RANCANG BANGUN KROMATOGRAFI GAS DENGAN TCD DAN SGP30 UNTUK KLASIFIKASI SENYAWA TOLUENA DAN ETANOL BERBASIS SUPPORT VECTOR MACHINE

ACHMAD NURNAAFI, . (2025) RANCANG BANGUN KROMATOGRAFI GAS DENGAN TCD DAN SGP30 UNTUK KLASIFIKASI SENYAWA TOLUENA DAN ETANOL BERBASIS SUPPORT VECTOR MACHINE. Sarjana thesis, UNIVERSITAS NEGERI JAKARTA.

	Text COVER.pdf Download (1MB)
	Text BAB I.pdf Download (277kB)
	Text BAB II.pdf Restricted to Registered users only Download (636kB) | Request a copy
	Text BAB III.pdf Restricted to Registered users only Download (1MB) | Request a copy
	Text BAB IV.pdf Restricted to Registered users only Download (3MB) | Request a copy
	Text BAB V.pdf Restricted to Registered users only Download (259kB) | Request a copy
	Text DAFTAR PUSTAKA.pdf Download (263kB)
	Text Lampiran.pdf Restricted to Registered users only Download (830kB) | Request a copy

Abstract

Keberadaan volatile organic compounds (VOC) seperti etanol dan toluena dalam lingkungan industri memerlukan sistem pemantauan yang portabel dan efisien. Sistem kromatografi gas konvensional cenderung mahal dan tidak praktis untuk penggunaan di lapangan. Penelitian ini mengembangkan sistem kromatografi gas portabel yang mengintegrasikan dua detektor, yaitu Thermal Conductivity Detector (TCD) dan sensor SGP30, tanpa penggunaan kolom berfase diam karena keterbatasan ketersediaan dan biaya komponen. Untuk mengatasi ketiadaan proses pemisahan senyawa, pendekatan klasifikasi berbasis machine learning diterapkan dengan memanfaatkan pola peluruhan sinyal dari sensor. Data sinyal diproses melalui tahapan trimming, penghapusan nilai hilang, filterisasi Savitzky-Golay, dan normalisasi, kemudian diekstraksi menjadi tujuh fitur statistik dan diseleksi menggunakan mutual information serta korelasi Pearson. Proses klasifikasi dilakukan menggunakan algoritma Support Vector Machine (SVM) dengan validasi hold-out dan Leave-One-Out Cross Validation (LOOCV). Hasil pengujian menunjukkan akurasi klasifikasi mencapai 100% pada kanal TVOC dan Combine, serta di atas 93% pada kanal TCD. Temuan ini menunjukkan bahwa sistem yang dikembangkan mampu mengidentifikasi senyawa VOC secara akurat melalui analisis pola sinyal, meskipun tanpa pemisahan kromatografis, dan berpotensi diterapkan untuk pemantauan gas secara portabel dan real-time. ***** Volatile organic compounds (VOCs) such as ethanol and toluene are commonly present in industrial environments and require portable and efficient monitoring systems. Conventional gas chromatography (GC) systems are often expensive and impractical for field use. This study developed a portable gas chromatography system integrating two detectors Thermal Conductivity Detector (TCD) and digital sensor SGP30 without using a stationary phase column due to limitations in component availability and cost. To compensate for the absence of chromatographic separation, a machine learning–based classification approach was applied by utilizing the decay pattern of sensor signals. The signal data underwent trimming, missing value removal, Savitzky-Golay filtering, and normalization, followed by extraction of seven statistical features. Feature selection was performed using mutual information and Pearson correlation, and classification was conducted using the Support Vector Machine (SVM) algorithm, validated with both hold-out and Leave-One-Out Cross Validation (LOOCV) methods. Experimental results showed classification accuracy reaching 100% for the TVOC and Combine channels, and over 93% for the TCD channel. These findings indicate that the developed system can accurately identify VOCs by analyzing signal patterns alone, and holds potential for use in real-time, portable gas monitoring applications.

Item Type:	Thesis (Sarjana)
Additional Information:	1). Dr.rer.nat. Bambang Heru Iswanto, M.Si. ; 2). Haris Suhendar, M.Sc.
Subjects:	Sains > Fisika
Divisions:	FMIPA > S1 Fisika
Depositing User:	Achmad Nurnaafi .
Date Deposited:	25 Aug 2025 03:42
Last Modified:	25 Aug 2025 03:42
URI:	http://repository.unj.ac.id/id/eprint/62022

Actions (login required)

View Item