Sistem Klasifikasi Kerusakan Bearing Roda Ship Unloader Menggunakan Metode Artificial Neural Network

Fahreza, Rakhya Rizq (2024) Sistem Klasifikasi Kerusakan Bearing Roda Ship Unloader Menggunakan Metode Artificial Neural Network. Diploma thesis, Institut Teknologi Sepuluh Nopember.

Text 2040201055-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (4MB) | Request a copy

Abstract

PT Adhi Guna Putera sebagai jasa pembongkaran batu bara bertanggung jawab atas kelancaran proses bongkar muat yang menjadi salah satu faktor yang perlu diberikan prioritas perhatian dengan cara menjaga agar kondisi ship unloader dapat beroperasi dengan baik. Ship unloader adalah suatu alat yang berfungsi untuk kegiatan pembongkaran batu bara dari tongkang menuju ke hopper kemudian batubara diteruskan untuk keperluan bahan bakar PLTU. Ship unloader merupakan salah satu perangkat penting dalam operasi pembongkaran muatan di pelabuhan, yang memainkan peran utama dalam proses transportasi batu bara. Dalam operasi sehari-hari, ship unloader bekerja dalam kondisi yang berkelanjutan, yang dapat menyebabkan peningkatan suhu dan getaran yang terjadi pada bearing roda ship unloader. Kerusakan akibat faktor-faktor seperti suhu tinggi dan getaran berlebih yang dapat terjadi selama operasi. Suhu berlebih disebabkan kurangnya pelumasan sedangkan getaran berlebih disebabkan bantalan yang mulai tidak rata bahkan retak atau pecah. Oleh karena itu, sangat penting untuk memiliki sistem monitoring yang efektif untuk mengukur dan memantau suhu dan getaran yang terjadi pada roda ship unloader. Sistem klasifikasi berdasarkan suhu dan getaran pada bearing roda ship unloader ini dapat memungkinkan deteksi dini masalah potensial yang dapat terjadi, memaksimalkan perawatan dan perbaikan tepat waktu. Sistem ini tidak hanya memantau kondisi bearing secara real-time, tetapi memungkinkan identifikasi dini terhadap potensi kerusakan, sehingga meminimalkan risiko kerusakan. Dengan menggunakan model Artificial Neural Network (ANN), sistem ini dapat mengklasifikasi kondisi bearing yang akan datang. Berdasarkan hasil pengujian, arsitektur neural network terbaik menggunakan pembagian data training sebesar 80% dan data testing sebesar 20% dengan hidden layer sebanyak 32 neuron dan 100 epoch. Hasil dari metode tersebut memperoleh nilai akurasi sebasar 96,27%. Dataset diklasifikasikan menjadi 4 jenis kondisi bearing yaitu normal, inner race fault, outer race fault, dan roller element fault. Hasil evaluasi model klasifikasi pada kelas normal mendapatkan nilai akurasi 1,00, precision 1,00, recall 1,00, dan F1-score 1,00. Kelas kerusakan inner race menghasilkan nilai akurasi 0,94, precision 0,90, recall 0,90, dan F1-score 0,90. Kelas kerusakan outer race menghasilkan nilai akurasi 0,98, precision 0,94, recall 0,94, dan F1-score 0,94. Kelas kerusakan roller element menghasilkan nilai akurasi 1,00, precision 1,00, recall 1,00, dan F1-score 1,00.
==================================================================================================================================
PT Adhi Guna Putera as a coal unloading service is responsible for the smooth loading and unloading process which is one of the factors that needs to be given priority attention by maintaining the condition of the ship unloader to operate properly. Ship unloader is a tool that functions for coal unloading activities from the barge to the hopper then the coal is forwarded for PLTU fuel purposes. Ship unloader is one of the important devices in unloading operations at the port, which plays a major role in the coal transportation process. In daily operation, the ship unloader works under continuous conditions, which can lead to an increase in temperature and vibration occurring in the ship unloader wheel bearings. Damage due to factors such as high temperature and excessive vibration that can occur during operation. Excessive temperature is due to lack of lubrication while excessive vibration is due to bearings that start to become uneven and even crack or break. Therefore, it is very important to have an effective monitoring system to measure and monitor the temperature and vibration that occur on the ship unloader wheels. This classification system based on temperature and vibration in ship unloader wheel bearings can enable early detection of potential problems that could occur, maximizing timely maintenance and repair. The system not only monitors the condition of the bearings in real-time, but allows early identification of potential damage, thus minimizing the risk of damage. By using an Artificial Neural Network (ANN) model, the system can classify the future condition of bearings. Based on the test results, the best neural network architecture uses a division of training data by 80% and testing data by 20% with a hidden layer of 32 neurons and 100 epochs. The results of the method obtained an accuracy value of 96.27%. The dataset is classified into 4 types of bearing conditions, namely normal, inner race fault, outer race fault, and roller element fault. The evaluation results of the classification model in the normal class get an accuracy value of 1.00, precision 1.00, recall 1.00, and F1-score 1.00. The inner race damage class produces an accuracy value of 0.94, precision 0.90, recall 0.90, and F1-score 0.90. The outer race damage class produces an accuracy value of 0.98, precision 0.94, recall 0.94, and F1-score 0.94. The roller element damage class produces an accuracy value of 1.00, precision 1.00, recall 1.00, and F1-score 1.00.

Item Type:	Thesis (Diploma)
Uncontrolled Keywords:	Klasifikasi, Bearing, Temperatur, Getaran, Neural Network, Classification, Temperature, Vibration
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105.546 Computer algorithms
Divisions:	Faculty of Vocational > 36304-Automation Electronic Engineering
Depositing User:	Rakhya Rizq Fahreza
Date Deposited:	11 Sep 2024 08:26
Last Modified:	11 Sep 2024 08:26
URI:	http://repository.its.ac.id/id/eprint/115623

Actions (login required)

View Item