Klasifikasi Aroma Kayu Gaharu Menggunakan Sistem Electronic Nose

Haqqi, Fahreza (2025) Klasifikasi Aroma Kayu Gaharu Menggunakan Sistem Electronic Nose. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6022212030-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2027. Download (4MB) | Request a copy

Abstract

Kayu gaharu merupakan komoditas bernilai tinggi yang banyak digunakan dalam industri parfum dan kegiatan keagamaan. Kualitasnya ditentukan berdasarkan aroma yang dihasilkan saat dipanaskan, tetapi penilaian aroma yang bersifat subjektif sering kali menimbulkan ketidakpuasan dalam proses pembelian. Untuk mengatasi masalah ini, penelitian ini mengembangkan sistem deteksi kualitas kayu gaharu menggunakan electronic nose dan machine learning. Sistem ini menggunakan deret sensor gas semikonduktor, yaitu MQ-2, MQ-4, TGS 813, TGS 822, TGS 2610, dan TGS 2620, untuk mendeteksi gas yang dihasilkan dari pemanasan kayu gaharu. Data dari sensor kemudian dianalisis menggunakan empat metode klasifikasi, yaitu neural network (NN), convolutional neural network (CNN), support vector machine (SVM), dan random forest (RF). Hasil pengujian menunjukkan bahwa metode CNN pada suhu 160°C memberikan tingkat akurasi tertinggi, yaitu 87,5%, dalam mengklasifikasi aroma kayu gaharu dari jenis Bintuni, Merauke, Sulawesi, dan Borneo. Penelitian ini diharapkan dapat menjadi solusi alternatif untuk menentukan kualitas kayu gaharu secara lebih objektif dan efektif, menggantikan penilaian berbasis indra penciuman manusia.
==================================================================================================================================
Agarwood (gaharu) is a high-value commodity widely used in the perfume industry and religious rituals. Its quality is determined by the aroma produced when heated, but the subjective nature of aroma assessment often leads to dissatisfaction in the purchasing process. To address this issue, this study developed a system to detect the quality of agarwood using electronic nose and machine learning methods. The system utilizes an array of gas sensors, including MQ-2, MQ-4, TGS 813, TGS 822, TGS 2610, and TGS 2620, to detect gases released during the heating process. Data from the sensors are analyzed using four classification methods: Neural Network (NN), Convolutional Neural Network (CNN), Support Vector Machine (SVM), and Random Forest (RF). Experimental results show that the CNN method at a temperature of 160°C achieved the highest accuracy, reaching 87.5%, in classifying the aroma of agarwood from the Bintuni, Merauke, Sulawesi, and Borneo varieties. This study is expected to provide an alternative solution for determining agarwood quality in a more objective and effective manner, replacing human scent-based assessment.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	agarwood, CNN, electronic nose, gaharu, NN, RF, SVM
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7871.674 Detectors. Sensors
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User:	Fahreza Haqqi
Date Deposited:	20 Jan 2025 07:26
Last Modified:	20 Jan 2025 07:26
URI:	http://repository.its.ac.id/id/eprint/116438

Actions (login required)

View Item