PENGEMBANGAN SISTEM KROMATOGRAFI GAS LOW-COST BERBASIS SENSOR SGP30: OPTIMASI KOLOM DAN KLASIFIKASI KOPI ARABIKA–ROBUSTA DENGAN MACHINE LEARNING

Sayid Mahmud Ibadirhaman Syah, . (2026) PENGEMBANGAN SISTEM KROMATOGRAFI GAS LOW-COST BERBASIS SENSOR SGP30: OPTIMASI KOLOM DAN KLASIFIKASI KOPI ARABIKA–ROBUSTA DENGAN MACHINE LEARNING. Sarjana thesis, UNIVERSITAS NEGERI JAKARTA.

	Text Cover.pdf Download (2MB)
	Text BAB I.pdf Download (289kB)
	Text BAB II.pdf Restricted to Registered users only Download (812kB) | Request a copy
	Text BAB III.pdf Restricted to Registered users only Download (976kB) | Request a copy
	Text BAB IV.pdf Restricted to Registered users only Download (2MB) | Request a copy
	Text BAB V.pdf Restricted to Registered users only Download (355kB) | Request a copy
	Text Daftar Pustaka.pdf Download (358kB)
	Text Lampiran.pdf Restricted to Registered users only Download (845kB) | Request a copy

Abstract

Kopi Arabika dan Robusta memiliki karakteristik senyawa volatil yang berbeda sehingga berpotensi dimanfaatkan sebagai dasar autentikasi. Namun, instrumen kromatografi gas konvensional masih memiliki biaya pengadaan dan operasional yang relatif tinggi. Penelitian ini bertujuan mengembangkan sistem kromatografi gas low-cost berbasis sensor SGP30 melalui optimasi konfigurasi kolom untuk menghasilkan fingerprint volatil kopi, serta mengevaluasi kinerja algoritma Support Vector Machine (SVM), Random Forest, dan Extreme Gradient Boosting (XGBoost) dalam mengklasifikasikan kopi Arabika dan Robusta. Sinyal hasil pengukuran diproses menggunakan metode Positive ReLU Decay, kemudian diekstraksi menjadi fitur statistik dan dinamis. Analisis data dilakukan menggunakan Principal Component Analysis (PCA) dan Mutual Information (MI), sedangkan optimasi model dilakukan menggunakan Grid Search dan dievaluasi menggunakan K-Fold Cross Validation. Hasil penelitian menunjukkan bahwa sistem yang dikembangkan mampu menghasilkan fingerprint volatil yang representatif untuk membedakan kopi Arabika dan Robusta. Ketiga algoritma memberikan performa klasifikasi yang sangat baik, dengan SVM menghasilkan kinerja terbaik. Penelitian ini menunjukkan bahwa sistem kromatografi gas low-cost berbasis sensor SGP30 berpotensi menjadi alternatif yang ekonomis untuk skrining awal autentikasi kopi. ***** Arabica and Robusta coffee possess distinct volatile compound profiles that can be utilized for authentication. However, conventional gas chromatography systems remain expensive for routine applications. This study aims to develop an SGP30-based low-cost gas chromatography system through column configuration optimization to generate volatile fingerprints of coffee and to evaluate the performance of Support Vector Machine (SVM), Random Forest, and Extreme Gradient Boosting (XGBoost) for classifying Arabica and Robusta coffee. The acquired signals were processed using the Positive ReLU Decay method and transformed into statistical and dynamic features. Feature analysis was performed using Principal Component Analysis (PCA) and Mutual Information (MI), while model optimization was conducted using Grid Search and evaluated through K-Fold Cross Validation. The results demonstrate that the proposed system successfully generated representative volatile fingerprints for distinguishing Arabica and Robusta coffee. All three algorithms achieved excellent classification performance, with SVM providing the best overall performance. These findings indicate that the proposed SGP30-based low-cost gas chromatography system is a promising and economical alternative for preliminary coffee authentication.

Item Type:	Thesis (Sarjana)
Additional Information:	1). Dr. Bambang Heru Iswanto, M.Si. ; 2). Haris Suhendar, S.Si., M.Sc.
Subjects:	Sains > Fisika
Divisions:	FMIPA > S1 Fisika
Depositing User:	Users 35855 not found.
Date Deposited:	10 Aug 2026 07:07
Last Modified:	10 Aug 2026 07:07
URI:	http://repository.unj.ac.id/id/eprint/69284

Actions (login required)

View Item