Electronic Theses and Dissertation

Ketidakstabilan suhu tubuh merupakan salah satu faktor risiko utama kematian neonatal, khususnya pada bayi prematur yang belum memiliki kemampuan termoregulasi mandiri. kondisi ini menuntut ketersediaan inkubator bayi dengan sistem kendali suhu yang akurat, responsif, dan mampu beradaptasi terhadap karakteristik fisiologis setiap bayi. penelitian ini merancang dan mengimplementasikan prototipe inkubator bayi berbasis internet of things (iot) dengan sistem kendali suhu otomatis menggunakan metode proportional integral derivative (pid) yang dikombinasikan dengan pendekatan gain scheduling. mikrokontroler esp32 digunakan sebagai unit pemroses utama, sensor dht22 sebagai pembaca suhu ruang inkubator, load cell sebagai parameter penentuan setpoint berdasarkan berat bayi, serta pemanas (heater) dan kipas (fan) sebagai aktuator pengendali suhu. parameter pid ditentukan melalui proses autotuning menggunakan library stune dan diterapkan secara adaptif pada empat zona berat badan bayi, yaitu zona 1 (setpoint 36°c) dengan kp = 36.0, ki = 0.10, dan kd = 6.5; zona 2 (setpoint 34°c) dengan kp = 32.0, ki = 0.08, dan kd = 5.8; zona 3 (setpoint 33,5°c) dengan kp = 30.0, ki = 0.06, dan kd = 5.5; serta zona 4 (setpoint 33°c) dengan kp = 28.0, ki = 0.05, dan kd = 5.2. hasil pengujian menunjukkan bahwa sistem mampu mempertahankan suhu inkubator mendekati nilai setpoint dengan rata-rata nilai error suhu sebesar 0,065°c, settling time sebesar 22 menit 39 detik, dan overshoot sebesar 0,27% pada seluruh zona pengujian. integrasi sistem monitoring berbasis iot juga berhasil menampilkan data suhu, berat badan, serta indicator heater dan fan secara real-time. hasil penelitian ini membuktikan bahwa penerapan kendali pid adaptif berbasis gain scheduling mampu menjaga kestabilan suhu pada prototipe inkubator bayi, sekaligus mendukung pemantauan jarak jauh. kata kunci: inkubator bayi, kestabilan suhu, pid, gain scheduling, internet of things

Body temperature instability is one of the major risk factors for neonatal mortality, particularly in premature infants who have not yet developed independent thermoregulatory capability. This condition demands the availability of infant incubators equipped with accurate, responsive temperature control systems capable of adapting to the physiological characteristics of each infant. This research designs and implements an Internet of Things (IoT)-based infant incubator prototype with an automatic temperature control system using the Proportional Integral Derivative (PID) method combined with a Gain Scheduling approach. An ESP32 microcontroller is used as the main processing unit, a DHT22 sensor as the incubator chamber temperature reader, a load cell as the parameter for setpoint determination based on infant weight, and a heater and fan as the temperature control actuators. The PID parameters were determined through an autotuning process using the sTune library and applied adaptively across four infant weight zones, namely Zone 1 (setpoint 36°C) Kp = 36.0, Ki = 0.10, dan Kd = 6.5; Zone 2 (setpoint 34°C) with Kp = 32.0, Ki = 0.08, dan Kd = 5.8; Zone 3 (setpoint 33.5°C) with Kp = 30.0, Ki = 0.06, dan Kd = 5.5; and Zone 4 (setpoint 33°C) with Kp = 28.0, Ki = 0.05, dan Kd = 5.2. The test results show that the system was able to maintain the incubator temperature close to the setpoint value, with an average temperature error of 0.065°C, a settling time of 22 minutes and 39 seconds, and an overshoot of 0.27% across all test zones. The integration of the IoT-based monitoring system also successfully displayed temperature, weight, and heater/fan status indicators in real time. These findings demonstrate that the implementation of adaptive PID control based on Gain Scheduling is able to maintain temperature stability in the infant incubator prototype while supporting remote monitoring. Keywords: Infant Incubator, Temperature Stability, PID, Gain Scheduling, Internet of Things