Localized indoor air quality monitoring for indoor pollutants’ healthy risk assessment using sub-principal component analysis driven model and engineering big data

Abstract

Indoor air quality (IAQ) in subway systems shows periodic dynamics due to the number of passengers, train schedules, and air pollutants accumulated in the system, which are considered as an engineering big data. We developed a new IAQ monitoring model using a sub-principal component analysis (sub-PCA) method to account for the periodic dynamics of the IAQ big data. In addition, the IAQ data in subway systems are different on the weekdays and weekend due to weekly effect, since the patterns of the number of passengers and their access time on the weekdays and weekend are different. Sub-PCA-based local monitoring was developed for separating the weekday and weekend environmental IAQ big data, respectively. The monitoring results for the test data at the Y-subway station clearly showed that the proposed method could analyze an environmental IAQ big data, improve the monitoring efficiency and greatly reduce the false alarm rate of the local on-line monitoring by comparison with the multi-way PCA.