Research suggests using neural networks to harness wind and solar energy

The ongoing transition from fossil fuels to renewable energy sources has never been more important as awareness of climate change and sustainability continues to grow.

This transition requires a solid information structure to ensure a smooth transition. Wind and solar energy are among the most abundant sources of renewable energy; however, engineers and researchers in this field need an established data pipeline to effectively integrate solar and wind energy into process design.

Abdullah Al-Aboosi is an interdisciplinary PhD student at the Department of Multidisciplinary Engineering. He is working with Dr. Aldo Jonathan Muñoz Vazquez, a multidisciplinary engineering professor at the McAllen Center for Higher Education, on a neural network that they hope will be able to provide such a pipeline.

The catalyst for this research was a discussion between Munoz Vazquez and Al-Aboosi. The idea evolved into a comprehensive project, drawing on the knowledge and experience of various collaborators, including Wei Zhan, Dr. Mahmoud El-Halwagi and Dr. Fadhil Al-Aboosi from the RAPID Manufacturing Institute for Process Intensification. The RAPID Manufacturing Institute provided an invaluable, publicly available source of accurate data that researchers used to test their model’s predictions.

As for the neural network itself, it can be used to provide an accurate overview of the performance of any renewable energy system and life cycle assessment by predicting daily and hourly wind speed and solar radiation intensity. By accurately predicting the availability of solar and wind energy, you can manage technology and supply resources more efficiently.

Al-Aboosi hopes that this project can better position renewable energy as the main source of electricity in the industrial sector. The project aims to enable researchers and renewable energy installation companies to determine the optimal number of solar panels and wind turbines needed to prevent over- or under-production. Such a reduction could encourage potential investors to adopt the technology and pursue a future with cleaner air and greener electricity.

This research was published in the journal Journal of Big Data and Cogitative Computing.