Analisis Keandalan Pada Sistem Pembangkit Listrik Hibrida Tenaga Surya Dan Tenaga Angin Dengan Pendekatan FMECA Berbasis ANFIS

Putri, Intan Parliani (2026) Analisis Keandalan Pada Sistem Pembangkit Listrik Hibrida Tenaga Surya Dan Tenaga Angin Dengan Pendekatan FMECA Berbasis ANFIS. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6009242002-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (29MB) | Request a copy

Abstract

Energi terbarukan khususnya tenaga surya dan tenaga angin terus dikembangkan di Indonesia, melalui penerapan sistem pembangkit listrik hibrida. Kompleksitas integrasi kedua sumber energi ini dapat meningkatkan potensi kegagalan sehingga diperlukan analisis keandalan dan risiko untuk menjamin kontinuitas operasi sistem. Penelitian ini bertujuan untuk menganalisis tingkat keandalan sistem pembangkit listrik hibrida tenaga surya-angin menggunakan pendekatan berbasis control chart, serta mengidentifikasi tingkat risiko dan mode kegagalan menggunakan metode Failure Mode, Effects, and Criticality Analysis (FMECA) berdasarkan parameter Severity (S), Occurrence (O), dan Detection (D). FMECA kemudian diintegrasikan dengan Adaptive Neuro-Fuzzy Inference System (ANFIS) untuk menghasilkan penilaian risiko yang lebih representatif dibandingkan metode konvensional. Analisis keandalan menunjukkan bahwa subsistem tenaga surya (PV) memiliki tingkat keandalan yang lebih tinggi dibandingkan subsistem tenaga angin dan sistem charging dengan nilai failure rate sebesar 0.02857 dan nilai reliability sebesar 97,18%. Turbin angin sendiri memiliki nilai failure rate sebesar 0,0352 dengan nilai reliability sebesar 96,53%, sedangkan untuk sistem charging memiliki failure rate 0.0597 dan reliability sebesar 94,204%. Keandalan keseluruhan sistem diperoleh sebesar 94,11%. Berdasarkan hasil analisis FMECA-ANFIS, beberapa mode kegagalan memiliki risk score yang lebih tinggi, yaitu pada kegagalan inverter baterai (73,2), masalah koneksi listrik PV (72,8), boost converter (72,59), dan generator (70,85), sehingga perlu diprioritaskan dalam kegiatan pemantauan, inspeksi, dan pemeliharaan. Hasil evaluasi model ANFIS menunjukkan bahwa fungsi keanggotaan generalized bell (gbell) memberikan performa terbaik dibandingkan fungsi keanggotaan lain dengan nilai RMSE training sebesar 2,991, RMSE testing sebesar 4,513, dan MAPE 4,79%. Integrasi FMECA dan ANFIS mampu menghasilkan pemeringkatan risiko yang lebih representatif sehingga dapat digunakan sebagai dasar dalam penyusunan strategi maintenance untuk meningkatkan keandalan sistem pembangkit listrik hibrida tenaga surya-angin.
=====================================================================================================================================
Renewable energy in Indonesia, especially solar and wind energy, has been increasingly developed in Indonesia through the implementation of hybrid power generation systems. The integration of these two renewable energy sources increases system complexity and the potential for failures, making reliability and risk analysis essential to ensure continuous system operation. This study aims to evaluate the reliability of a hybrid solar-wind power generation system using a control chart-based approach and to identify failure risks and failure modes using Failure Mode, Effects, and Criticality Analysis (FMECA) based on the parameters of Severity (S), Occurrence (O), and Detection (D). FMECA was integrated with the Adaptive Neuro-Fuzzy Inference System (ANFIS) to provide a more representative risk assessment than the conventional FMECA approach. The reliability analysis showed that the photovoltaic (PV) subsystem exhibited higher reliability than the wind turbine subsystem and the charging system, with a failure rate of 0.0285 and a reliability of 97.12%. The wind turbine subsystem had a failure rate of 0.0352 and a reliability of 96.53%, while the charging system had a failure rate of 0.04167 and a reliability of 95.92%. The overall system reliability was calculated to be 95.82%. Based on the FMECA-ANFIS analysis, several failure modes were identified as having the highest risk scores, including battery inverter failure (73.2), PV electrical connection faults (72.8), boost converter failure (72.59), and generator failure (70.85). These failure modes should be prioritized in system monitoring, inspection, and maintenance activities. The ANFIS model evaluation demonstrated that the generalized bell (gbell) membership function achieved the best performance among the evaluated membership functions, with a training RMSE of 2.991, a testing RMSE of 4.513, and a MAPE of 4.79%. The integration of FMECA and ANFIS produced a more representative risk ranking than the conventional approach and can serve as a basis for developing maintenance strategies to improve the reliability of hybrid solar-wind power generation systems.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Sistem hibrida surya-angin, keandalan sistem, FMECA, ANFIS, penilaian risiko, hybrid solar-wind power generation system, system reliability, FMECA, ANFIS, risk assessment
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7870.23 Reliability. Failures
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30101-(S2) Master Thesis
Depositing User:	Intan Parliani Putri
Date Deposited:	06 Aug 2026 06:06
Last Modified:	06 Aug 2026 06:06
URI:	http://repository.its.ac.id/id/eprint/144057

Actions (login required)

View Item