Sistem Monitoring Elektrolisis Air Berbasis IOT Dengan Implementasi Artificial Neural Network (ANN) Untuk Prediksi Parameter PID (KP, KI, KD)

Nisa', Nadia Kamilatun (2026) Sistem Monitoring Elektrolisis Air Berbasis IOT Dengan Implementasi Artificial Neural Network (ANN) Untuk Prediksi Parameter PID (KP, KI, KD). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 2042221066-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (6MB) | Request a copy

Abstract

Kebutuhan hidrogen global diperkirakan meningkat dari 90 juta ton per tahun menjadi 530–810 juta ton pada tahun 2050 untuk mendukung target net zero emission. Alkaline Water Electrolyzer (AWE) merupakan teknologi yang berpotensi memenuhi kebutuhan tersebut melalui proses elektrolisis air. Performa AWE dipengaruhi oleh parameter operasi yang perlu dipantau secara real-time. Penelitian ini merancang sistem monitoring elektrolisis air berbasis Internet of Things (IoT) yang terintegrasi dengan Firebase Cloud Platform melalui protokol HTTP. Sistem menggunakan mikrokontroler ESP32-S3 untuk mengakuisisi data dari sensor DS18B20 (temperatur), INA219 (tegangan elektroda), dan F1031V (laju aliran gas). Data dikirim ke Firebase Realtime Database, disinkronisasi otomatis ke Google Spreadsheet sebagai penyimpanan historis, dan divisualisasikan pada dashboard website secara real-time. Hasil validasi sensor menunjukkan sensor DS18B20 memiliki error 3,88% dengan akurasi 96,12%, sensor F1031V memiliki error 4,60% dengan akurasi 95,40%, dan sensor INA219 memiliki error 2,45% dengan akurasi 97,55%. Pengujian komunikasi IoT menunjukkan rata-rata delay sebesar 0,9333 detik dengan rentang 0–3 detik. Model Artificial Neural Network (ANN) dikembangkan menggunakan arsitektur 2 hidden layer (12,12) dengan aktivasi tanh untuk memprediksi parameter PID (Kp, Ki, Kd) dari data temperatur. Dataset berjumlah 15 data dibagi dengan rasio 80% training dan 20% testing. Hasil evaluasi model menunjukkan Training MAE 0,1057 dan Testing MAE 0,9636.
===================================================================================================================================
Global hydrogen demand is estimated to increase from 90 million tons per year to 530–810 million tons by 2050 to support net zero emission targets. Alkaline Water Electrolyzer (AWE) is a technology that has the potential to meet this demand through the process of water electrolysis. AWE performance is influenced by operating parameters that need to be monitored in real time. This study designs an Internet of Things (IoT)-based water electrolysis monitoring system integrated with the Firebase Cloud Platform via the HTTP protocol. The system uses an ESP32-S3 microcontroller to acquire data from DS18B20 (temperature), INA219 (electrode voltage), and F1031V (gas flow rate) sensors. The data is sent to the Firebase Realtime Database, automatically synchronized to Google Spreadsheet for historical storage, and visualized on a website dashboard in real-time. Sensor validation results show that the DS18B20 sensor has an error of 3.88% with an accuracy of 96.12%, the F1031V sensor has an error of 4.60% with an accuracy of 95.40%, and the INA219 sensor has an error of 2.45% with an accuracy of 97.55%. IoT communication testing showed an average delay of 0.9333 seconds with a range of 0–3 seconds. An Artificial Neural Network (ANN) model was developed using a 2 hidden layer (12,12) architecture with tanh activation to predict PID parameters (Kp, Ki, Kd) from temperature data. The dataset consisted of 15 data points divided into an 80% training and 20% testing ratio. Model evaluation results showed a Training MAE of 0.1057 and a Testing MAE of 0.9636

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Elektrolisis Air, Internet Of Things (IoT), Firebase Realtime Database, Artificial Neural Network (ANN) Water Electrolysis, Internet of Things (IoT), Firebase Realtime Database, Artificial Neural Network (ANN)
Subjects:	T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T57.84 Heuristic algorithms. T Technology > TP Chemical technology > TP255 Electrochemistry, Industrial.
Divisions:	Faculty of Vocational > Instrumentation Engineering
Depositing User:	Nadia Kamilatun Nisa
Date Deposited:	02 Aug 2026 18:58
Last Modified:	02 Aug 2026 18:58
URI:	http://repository.its.ac.id/id/eprint/141054

Actions (login required)

View Item