Analisis Karakteristik Pengeringan Kunyit Menggunakan Menggunakan Mesin Pengering Sistem Rotasi Berbasis ANN-FL

Abstract

Turmeric is one of the spices that has many health benefits so that it is widely consumed by humans, both for cooking spices, health therapy, and traditional cosmetics. Therefore, drying turmeric is one way to preserve turmeric so that it can be used for a long time. The purpose of this study was to analyze the properties or characteristics of turmeric drying using an automatic rotary drying machine. The Artificial Neural Network-Fuzzy Logic (ANN-FL) method was developed to model the relationship between drying parameters such as turmeric silage slice thickness, drying temperature, air flow rate, and drying time to the remaining water content and air humidity ratio. The first experiment showed a remaining water content of 10% from the initial water content of 75% with a final humidity of 40%. The second and third experiments showed water content to be 8% and 5% respectively from the initial water content of 80 and 85% with humidity of 30% and 15%. Artificial Neural Network (ANN) and Fuzzy Logic (FL) were modeled to learn patterns from experimental data obtained through experiments. The histogram of the model error shows that the distribution of errors between predictions and targets. Most of the errors are close to zero, indicating that the ANN-FL model has an accurate prediction rate. In addition, the training performance shows a decrease in Mean Squared Error (MSE) as the number of epochs increases, which means that the ANN-FL model learns well and reduces the prediction error as the number of epochs increases. The results of the study indicate that the ANN-FL method can be used to predict the water content of turmeric with a high level of accuracy. In addition, the ANN-FL method can also be used to control the temperature and humidity of turmeric drying so that more uniform and high-quality drying results are obtained.