PROTOTYPE SISTEM MONITORING DAN KENDALI SUHU INKUBATOR BAYI BERBASIS INTERNET OF THINGS MENGGUNAKAN SENSOR INFRAMERAH GY-906 DENGAN METODE FUZZY LOGIC

EBIN DIONATAN, . (2026) PROTOTYPE SISTEM MONITORING DAN KENDALI SUHU INKUBATOR BAYI BERBASIS INTERNET OF THINGS MENGGUNAKAN SENSOR INFRAMERAH GY-906 DENGAN METODE FUZZY LOGIC. Sarjana thesis, UNIVERSITAS NEGERI JAKARTA.

	Text COVER.pdf Download (1MB)
	Text BAB 1.pdf Download (326kB)
	Text BAB 2.pdf Restricted to Registered users only Download (824kB) | Request a copy
	Text BAB 3.pdf Restricted to Registered users only Download (1MB) | Request a copy
	Text BAB 4.pdf Restricted to Registered users only Download (1MB) | Request a copy
	Text BAB 5.pdf Restricted to Registered users only Download (245kB) | Request a copy
	Text Daftar Pustaka.pdf Download (237kB)
	Text Lampiran.pdf Restricted to Registered users only Download (824kB) | Request a copy

Abstract

Bayi prematur memiliki sistem termoregulasi yang belum matang sehingga rentan mengalami hipotermia maupun kondisi suhu tubuh yang tidak stabil. Inkubator bayi konvensional umumnya masih menggunakan pengaturan suhu manual atau kendali ON-OFF sederhana yang belum menyesuaikan setpoint secara otomatis terhadap berat badan bayi dan kondisi suhu ruang. Penelitian ini merancang sebuah prototipe sistem monitoring dan kendali suhu inkubator bayi berbasis Internet of Things (IoT) yang menggunakan metode Fuzzy Mamdani untuk menentukan setpoint suhu secara otomatis berdasarkan dua variabel input, yaitu berat badan bayi dan suhu ruang inkubator. Basis aturan fuzzy yang disusun berdasarkan sintesis rekomendasi World Health Organization (WHO) dan literatur neonatal selanjutnya dioptimasi menggunakan algoritme Ant Colony Optimization (ACO) untuk memperoleh kombinasi aturan dengan galat prediksi yang lebih kecil. Pengendalian aktuator pemanas Positive Temperature Coefficient (PTC) dan pendingin Peltier dilakukan secara non-fuzzy berbasis ambang batas (threshold), sedangkan kelembapan ruang dikendalikan melalui mekanisme histeresis. Seluruh data hasil pembacaan sensor dan status aktuator dikirimkan ke Firebase Realtime Database dan ditampilkan melalui dashboard web untuk keperluan pemantauan oleh tenaga medis secara real-time. Hasil optimasi ACO menunjukkan penurunan Mean Squared Error (MSE) sebesar 45,97% pada evaluasi weighted-average dan 42,26% pada evaluasi Mamdani penuh (apple-to-apple) dibandingkan basis aturan awal, dengan tiga dari sembilan aturan mengalami penyesuaian pada kondisi suhu ruang ekstrem. Pengujian sensor menunjukkan akurasi tinggi terhadap alat ukur pembanding, pengujian kendali suhu closed-loop pada mode Manual menghasilkan steady-state error rata-rata 0,15°C dengan overshoot di bawah 1%, dan implementasi algoritma pada ESP32 menunjukkan deviasi rata-rata hanya 0,064% terhadap simulasi MATLAB. Hasil keseluruhan membuktikan bahwa sistem yang dirancang mampu menentukan dan mempertahankan setpoint suhu inkubator secara akurat dan stabil, dengan optimasi berbasis ACO meningkatkan akurasi sistem fuzzy dibandingkan basis aturan yang hanya disusun berdasarkan referensi literatur tanpa optimasi lebih lanjut.*****Preterm infants have an immature thermoregulatory system, making them susceptible to hypothermia and unstable body temperature. Conventional baby incubators generally still rely on manual temperature settings or simple ON-OFF control that does not automatically adjust the setpoint to the infant's body weight and room temperature conditions. This research designs a prototype Internet of Things (IoT)-based monitoring and temperature control system for baby incubators using the Fuzzy Mamdani method to automatically determine the temperature setpoint based on two input variables: infant body weight and incubator room temperature. The fuzzy rule base, synthesized from World Health Organization (WHO) recommendations and neonatal literature, was further optimized using the Ant Colony Optimization (ACO) algorithm to obtain a rule combination with a smaller prediction error. The Positive Temperature Coefficient (PTC) heater and Peltier cooler actuators are controlled non-fuzzily based on staged thresholds, while chamber humidity is regulated through a hysteresis mechanism. All sensor reading data and actuator status are transmitted to the Firebase Realtime Database and displayed via a web dashboard for real-time monitoring by medical personnel. The ACO optimization results show a Mean Squared Error (MSE) reduction of 45.97% under weighted-average evaluation and 42.26% under full Mamdani (apple-to-apple) evaluation compared to the initial rule base, with three of nine rules adjusted at extreme room temperature conditions. Sensor testing shows high accuracy against reference measuring instruments, closed-loop temperature control testing in Manual mode yields an average steady-state error of 0.15°C with overshoot below 1%, and the algorithm implementation on the ESP32 shows an average deviation of only 0.064% from the MATLAB simulation. The overall results demonstrate that the designed system is able to accurately and stably determine and maintain the incubator temperature setpoint, with ACO-based optimization improving the accuracy of the fuzzy system compared to a rule base constructed solely from literature references without further optimization.

Item Type:	Thesis (Sarjana)
Additional Information:	1). Rafiuddin Syam, S.T., M.Eng., PhD. ; 2). Churnia Sari, S.T,M.T.
Subjects:	Teknologi dan Ilmu Terapan > Teknik Elektronika
Divisions:	FT > D IV Teknologi Rekayasa Otomasi
Depositing User:	EBIN DIONATAN .
Date Deposited:	12 Aug 2026 04:27
Last Modified:	12 Aug 2026 04:27
URI:	http://repository.unj.ac.id/id/eprint/70049

Actions (login required)

View Item