Article Per Year (5 Year)
p-Index From 2021 - 2026
0.408
P-Index
This Author published in this journals
All Journal International Journal of Electrical and Computer Engineering Journal of Applied Smart Electrical Network and System (JASENS)
Dimas Pristovani Riananda
Politeknik Perkapalan Negeri Surabaya
Computer Science & IT Control & Systems Engineering Electrical & Electronics Engineering Energy Engineering

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

 1 
Simulasi Perbandingan Training Dataset YOLOv5 Dan YOLOv8 Menggunakan Prototipe AUV Ryan Yudha Adhitya; Erutan Dwi Sajiwo; Zindhu Maulana Ahmad Putra; M. Yusuf Santoso; M. Khoirul Hasin; Adianto; Dimas Pristovani Riananda; Isa Rachman; Agus Khumaidi; Yuning Widiarti
Journal of Applied Smart Electrical Network and Systems Vol. 7 No. 1 (2026): JASENS Vol. 7 No. 1 (2026) : Vol. 07 No. 01, Juni 2026
Publisher : Indonesian Society of Applied Science (ISAS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52158/pkrjg903

Abstract

Penelitian ini mengevaluasi performa pelatihan model YOLOv5n dan YOLOv8n menggunakan dataset Kompetisi Robot Berbasis Kecerdasan Buatan (KRBAI) 2024 untuk deteksi objek pada prototipe Autonomous Underwater Vehicle (AUV). Dataset yang terdiri atas 4.114 citra dengan tiga kelas objek (gerbang, rintangan, drum) dilatih melalui Google Colab dengan konfigurasi 200 epoch, ukuran batch 16, dan resolusi citra 640x640. YOLOv5n memanfaatkan struktur berbasis anchor, sedangkan YOLOv8n menggunakan pendekatan tanpa anchor dengan modul C2f untuk efisiensi lebih tinggi. Hasil pelatihan menunjukkan YOLOv8n unggul dalam presisi (0,994 vs. 0,990) dan mAP50-95 (0,940 vs. 0,924), sementara YOLOv5n lebih baik pada recall kelas gerbang (0,999). Kedua model memiliki mAP50 serupa (0,993). Waktu pelatihan YOLOv8n sedikit lebih lama (3 jam 17 menit vs. 3 jam 31 menit). YOLOv8n terbukti lebih efektif untuk deteksi objek real-time pada AUV, mendukung navigasi dan pengenalan lingkungan bawah air. Penelitian mendatang perlu menguji dataset yang lebih beragam dan kondisi lingkungan yang lebih kompleks.
Implementation of support vector machine on LVMDP panel with overheating protection system Annas Singgih Setiyoko; Dimas Pristovani Riananda; Adianto Adianto
International Journal of Electrical and Computer Engineering (IJECE) Vol 16, No 4: August 2026
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijece.v16i4.pp1755-1766

Abstract

Electricity is a critical requirement in industrial operations, where the continuity and stability of power distribution directly affect safety and productivity. The low voltage main distribution panel (LVMDP) functions as the main node of electrical power distribution; however, conventional LVMDP systems generally lack intelligent protection mechanisms capable of detecting overheating-related fire hazards and initiating preventive action before failure occurs. This study proposes an intelligent monitoring and protection system for LVMDP panels that combines real-time multi-sensor monitoring, support vector machine (SVM)-based hazard classification, and an automatic shutdown mechanism. The main contribution of this work lies in the integration of predictive thermal risk detection with autonomous protective action, enabling the system not only to monitor panel conditions but also to respond immediately to hazardous states before they escalate into fire incidents. SVM was selected because of its strong capability to classify complex and nonlinear patterns from sensor data with high reliability. The developed system continuously evaluates panel conditions and triggers auto-shutdown when an overheating risk is identified, thereby improving preventive protection compared with conventional alarm-based monitoring systems. Experimental results show that the sensor measurements achieved error rates mostly below 5% compared with calibrated instruments, indicating good accuracy. In addition, the SVM model obtained an overall accuracy of 93%, with a macro-average F1-score of 92% and a weighted-average F1-score of 93%. These results demonstrate that the proposed system is effective for early detection and active protection of LVMDP panels against overheating hazards.
Co-Authors Adianto Adianto Adianto Agus Khumaidi Annas Singgih Setiyoko Erutan Dwi Sajiwo M. Khoirul Hasin M. Yusuf Santoso Ryan Yudha Adhitya Yuning Widiarti Zindhu Maulana Ahmad Putra