Gozali, Muhammad Fahmi Gymnastiar and Dahnial Syauqy, S.T., M.T., M.Sc., and Wijaya Kurniawan, S.T., M.T., (2024) Klasifikasi Kelayakan Konsumsi Susu Kambing Etawa asteurisasi Berdasarkan Warna, Aroma, dan pH dengan Metode Random Forest Berbasis Arduino. Sarjana thesis, Universitas Brawijaya.
Abstract
Peningkatan konsumsi produk sehat dan alami menjadikan susu kambing Etawa pasteurisasi pilihan populer di masyarakat, karena kandungan nutrisinya yang kaya seperti protein, kalsium, dan fosfor. Namun, susu pasteurisasi rentan mengalami penurunan kualitas jika penyimpanan tidak sesuai, yang ditandai dengan perubahan warna, aroma, dan pH. Penilaian manual seperti metode organoleptik, yang melibatkan penciuman aroma, pengamatan warna, dan pengecapan rasa, sering kali bersifat subjektif dan kurang akurat. Penelitian ini bertujuan merancang sistem otomatis untuk menentukan kelayakan konsumsi susu kambing Etawa pasteurisasi berdasarkan parameter warna, kadar gas amonia, dan pH menggunakan algoritma Random Forest berbasis Arduino. Sistem menggunakan sensor warna TCS3200 untuk membaca nilai RGB, sensor gas MQ135 untuk mendeteksi kadar gas ammonia (NH3), dan sensor PH4502C untuk mengukur pH. Data yang diperoleh diproses menggunakan model Random Forest yang sudah diimplementasikan pada Arduino Mega 2560 dan diklasifikasikan menjadi tiga kategori: "Layak Konsumsi", "Segera Konsumsi", dan "Tidak Layak". Pengujian dilakukan dengan menggunakan 112 data latih dan 28 data uji untuk mengevaluasi akurasi sistem. Hasil klasifikasi ditampilkan melalui LCD 16×2 I2C beserta data sampel yang ditampilkan diatas hasil klasifikasi kelayakan konsumsi susu kambing etawa pasteurisasi. Hasil penelitian menunjukkan bahwa sistem mampu mengklasifikasi kelayakan konsumsi susu kambing Etawa dengan akurasi 92% dan waktu komputasi rata-rata 1.076 detik.
English Abstract
The increasing consumption of healthy and natural products has made pasteurized Etawa goat milk a popular choice in the community, due to its rich nutritional content such as protein, calcium, and phosphorus. However, pasteurized milk is prone to deterioration in quality if stored improperly, which is characterized by changes in color, aroma, and pH. Manual assessments such as organoleptic methods, which involve smelling aromas, observing colors, and tasting flavors, are often subjective and less accurate. This research aims to design an automated system to determine the consumption feasibility of pasteurized Etawa goat milk based on color parameters, ammonia gas levels, and pH using the Arduino-based Random Forest algorithm. The system uses TCS3200 color sensor to read RGB value, MQ135 gas sensor to detect ammonia (NH3) gas level, and PH4502C sensor to measure pH. The data obtained is processed using Random Forest model that has been implemented on Arduino Mega 2560 and classified into three categories: "Worth Consuming", "Consume Immediately", and "Not Worth Consuming". Tests were conducted using 112 training data and 28 test data to evaluate the accuracy of the system. Classification results are displayed through a 16×2 I2C LCD along with sample data displayed above the classification results of the feasibility of pasteurized Etawa goat milk consumption. The results showed that the system was able to classify the feasibility of Etawa goat milk consumption with an accuracy of 92% and an average computation time of 1.076 seconds.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | 052415 |
| Uncontrolled Keywords: | susu kambing Etawa pasteurisasi, klasifikasi kelayakan konsumsi, sensor, arduino mega2560, random forest, akurasi sistem |
| Divisions: | Fakultas Ilmu Komputer > Teknik Komputer |
| Depositing User: | S Sucipto |
| Date Deposited: | 13 Feb 2025 01:38 |
| Last Modified: | 13 Feb 2025 01:38 |
| URI: | http://repository.ub.ac.id/id/eprint/236532 |
![[thumbnail of DALAM MASA EMBARGO]](http://repository.ub.ac.id/style/images/fileicons/text.png) |
Text (DALAM MASA EMBARGO)
Muhammad Fahmi Gymnastiar Gozali.pdf Restricted to Registered users only Download (5MB) |
Actions (login required)
 |
View Item |