Rancang Bangun Sistem Monitoring Kualitas Udara Berdasarkan Indeks Standar Pencemaran Udara Berbasis IoT dan Machine Learning

Armandito, Axellino Anggoro (2024) Rancang Bangun Sistem Monitoring Kualitas Udara Berdasarkan Indeks Standar Pencemaran Udara Berbasis IoT dan Machine Learning. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5027201040-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2026. Download (2MB) | Request a copy

Abstract

Polusi udara telah menjadi masalah lingkungan yang mendesak di seluruh dunia karena memiliki dampak signifikan terhadap kesehatan manusia. Menurut Indeks Standar Pencemaran Udara (ISPU), terdapat 7 parameter partikulat dan gas, seperti PM10, PM2.5, karbon monoksida (CO), nitrogen dioksida (NO2), sulfur dioksida (SO2), ozon (O3), dan hidrokarbon (HC). Untuk mengatasi permasalahan ini penulis mengajukan sistem pemantauan kualitas udara yang dapat diakses secara real time yang mampu mendeteksi 6 parameter gas, yaitu PM1, PM10, PM2.5, CO, NO2, dan O3. Sistem ini juga dilengkapi dengan model kecerdasan buatan algoritma eXtreme Gradient Boosting (XGBoost) untuk memprediksi 6 polutan gas di dalam ruangan selama 1 jam kedepan. Arsitektur sistem mengintegrasikan mikrokontroler yang terhubung dengan sensor untuk mengukur dan mengirimkan data ke MQTT Broker, memungkinkan web server untuk mengolah data sesuai pedoman ISPU. Model algoritma XGBoost menunjukkan performa dengan akurasi yang lebih baik dibandingkan Random Forest dan SVM menggunakan default parameter dengan hasil CO memiliki MAE sebesar 3,5166, RMSE 4,4629, dan MAPE 1,2259%. NO2 menunjukkan MAE 0,0491, RMSE 0,0541, dan MAPE 0,3865%, O3 memiliki MAE 1,5457, RMSE 1,8233, dan MAPE 2,3848%, PM1 memiliki MAE 1,1789, RMSE 1,3505, MAPE 2,9976%, PM2.5 memiliki MAE 1,6855, RMSE 1,9016, MAPE 3,4948%, dan PM10 memiliki MAE 1,5301, RMSE 1,7601, dan MAPE 2,7949%.
=================================================================================================================================
Air pollution has become an urgent environmental issue worldwide due to its significant impact on human health. According to the Air Pollution Standard Index (ISPU), there are 7 particulate and gas parameters, such as PM10, PM2.5, Carbon Monoxide (CO), Nitrogen Dioxide (NO2), Sulfur Dioxide (SO2), Ozone (O3), and Hydrocarbon (HC). To address this issue, the authors propose a real-time accessible air quality monitoring system capable of detecting 6 gas parameters, namely PM1, PM10, PM2.5, CO, NO2, and O3. This system is also equipped with an Artificial Intelligence model using the eXtreme Gradient Boosting (XGBoost) algorithm to predict 6 indoor gas pollutants for the next hour. The system architecture integrates a microcontroller connected to sensors for measuring and transmitting data to the MQTT Broker, enabling the web server to process data according to the ISPU guidelines. Using default parameters, the XGBoost algorithm model demonstrates better performance accuracy than Random Forest and SVM, with CO yielding an MAE of 3.5166, RMSE of 4.4629, and MAPE of 1.2259%. NO2 exhibits an MAE of 0.0491, RMSE of 0.0541, and MAPE of 0.3865%, O3 shows an MAE of 1.5457, RMSE of 1.8233, and MAPE of 2.3848%, PM1 has an MAE of 1.1789, RMSE of 1.3505, MAPE of 2.9976%, PM2.5 has an MAE of 1.6855, RMSE of 1.9016, MAPE of 3.4948%, and PM10 has an MAE of 1.5301, RMSE of 1.7601, and MAPE of 2.7949%.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Internet of Things, MQTT, Kualitas Udara, Dashboard Monitoring, ISPU, Air Quality, XGBoost.
Subjects:	T Technology > T Technology (General) > T174 Technological forecasting T Technology > T Technology (General) > T57.8 Nonlinear programming. Support vector machine. Wavelets. Hidden Markov models.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information Technology > 59201-(S1) Undergraduate Thesis
Depositing User:	Axellino Anggoro Armandito
Date Deposited:	06 Feb 2024 02:26
Last Modified:	06 Feb 2024 02:26
URI:	http://repository.its.ac.id/id/eprint/106154

Actions (login required)

View Item