Thanks to the GREEN-LOOP project activities, IDENER’s researchers have successfully applied artificial intelligence (AI) to predict temperature profiles in microwave heating systems, according to a new study published in the open-access journal Thermo. The findings, led by Victor Rosario Núñez, could lead to more efficient and uniform heating in various applications, from food processing to industrial material curing.
“Microwaves offer a rapid and efficient heating method,” explains Victor Rosario Núñez, “[but] achieving a uniform temperature distribution can be challenging.” This new research addresses this challenge by using convolutional neural networks (CNNs), a type of AI algorithm adept at image recognition, to analyze data from microwave systems and predict the resulting temperature profile.
In the study, the researchers trained a CNN model on data collected from a microwave heating system. The model learned to identify patterns in the data that correlated with specific temperature distributions. Once trained, the model could then predict the temperature profile for new heating scenarios.
“This approach has the potential to revolutionize microwave heating technology,” says Victor. “By accurately predicting temperature profiles, we can optimize heating processes to achieve more uniform heating and improve efficiency.”
The findings have implications for a wide range of applications using microwave heating, with the following benefits:
- More uniform heating can improve material quality and safety.
- Precise temperature control can enhance material properties.
- Targeted heating can be used to increase the selectivity of the heat transfer effects.
About the Article and the Researchers
The article “Artificial Intelligence Applied to Microwave Heating Systems: Prediction of Temperature Profile through Convolutional Neural Networks” is a collaborative work by Victor Rosario Núñez, Alfonso Hernández, Iván Rodríguez, Ignacio Fernández-Pacheco Ruiz and Luis Enrique Acevedo Galicia, from IDENER Research and Development – Sevilla (SPAIN).
About the Journal
Thermo is an open-access journal that publishes high-quality research on all aspects of thermal science and engineering.
Here is the link to the article https://www.mdpi.com/2899052