Revanska, Athallah Iqbal Farellio (2026) Desain Dan Implementasi Sistem Adaptive Charging Dengan Metode CC-CV Menggunakan Buck Converter Berbasis Artificial Neural Network Pada Baterai Li-ion. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5022221126-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5022221126-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (6MB) | Request a copy |
Abstract
Peningkatan tren penggunaan kendaraan listrik mendorong optimalisasi teknologi penyimpanan energi, di mana baterai Li-ion menjadi pilihan utama karena densitas energi dan fleksibilitas desain yang tinggi. Namun, metode pengisian daya Constant Current-Constant Voltage (CC-CV) konvensional dengan kontrol PI sering kali mengalami fenomena stalling saat transisi tingkat arus, serta overshoot arus dan tegangan yang signifikan pada fase transisi CC ke CV. Penelitian ini mengusulkan sistem pengisian daya adaptif berbasis Artificial Neural Network (ANN) untuk mengatur duty cycle pada DC-DC buck converter. Sistem ini diterapkan pada baterai Li-ion 2500 mAh yang diseri 3 sel (12,6 V). Adaptive charging digunakan dengan menggunakan metode dua tingkatan mode CC (CC1 sebesar 2 A dan CC2 sebesar 1,5 A). Hasil simulasi menunjukkan kontrol ANN mampu merespons transisi CC1-CC2 secara instan dalam waktu 437 µs tanpa kondisi idle yang dialami oleh kontrol PI. Selain itu, pada fase transisi CC2 ke CV, kontrol ANN berhasil mereduksi overshoot arus sebesar 93,08% dan mengeliminasi overshoot tegangan sepenuhnya dibandingkan dengan penggunaan kontrol PI. Hasil implementasi menunjukkan DC-DC buck converter menghasilkan efisiensi rata-rata sebesar 81,81%. Pengujian pengisian daya dimulai dari tegangan awal baterai 10,1 V berhasil mengeksekusi fase CC1 selama 28 menit, CC2 selama 37 menit, dan fase CV (12,6 V) selama 47 menit tanpa gejala overshoot sebelum pengisian berhenti otomatis saat terbaca arus 0,5 A. Total waktu pengisian daya hingga baterai penuh adalah 1 jam 52 menit.
====================================================================================================================================
The rising trend in electric vehicle adoption has driven the optimization of energy storage technologies, where Lithium-ion (Li-ion) batteries serve as the primary choice due to their high energy density and design flexibility. However, the conventional Constant Current-Constant Voltage (CC-CV) charging method regulated by PI control often suffers from stalling phenomena during current level transitions, as well as significant current and voltage overshoots during the CC-to-CV transition phase. This study proposes an Artificial Neural Network (ANN)-based adaptive charging system to regulate the duty cycle of a DC-DC buck converter. The system is implemented on a 3-cell series (12.6 V) 2500 mAh Li-ion battery. Adaptive charging is executed utilizing a dual-stage CC mode method (CC1 at 2 A and CC2 at 1.5 A). Simulation results demonstrate that the ANN control is capable of responding to the CC1-CC2 transition instantaneously within 437 µs, completely eliminating the idle state experienced under PI control. Furthermore, during the CC2-to-CV transition phase, the ANN control successfully reduces current overshoot by 93.08% and completely eliminates voltage overshoot compared to the deployment of PI control. Implementation results indicate that the DC-DC buck converter achieves an average efficiency of 81.81%. Charging tests initiated from an initial battery voltage of 10.1 V successfully executed the CC1 phase for 28 minutes, the CC2 phase for 37 minutes, and the CV phase (12.6 V) for 47 minutes without any overshoot symptoms before the charging automatically terminated when a current of 0.5 A was detected. The total charging time to full capacity is 1 hour and 52 minutes.
Actions (login required)
 |
View Item |