Sistem Deteksi Kesalahan pada Bearing Generator Hydropower Plant Menggunakan Similarity Based Model

Sri, Rohmawanto (2024) Sistem Deteksi Kesalahan pada Bearing Generator Hydropower Plant Menggunakan Similarity Based Model. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5009211161-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until April 2026. Download (3MB) | Request a copy

Abstract

Penelitian ini mengusulkan pengembangan sistem deteksi kesalahan pada sistem mekanis dalam hydropower plant. Deteksi kesalahan menjadi aspek penting dalam menjaga keandalan dan kinerja sistem, memungkinkan identifikasi dini terhadap peristiwa tidak terduga dari kumpulan data. Proses deteksi kesalahan menjadi sarana pencegahan kerugian yang lebih parah akibat dampak dari kesalahan tersebut. Sistem deteksi yang diusulkan terdiri dari tiga tahap utama: pengambilan data, pemrosesan sinyal, dan klasifikasi. Klasifikasi menjadi tahap paling penting karena memberikan prediksi kondisi suatu mesin. Dalam penelitian ini, SVM digunakan sebagai alat utama untuk proses klasifikasi data. Penelitian ini menggunakan 81 data hasil simulasi diantaranya 9 data kondisi normal, 27 data kondisi kesalahan inner race bearing, dan 45 data kondisi kesalahan angular misalignment. Data tersebut dibagi menjadi dua kelompok, yaitu 80% untuk data training dan 20% untuk data testing. Hasil confusion matrix untuk data testing menunjukkan nilai akurasi sebesar 98,8%. Analisis seleksi fitur menunjukkan bahwa hasil klasifikasi sangat dipengaruhi oleh pemilihan fitur. Hasil optimal, baik dari segi training maupun testing, ditemukan pada penggunaan 4 fitur khususnya dengan algoritma mRMR.
=================================================================================================================================
This research proposes the development of a fault detection system for mechanical systems in a hydropower plant. Fault detection is an important aspect in maintaining the reliability and performance of a system, enabling early identification of unforeseen events from the dataset. Fault detection process serves as a preventive measure to mitigate more severe losses resulting from these faults. Fault detection system consists of three main stages: data acquisition, signal processing, and classification. The classification stage is the most critical as it provides predictions regarding the condition of a machine. Support Vector Machine (SVM) as the primary tool for the data classification process. The research involves 81 simulation data, they are 9 datasets for normal condition, 27 datasets for inner race bearing faults, and 45 datasets for angular misalignment faults. The data is divided into two groups: 80% for training and 20% for testing. The confusion matrix results for the testing data demonstrate an accuracy of 98.8%. Feature selection analysis indicates that the classification outcomes are highly influenced by the feature selection process. Optimal results, both in terms of training and testing, are achieved using 4 features, particularly with the mRMR algorithm.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	angular misalignment, hydropower plant, inner race bearing, sistem deteksi, SVM, angular misalignment, fault detection system
Subjects:	Q Science > QC Physics > QC20.7.F67 Fourier transformations T Technology > T Technology (General) > T57.62 Simulation T Technology > TA Engineering (General). Civil engineering (General) > TA169.5 Failure analysis T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK1519.S68 Hydroelectric power plants
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis
Depositing User:	Sri Rohmawanto
Date Deposited:	07 Feb 2024 06:56
Last Modified:	07 Feb 2024 06:56
URI:	http://repository.its.ac.id/id/eprint/106392

Actions (login required)

View Item