Study on drying of bitter gourd slices based on halogen dryer

Abstract

In this study, the drying of bitter gourd slices with a halogen dryer was done at different thicknesses of bitter gourd (3, 5, and 7 mm) and temperatures (60, 65, and 70 °C). The effect of varying drying characteristics in the experiment was explored. Experimental results were evaluated based on the drying time and moisture content. The results indicate that the material drying thickness and drying temperature significantly impact the drying time and the equilibrium moisture content. Furthermore, the Multivariate Adaptive Regression Splines (MARS) model is also used to train and predict the moisture content of bitter gourd in this research. The temperature, thickness of the bitter gourd, and drying time were used as input parameters for the model. Two measures R2 and Root Mean Ssquare Error (RMSE) were used to determine the accuracy of the trained MARS model. During training, the values of R2 and RMSE obtained were 0.9846 and 3.7324, respectively. The test of trained MARS was successful, with a satisfactory correlation between experimental data points and predicted points. The results show that MARS can accurately predict the moisture content of bitter gourd in a halogen dryer.

Keywords

• ann
• drying temperature
• machine learning
• mars model
• moisture content