Predictive Modeling of Indoor Environmental Parameters for Assessing Comfort Conditions in a Kindergarten Setting

People tend to spend the majority of their time indoors. Indoor air properties can significantly affect humans' comfort, health, and productivity. This study utilizes measurement data of indoor conditions in a kindergarten in Sofia, Bulgaria. Autoregressive integrated moving average (ARIMA) and long short-term memory (LSTM) recurrent neural network (RNN) models were developed to predict CO2 levels in the educational facility over the next hour based on 2.5 h of past data and allow for near real-time decision-making. The better-performing model, LSTM, is also used for temperature and relative humidity forecasting. Global comfort is then estimated based on threshold values for temperature, humidity, and CO2. The predicted R-2 values ranged between 0.938 and 0.981 for the three parameters, while the prediction of global comfort conditions achieved a 91/100 accuracy.

Automated summary

People spend most of their time indoors, and indoor air quality can greatly impact comfort, health, and productivity. This study uses measurement data from a kindergarten in Sofia, Bulgaria to develop models (ARIMA and LSTM) for predicting CO2 levels in real-time. The LSTM model also predicts temperature and humidity, and global comfort is estimated based on threshold values. The models achieve high prediction accuracy for CO2 levels and global comfort conditions.