Sistem Deteksi Kesalahan Pada Penukar Panas Menggunakan Suport Vector Machine

Artanita, Madarina (2024) Sistem Deteksi Kesalahan Pada Penukar Panas Menggunakan Suport Vector Machine. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5009211155-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 July 2026. Download (1MB) | Request a copy

Abstract

Penelitian ini fokus pada pengembangan sistem deteksi kesalahan pada penukar panas menggunakan metode klasifikasi data berbasis support vector machine (SVM). Metode yang digunakan dalam pengumpulan data adalah dengan melakukan pemodelan penukar panas menggunakan toolkit Simulink pada Matlab. Terdapat tiga pemodelan yang dilakukan yaitu keadaan normal, fouling dan kerusakan permukaan perpindahan panas. Dengan melibatkan 19 pengambilan data normal, 41 pengambilan data fouling, dan 41 pengambilan data kerusakan permukaan perpindahan panas, penelitian ini membagi data menjadi 85% untuk proses training dan 15% untuk proses testing. Hasil klasifikasi pada tahap training mencapai nilai precision 96,5% dan recall 97%, sementara pada tahap testing, nilai precision dan recall mencapai 100%. Temuan ini menunjukkan bahwa sistem machine learning yang digunakan dapat mengklasifikasikan input data dengan cukup akurat, memberikan dasar yang cukup untuk mengantisipasi dan mencegah potensi kesalahan pada penukar panas dalam kondisi operasional
====================================================================================================================================
This research focuses on developing an error detection system in heat exchangers using a support vector machine (SVM) based data classification method. The method used in data collection is modeling the heat exchanger using the Simulink toolkit in Matlab. There are three models carried out, namely normal conditions, fouling and damage to the heat transfer surface. Involving 19 normal data collections, 41 fouling data collections, and 41 heat transfer surface damage data collections, this research divided the data into 85% for the training process and 15% for the testing process. The classification results at the training stage reached a precision value of 96.5% and a recall of 97%, while at the testing stage, the precision and recall values reached 100%. These findings indicate that the machine learning system used can classify input data quite accurately, providing a sufficient basis for anticipating and preventing potential errors in heat exchangers under operational conditions

Item Type:	Thesis (Other)
Uncontrolled Keywords:	penukar panas, kesalahan, fouling, kerusakan permukaan, SVM; heat exchanger, faults, fouling, surface damage, SVM
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7870.23 Reliability. Failures
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis
Depositing User:	Madarina Artanita
Date Deposited:	06 Feb 2024 04:26
Last Modified:	06 Feb 2024 04:26
URI:	http://repository.its.ac.id/id/eprint/106152

Actions (login required)

View Item