15/8/2025 #NOPUBLISH# Klasifikasi tingkat kematangan durian berbasis e-nose dan raspberry pi dengan metode PCA, LDA, dan neural network

ABSTRAK:

Electronic Nose (E-Nose) is a gas sensor-based instrumentation system designed to detect, recognize, and classify volatile compound patterns from aroma samples in a non-destructive manner. This technology mimics the function of the human olfactory system and has been widely applied in various fields, including the food and agricultural industries. This study developed an E-Nose system based on Raspberry Pi equipped with eight gas sensors (MQ and TGS series) to classify the ripeness levels of durian fruit based on the aroma produced during the ripening process. Sensor data were recorded in real-time and analyzed using dimensionality reduction techniques such as Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA), followed by classification using a Neural Network algorithm. The results showed that the E-Nose system could significantly differentiate the aroma patterns of durians at various ripeness levels. PCA and LDA demonstrated clear cluster separation between ripeness stages, while the Neural Network model achieved classification accuracy above 90%. This research demonstrates that E-Nose is an effective tool for rapid and objective classification of durian ripeness, with strong potential for application in the fruit supply chain to maintain product quality.

ABSTRACT:
Electronic Nose (E-Nose) is a gas sensor-based instrumentation system designedto detect, recognize, and classify volatile compound patterns from aroma samples in a non-destructive manner. This technology mimics the function of the human olfactory systemand has been widely applied in various fields, including the food and agricultural industries.This study developed an E-Nose system based on Raspberry Pi equipped with eight gassensors (MQ and TGS series) to classify the ripeness levels of durian fruit based on thearoma produced during the ripening process. Sensor data were recorded in real-time andanalyzed using dimensionality reduction techniques such as Principal Component Analysis(PCA) and Linear Discriminant Analysis (LDA), followed by classification using a NeuralNetwork algorithm. The results showed that the E-Nose system could significantlydifferentiate the aroma patterns of durians at various ripeness levels. PCA and LDAdemonstrated clear cluster separation between ripeness stages, while the Neural Networkmodel achieved classification accuracy above 90%. This research demonstrates that E-Nose is an effective tool for rapid and objective classification of durian ripeness, withstrong potential for application in the fruit supply chain to maintain product quality

مستخلص البحث: