<![CDATA[With the rapid development of technology, individuals forget about their health, plus during the pandemic, the indoor air quality becomes more of a concern. Maintaining air quality to be healthy and good for humans is by keeping the amount of pollutants in the air, such as Carbon Dioxide (CO2), Volatile Organic Compound (VOC), and Formaldehyde (HCHO), at a predetermined and agreed threshold. We propose an on-board air quality index detection system for indoor and forecast the AQI in the future. The system will use a Raspberry Pi 4 and a WP6003 sensor device that will capture parameters for the AQI. The parameter data is analyzed using a correlation matrix to determine the parameters that affect each other. Then classified using fuzzy logic to determine the quality index based on the value of each parameter. Then forecast using the ARIMA and LSTM methods for the next 30 minutes. The forecasting accuracy is calculated using the RMSE and MAPE metrics. The result shows that CO2, VOC, and HCHO are related. Comparison of the forecasting results of the two methods concluded that the LSTM outperformed ARIMA to forecast the AQI for the next 30 minutes based on the previous 10 hours of data.]]>
Copyrights © 2023
