Data-analytics Based Monitoring Approach for Indoor Air Quality of Urban Railway System

초록

Urban railway system has been considered an air quality hot spot, since a large portion of the urban railway system are constructed underground and air pollutants are originated from internal sources. Recently, sensors have been used to monitor indoor air quality (IAQ) of the railway systems, however, they suffer reliability problems due to a hostile environment where the sensors are installed. An objective of this study is to develop a data-analytics based monitoring approach for the railway system IAQ. The data-analytics based monitoring approach has two specific purposes: (1) to investigate influence of physical/operational properties of the railway system on particulate matters, and (2) to detect, identify and reconstruct IAQ sensor faults. For this, a multivariate statistical analysis which extracts essential data variance by reducing data dimensionality (e.g., principal component analysis (PCA), independent component analysis (ICA)) is applied.