpose, the first directional wave recorder was designed and built with the aim of understanding the wave regime of marine and oceanic environments,” said Seyed Massoud Mahmoudov, a member of the National Research Institute of Oceanography and the executive of the first directional wave recorder project.

Using this device, we will be able to extract the two-dimensional spectrum of waves. This device has gone through its initial testing period and has reached a stage where it is going to be launched in the sea for the first time and the main tests will be carried out in the Southern coasts of the country and the Caspian Sea,” he added.

Mahmoudov explained that the device was made in collaboration with the University of Science and Technology in Tehran.

The basic principles of wave measurement rely on the understanding that a wave is not a single defined entity, but the result of a series of individual waveforms superimposed on top of each other, all with different wavelengths, frequencies and amplitudes. Measurement of the wave activity therefore requires measurement of the pressure (and current) variations for a period of time, then “decomposing” the pattern into the constituent waveforms before analysing them and interpreting the data as a set.

The directional wave recorder calculates the direction from which the waves are coming by measuring the current oscillations caused by wave motion.

4155/v