An IoT-based Discrete Time Markov Chain Model for Analysis and Prediction of Indoor Air Quality Index

Abstract

Humans generally spend most of their time indoors, therefore, having good Indoor Air Quality (IAQ) and its real time information is critical for maintaining human health and productivity. According to United States Environmental Protection Agency, indoor air even in centrally air-conditioned buildings is several times more polluted than outdoor air, primarily due to change in occupancy pattern, old or ill maintained ventilation systems, and cracks in buildings. In this work, we have proposed an Internet of Things (IoT) based Discrete Time Markov Chain (DTMC) model for analysis and forecasting of IAQ. The IoT architecture used for collecting IAQ data consists of sensing nodes deployed in different rooms of the University building. This sensed data is transferred and stored in IoT cloud and used to generate the IAQ state transition matrix and compute return periods for each state. The predicted and actual return periods have been compared and the accuracy of the proposed model is found to be satisfactory with a low average absolute prediction error of 4.75%.