IMPLEMENTASI ALGORITMA JARINGAN SARAF TIRUAN BACKPROPAGATION UNTUK PREDIKSI DAYA FOTOVOLTAIK PADA SISTEM SMART STREET LIGHTING

Ridwan Riswana, - and Ade Gaffar Abdullah, - (2025) IMPLEMENTASI ALGORITMA JARINGAN SARAF TIRUAN BACKPROPAGATION UNTUK PREDIKSI DAYA FOTOVOLTAIK PADA SISTEM SMART STREET LIGHTING. S1 thesis, Universitas Pendidikan Indonesia.

Abstract

Salah satu tantangan utama dalam penerapan Smart Street Lighting (SSL) adalah ketidakstabilan output daya dari sistem fotovoltaik. Ketidakstabilan ini dipengaruhi oleh berbagai faktor lingkungan seperti intensitas cahaya matahari, suhu, kelembaban udara, dan kecepatan angin. Ketidakpastian yang diakibatkan oleh faktor-faktor tersebut dapat berdampak signifikan terhadap kinerja sistem penerangan jalan apabila tidak ditangani secara optimal. Penelitian ini mengimplementasikan algoritma Jaringan Saraf Tiruan (JST) untuk memprediksi daya fotovoltaik pada sistem SSL. Metode penelitian yang digunakan bersifat kuantitatif dengan melibatkan variabel intensitas cahaya matahari, suhu, kelembaban udara, kecepatan angin, dan daya fotovoltaik. Data yang digunakan dalam penelitian ini sebanyak 158.300 sampel yang tercatat dari 1 – 4 November 2024. Data tersebut dibagi menjadi 3 bagian, yaitu 70 % sebagai data pelatihan, 15% sebagai data validasi, dan 15% untuk data pengujian. Struktur jaringan yang digunakan terdiri dari 4 – 70 – 1 dan 4 – 90 – 1. Hasil evaluasi menunjukan bahwa arsitektur 4 – 70 – 1 memperoleh akurasi 98,08% dengan nilai presentase kesalahan rata-rata (MAPE) 1,92% sedangkan arsitektur 4 – 90 – 1 mencapai akurasi 98,13% dengan MAPE 1,87%. Nilai MAPE yang rendah <10% menandakan bahwa kedua model memiliki tingkat prediksi yang sangat baik, dengan arsitektur 4 – 90 – 1 memberikan performa terbaik. One of the major challenges in the implementation of Smart Street Lighting (SSL) is the instability of the power output of the photovoltaic system. This instability is affected by various environmental factors such as sunlight intensity, temperature, air humidity, and wind speed. The uncertainty caused by these factors can have a significant impact on the performance of the street lighting system if not handled optimally. This research implements an Artificial Neural Network (ANN) algorithm to predict photovoltaic power in SSL systems. The research method used is quantitative by involving variables of sunlight intensity, temperature, air humidity, wind speed, and photovoltaic power. The data used in this study were 158,300 samples recorded from November 1 to 4, 2024. The data is divided into 3 parts, namely 70% as training data, 15% as validation data, and 15% as testing data. The network structure used consists of 4 - 70 - 1 and 4 - 90 - 1. The results of this study show that the architecture with the number of 70 neurons produces a MAPE value of 1.92% with an accuracy of 98.08%, while the architecture with the number of 90 neurons produces a MAPE value of 1.87% with an accuracy of 98.13%. This illustrates that the architecture model with 90 neurons is better at predicting photovoltaic power. Based on this, the two models that have been tested produce a MAPE value <10%, indicating that the two models are in the excellent prediction category.

	Text S_TE_2108009_Title.pdf Download (1MB)
	Text S_TE_2108009_Chapter 1.pdf Download (142kB)
	Text S_TE_2108009_Chapter 2.pdf Restricted to Staf Perpustakaan Download (295kB)
	Text S_TE_2108009_Chapter 3.pdf Download (299kB)
	Text S_TE_2108009_Chapter 4.pdf Restricted to Staf Perpustakaan Download (507kB)
	Text S_TE_2108009_Chapter 5.pdf Download (115kB)
	Text S_TE_2108009_Appendix.pdf Restricted to Staf Perpustakaan Download (540kB)

Item Type:	Thesis (S1)
Additional Information:	https://scholar.google.com/citations?hl=en&user=8iNMkxwAAAAJ ID SINTA Dosen Pembimbing: Ade Gaffar Abdullah: 257412
Uncontrolled Keywords:	Jaringan saraf tiruan, daya fotovoltaik, backpropagation, smart street lighting. Artificial neural network, backpropagation, smart street lighting, photovoltaic system
Subjects:	L Education > L Education (General) Q Science > Q Science (General) T Technology > TA Engineering (General). Civil engineering (General) T Technology > TK Electrical engineering. Electronics Nuclear engineering
Divisions:	Fakultas Pendidikan Teknik dan Industri > Jurusan Pendidikan Teknik Elektro
Depositing User:	Ridwan Riswana
Date Deposited:	12 Nov 2025 07:15
Last Modified:	12 Nov 2025 07:15
URI:	http://repository.upi.edu/id/eprint/145083

Actions (login required)

View Item