Indoor Air Quality Control and Intelligent Design Combined with Machine Learning

This paper presents an indoor air quality control strategy integrating machine learning and intelligent design to enhance indoor comfort and health of the indoor environment. Through building a long short-term memory network (LSTM) based prediction model, precise forecasting of multi-dimensional air quality indicators like PM2.5 and carbon dioxide can be ensured, and the mean absolute error is regulated below <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$5 \mu ~\mathrm{g} / \mathrm{m}^{3}$</tex>, providing a sound foundation for real-time control. Intelligent design incorporates sensor networks and air purification and ventilation devices to establish an efficient real-time monitoring and feedback mechanism. Experiment results demonstrate that the system can react to over-standard pollution in less than 6 seconds, and control efficiency increases by about 20 %. Performance analysis of the system under different periods of time demonstrates its stability and adaptability, and its improved running strategy strengthens balance between effectiveness and load. The study proves the synergetic effects of machine learning and intelligent design in air quality control, brings new ideas to the concept of smart buildings, and has certain application prospects.