Sistem Pemantauan Udara dengan Klasifikasi Kualitas Udara Berbasis KNN
AHNAF NAUFAL, Prof. Dr.techn. Ahmad Ashari, M.I.Kom; Dr. Danang Lelono, S.Si., M.T.
2024 | Skripsi | ELEKTRONIKA DAN INSTRUMENTASI
Air quality is a crucial factor for the life of living beings everywhere. However, the increase in population, motor vehicles, and industries has significantly elevated pollution levels. Pollution at certain concentrations can disrupt human health, even endangering it. Therefore, an air monitoring system that can provide indications of air quality is necessary to educate the surrounding community.
This research develops an air monitoring system using low-cost sensors accompanied by the K-Nearest Neighbor (KNN) algorithm to obtain air quality categories. The pollutant concentrations are measured using MiCS-6814 and ZH03B sensors. The pollutants measured by these sensors are CO, NO2, PM2.5, and PM10. The concentration data obtained from the sensors are then classified using the KNN algorithm based on the Indeks Standar Pencemaran Udara (ISPU).
From the testing of the KNN algorithm for air quality category classification, the highest prediction accuracy obtained is 0.88 at K = 8. Additionally, the implementation of the KNN algorithm on the Wemos D1 Mini takes only 22 to 27 milliseconds. Overall, this research shows that the system can measure pollutant concentrations, determine air quality categories, and be monitored in real-time through Thingspeak. This study offers an effective and affordable solution also portable for air quality monitoring in areas that lack adequate air monitoring facilities.
Kata Kunci : Kualitas Udara, Konsentrasi Polutan, Sensor, K-Nearest Neighbor