Development of data correction model for low-cost particle sensors under ambient conditions

点击翻译按钮获取中文摘要

The advancement of the Internet of Things (IoT) technology makes it possible for mapping air qualityvariations in large spatial scale regions. Deployment of such technology requires large number low-costsensors. To monitor PM2.5 (suspended aerosol particles with an aerodynamic diameter no more than 2.5micrometers) distribution in a large area such as the urban area of a city, hundreds of low-cost PM2.5 sensorsare required. Such sensors mostly are based on light scattering technology, and the readings were calibratedwith standard aerosols by their manufacture. The calibration curves are only valid for certain aerosols undercertain conditions (e.g., humidity, temperature, etc.), which makes their readings questionable when comparingwith the results from mass or beta-attenuation -based technologies used by environmental monitoring agencies.Therefore, there is need to calibration the low-cost PM2.5 sensors against the certified technologies byenvironmental authorities under the real-world conditions so that the sensor network could provide acceptablemonitoring data for air quality assessment.SensorUp is a leading sensor network provider based on IoT technology that has deployed an air qualitymonitoring network consisting of 200+ PM2.5 sensors to cover 75% of the area of metropolitan Calgary. Thecompany is interested in an in-depth, quantitative study of the impacts of aerosols properties andmeteorological factors on the readings of their sensors and developing a method to calibrate the sensors so thatthey can provide comparable reading with the certified instrument being used by Calgary Regional AirshedZone (CRAZ). The outcome of this study will not only improve SensorUp's capacity in PM2.5 monitoring, butthe method is also transferable to other instrument/pollutant providing environmental monitoring agencies andservice providers a low-cost option for conducting high-resolution monitoring.