Munawiri, Munawiri (2026) ANALISIS PERBANDINGAN PREDIKSI PRODUKSI ENERGI PLTS ON – GRID UNIVERSITAS MERCU BUANA MENGGUNAKAN POLYNOMIAL REGRESSION, RANDOM FOREST DAN AUTOREGRESSIVE INTEGRATED MOVING AVERAGE (ARIMA). S2 thesis, Universitas Mercu Buana Jakarta - Menteng.
|
Text (COVER)
55424110004-Munawiri-01 Cover - Munawiri.pdf Download (8MB) | Preview |
|
![[img]](https://repository.mercubuana.ac.id/style/images/fileicons/text.png) |
Text (BAB I)
55424110004-Munawiri-02 BAB 1 - Munawiri.pdf Restricted to Registered users only Download (596kB) |
|
|
Text (BAB II)
55424110004-Munawiri-03 BAB 2 - Munawiri.pdf Restricted to Registered users only Download (1MB) |
|
|
Text (BAB III)
55424110004-Munawiri-04 BAB 3 - Munawiri.pdf Restricted to Registered users only Download (1MB) |
|
|
Text (BAB IV)
55424110004-Munawiri-05 BAB 4 - Munawiri.pdf Restricted to Registered users only Download (1MB) |
|
|
Text (BAB V)
55424110004-Munawiri-06 BAB 5 - Munawiri.pdf Restricted to Registered users only Download (509kB) |
|
|
Text (DAFTAR PUSTAKA)
55424110004-Munawiri-08 Daftar Pustaka - Munawiri.pdf Restricted to Registered users only Download (492kB) |
|
|
Text (LAMPIRAN)
55424110004-Munawiri-09 Daftar Lampiran - Munawiri.pdf Restricted to Registered users only Download (1MB) |
Abstract
Peningkatan konsumsi energi listrik di lingkungan perguruan tinggi menuntut pengelolaan sistem energi yang lebih efisien dan terencana, khususnya pada sistem Pembangkit Listrik Tenaga Surya (PLTS) on-grid yang terintegrasi dengan jaringan Perusahaan Listrik Negara (PLN). Karakteristik produksi energi PLTS yang bersifat fluktuatif dan dipengaruhi oleh kondisi meteorologi menyebabkan perlunya metode prediksi yang mampu menangkap pola data secara akurat. Oleh karena itu, penelitian ini bertujuan untuk melakukan analisis perbandingan kinerja tiga metode prediksi produksi energi, yaitu Polynomial Regression, Random Forest, dan Autoregressive Integrated Moving Average (ARIMA), pada sistem PLTS on – grid di Universitas Mercu Buana. Penelitian ini menggunakan data produksi energi listrik harian selama periode September 2023 hingga September 2024, dengan total energi sebesar 65.475,9 kWh, rata-rata bulanan 171,85 kWh, serta variasi musiman sekitar 36% yang ditunjukkan oleh perbedaan produksi tertinggi 197,93 kWh dan terendah 144,66 kWh. Selain itu, produksi harian maksimum mencapai 309,4 kWh dan minimum sebesar 1,5 kWh, yang menunjukkan pengaruh signifikan faktor meteorologi. Evaluasi kinerja model dilakukan menggunakan Root Mean Square Error (RMSE) dan koefisien determinasi (R²), di mana hasil penelitian menunjukkan bahwa Polynomial Regression orde 5 memberikan performa terbaik dengan RMSE 2,77 dan R² 0,914, diikuti oleh Random Forest dengan RMSE 3,11 dan R² 0,831, sedangkan ARIMA menunjukkan performa yang kurang optimal dengan RMSE 9,48 dan R² -0,008. Hasil ini menunjukkan bahwa model berbasis hubungan nonlinier lebih unggul dibandingkan pendekatan deret waktu murni dalam memprediksi produksi energi PLTS, sehingga Polynomial Regression direkomendasikan sebagai metode yang lebih efektif untuk mendukung manajemen energi prediktif pada sistem PLTS on – grid di lingkungan perguruan tinggi The increasing electricity consumption in higher education institutions requires more efficient and well-planned energy management systems, particularly for gridconnected Solar Power Plants (Photovoltaic/PLTS on-grid) integrated with the national electricity grid (PLN). The fluctuating nature of solar energy production, influenced by meteorological conditions, necessitates the use of predictive models capable of accurately capturing data patterns. Therefore, this study aims to analyze and compare the performance of three energy production forecasting methods, namely Polynomial Regression, Random Forest, and Autoregressive Integrated Moving Average (ARIMA), applied to a grid-connected PLTS system at Universitas Mercu Buana. The study utilizes daily energy production data from September 2023 to September 2024, with a total energy output of 65,475.9 kWh, an average monthly production of 171.85 kWh, and seasonal variation of approximately 36%, indicated by the highest production of 197.93 kWh and the lowest of 144.66 kWh. Additionally, the maximum daily production reached 309.4 kWh, while the minimum was 1.5 kWh, reflecting significant meteorological influence. Model performance was evaluated using Root Mean Square Error (RMSE) and coefficient of determination (R²), where the results show that the 5th-order Polynomial Regression achieved the best performance with RMSE of 2.77 and R² of 0.914, followed by Random Forest with RMSE of 3.11 and R² of 0.831, while ARIMA showed poor performance with RMSE of 9.48 and R² of -0.008. These findings indicate that nonlinear-based models outperform pure time-series approaches in predicting solar energy production, and thus Polynomial Regression is recommended as the most effective method to support predictive energy management in grid-connected PV systems in higher education environments.
| Item Type: | Thesis (S2) |
|---|---|
| NIM/NIDN Creators: | 55424110004 |
| Uncontrolled Keywords: | PLTS on-grid, prediksi energi, Polynomial Regression, Random Forest, ARIMA, RMSE, R². grid-connected solar power plant, energy forecasting, Polynomial Regression, Random Forest, ARIMA, RMSE, R². |
| Subjects: | 600 Technology/Teknologi > 620 Engineering and Applied Operations/Ilmu Teknik dan operasi Terapan > 621 Applied Physics/Fisika terapan > 621.3 Electrical Engineering, Lighting, Superconductivity, Magnetic Engineering, Applied Optics, Paraphotic Technology, Electronics Communications Engineering, Computers/Teknik Elektro, Pencahayaan, Superkonduktivitas, Teknik Magnetik, Optik Terapan, Tekn |
| Divisions: | Pascasarjana > Magister Teknik Elektro |
| Depositing User: | ARDIFTA DWI AFRIANI |
| Date Deposited: | 12 May 2026 07:16 |
| Last Modified: | 12 May 2026 07:16 |
| URI: | http://repository.mercubuana.ac.id/id/eprint/102161 |
Actions (login required)
 |
View Item |
