Sistem Monitoring dan Kontrol Energi Cerdas Panel Surya pada Alat Peringatan Dini Banjir Berbasis IoT

Abstract

Keberlanjutan pasokan energi off-grid menjadi tantangan utama operasional Early Warning System (EWS) banjir di Kali Ciluar. Fluktuasi panel surya dan ketiadaan pengawasan real-time berisiko memicu deep discharge pada aki, yang berpotensi menyebabkan kegagalan sistem. Penelitian ini merancang sistem monitoring neraca energi berbasis mikrokontroler ESP32-S3 dan dual-sensor INA219, mengimplementasikan kontrol histeresis baterai (hold data mode pada SOC = 20% dan resume pada SOC = 80%), serta mengintegrasikan ThingSpeak dan Telegram Bot API. Pengujian lapangan dilakukan off-grid selama lima hari di Kali Ciluar. Kalibrasi sensor INA219 terhadap multimeter Sanwa CD800a menunjukkan akurasi 99,85% untuk tegangan dan 98,51% untuk arus. Evaluasi neraca energi harian membuktikan sistem selalu mengalami surplus positif dengan sisa energi bersih +23,46 Wh hingga +69,29 Wh per hari. Algoritma histeresis berhasil memutus transmisi wi-fi saat kapasitas kritis dan memulihkannya secara otomatis setelah terisi. Efisiensi pemanfaatan energi harian berkisar antara 42,06% hingga 56,62%, dengan ±52% sisa energi dialokasikan sebagai cadangan baterai harian. Sistem ini terbukti andal menjaga keberlanjutan operasional alat kebencanaan.

Off-grid power supply is a major challenge for flood Early Warning System (EWS) operations at Kali Ciluar. Solar panel fluctuations and lack of real-time monitoring risk triggering battery deep discharge, potentially causing system failure. This study designs an energy balance monitoring system based on ESP32-S3 microcontroller and dual INA219 sensors, implements battery hysteresis control (hold data mode at SOC =20% and resume at SOC =80%), and integrates ThingSpeak and Telegram Bot API. Field testing was conducted off-grid for five days at Kali Ciluar. INA219 sensor calibration against Sanwa CD800a multimeter showed 99.85% accuracy for voltage and 98.51% for current. Daily energy balance evaluation proved the system always experiences a positive surplus with net energy remaining between +23.46 Wh to +69.29 Wh daily. Hysteresis algorithm successfully cut Wi-Fi transmission during critical capacity and automatically restored it after charging. Daily energy utilization efficiency ranged from 42.06% to 56.62%, with ±52% remaining energy allocated as daily battery reserve. This system is proven reliable in maintaining operational sustainability of disaster equipment.