RAHMADI, Lendy and Hadiyanto, Hadiyanto and Sanjaya, Ridwan (2025) PENGEMBANGAN MODEL SIMULASI ОТОМATISASI HIDROPONIK NUTRIENT FILM TECHNIQUE (NFT) SYSTEMS TANAMAN SELADA MENGGUNAKAN HASIL PERBANDINGAN ARIMA DAN PROPHЕТ. Doctoral thesis, UNIVERSITAS DIPONEGORO.

	Text cover.pdf Download (96kB)
	Text COVER LENGKAP.pdf Restricted to Repository staff only Download (1MB)
	Text BAB I.pdf Download (541kB)
	Text BAB II.pdf Download (892kB)
	Text BAB III.pdf Restricted to Repository staff only Download (806kB)
	Text BAB IV.pdf Restricted to Repository staff only Download (2MB)
	Text BAB V.pdf Restricted to Repository staff only Download (227kB)
	Text DAFTAR PUSTAKA.pdf Download (298kB)
	Text LAMPIRAN.pdf Restricted to Repository staff only Download (428kB)

Abstract

Meningkatnya kebutuhan pangan global mendorong inovasi teknologi pertanian modern seperti hidroponik Nutrient Film Technique (NFT), yang memungkinkan pertumbuhan tanaman tanpa tanah secara efisien. Penelitian ini bertujuan mengembangkan model simulasi otomatisasi pertumbuhan tanaman hidroponik NFT dengan fokus pada prediksi pertumbuhan selada pada periode penelitian yang telah ditentukan di awal tahun 2024. Data variabel lingkungan dan nutrisi air, seperti suhu, kelembapan, pH, TDS, EC, dan intensitas cahaya, dikumpulkan selama empat periode panen. Variabel jumlah daun dipilih sebagai indikator utama pertumbuhan karena dapat diukur harian tanpa merusak tanaman dan memiliki korelasi tinggi dengan hasil akhir. Model dibangun menggunakan algoritma ARIMA dan Prophet, dengan evaluasi menggunakan Mean Absolute Error (MAE) dan Root Mean Square Error (RMSE). Hasil menunjukkan Prophet lebih unggul dengan MAE 1,54 dan RMSE 1,84, memberikan prediksi akurat dalam data dinamis. Model ini menggunakan data variabel lingkungan dan jumlah daun sebagai input, sementara output berupa simulasi prediksi jumlah daun dan rekomendasi pengelolaan variabel lingkungan untuk hasil optimal. Model Prophet diintegrasikan ke dalam aplikasi web HydroSim, yang memungkinkan simulasi prediksi real-time dan pengelolaan hidroponik berbasis data. Perspektif otomatisasi dalam penelitian ini mencakup prediksi dan pengelolaan variabel lingkungan secara otomatis untuk menjaga stabilitas sistem. Studi ini mendukung otomatisasi pertanian modern, mengurangi risiko operasional, dan meningkatkan efisiensi. Penelitian lanjutan dapat mencakup data dengan variasi iklim, tanaman lain, dan kombinasi metode machine learning untuk meningkatkan performa dan efektivitas otomatisasi serta dapat melakukan integrasi dengan teknologi IoT dalam melakukan otomatisasi.
Kata kunci: ARIMA, Hidroponik, HydroSim, NFT, Simulasi, Prophet

The increasing global food demand drives innovations in modern agricultural technology such as the Nutrient Film Technique (NFT) hydroponics, which enables efficient soil-less plant growth.This research aims to develop a simulation model for the automation of NFT hydroponic plant growth, focusing on predicting lettuce growth during the specified research period. Environmental and water nutrient variable data, such as temperature, humidity, pH, TDS, EC, and light intensity, were collected over four harvest periods.The variable of leaf count was chosen as the main growth indicator because it can be measured daily without damaging the plants and has a high correlation with the final yield.The model was built using the ARIMA and Prophet algorithms, with evaluation using Mean Absolute Error (MAE) and Root Mean Square Error (RMSE).The results show that Prophet outperforms with an MAE of 1.54 and an RMSE of 1.84, providing accurate predictions in dynamic data.This model uses environmental variable data and leaf count as input, while the output consists of simulations predicting leaf count and recommendations for managing environmental variables for optimal results.The Prophet model is integrated into the HydroSim web application, enabling real-time prediction simulations and data-based hydroponic management.The automation perspective in this research includes the automatic prediction and management of environmental variables to maintain system stability. This study supports modern agricultural automation, reduces operational risks, and increases efficiency.Further research could include data with climate variations, other crops, and combinations of machine learning methods to enhance the performance and effectiveness of automation, as well as the integration with IoT technology in automation.
Kata kunci: ARIMA, Hydroponics, HydroSim, NFT, Simulation, Prophet

Actions (login required)

View Item