Indoor Air Quality Predictions For Automation

Abstract: This study examines the implementation of home automation systems to predict indoor air quality using real-time data such as temperature, humidity, pressure, occupancy status, energy consumption, and window conditions. Due to the superior pattern recognition performance of recurrent neural networks, the study employs deep learning techniques for air quality prediction. A comparative analysis of GRU, LSTM and BiGRU models highlights GRU’s superior performance across various metrics, emphasizing its generalization capability. The study also introduces an Air-Smart Control Device, enabling users to monitor predictions and control home automation systems. In conclusion, the research underscores the potential of home automation in air quality prediction, provides insights into neural network architectures, and contributes to advancements in automation technology and air quality management.Keywords: Automation System, Air Quality, Deep Learning, Machine Learning