Modeling and developing a smart interface for various drying methods of pomelo fruit (Citrus maxima) peel using machine learning approaches

Freeze-drying is a method used for valuable and heat-sensitive products, ensuring that the product features are best protected. In this study, the pomelo fruit (Citrus maxima) peel samples were dried in two different thicknesses (5 x 1 x 1 cm and 5 x 1 x 0.5 cm) by freeze drying (FD) as well as forced convection (FCD) and microwave drying. According to the experimental results, it was observed that the thin sample dried in a shorter time in all drying methods. Besides, the shortest drying time was seen in microwave drying method. The experimental results were modeled with artificial neural networks which is one of the machine learning approaches. Two different models were developed to predict drying parameters. The first model predicts mass dependent parameters (sample mass, moisture content, and moisture ratio). The second model was developed for drying time predictions. At the same time, it is intended that users can quickly and simply predict the specified parameters thanks to the smart interface developed.