Predicting airborne pollutant concentrations and events in a commercial building using low-cost pollutant sensors and machine learning: A case study

Mohammadshirazi, Ahmad; Kalkhorani, Vahid Ahmadi; Humes, Joseph; Speno, Benjamin; Rike, Juliette; Ramnath, Rajiv; Clark, Jordan

doi:10.1016/j.buildenv.2022.108833

Cited by 14 publications

(9 citation statements)

References 70 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Zhang et al also developed another approach using deep learning model and tested it in an occupied classroom [ 42 ]. Similarly, Mohammadshirazi et al also used a LSTM deep learning approach to predict formaldehyde concentration in an occupied office based on historical measured data [ 43 ] and compared to other three forecasting models: rolling average, Random Forest, and Gradient Boosting. The data-driven methods do not require detailed mass transfer parameters of the emitting materials, but the approaches typically need massive data for training and the approaches have not been applied to any residential buildings, where the environmental and occupancy pattern are more complex than commercial buildings.…”

Section: Introductionmentioning

confidence: 99%

A Time-Varying Model for Predicting Formaldehyde Emission Rates in Homes

Zhao

Walker

Sohn

et al. 2022

IJERPH

View full text Add to dashboard Cite

Recent studies have succeeded in relating emissions of various volatile organic compounds to material mass diffusion transfer using detailed empirical characteristics of each of the individual emitting materials. While significant, the resulting models are often scenario specific and/or require a host of individual component parameters to estimate emission rates. This study developed an approach to estimate aggregated emissions rates based on a wide number of field measurements. We used a multi-parameter regression model based on previous mass transfer models to predict formaldehyde emission rate for a whole dwelling using field-measured, time-resolved formaldehyde concentrations, air exchange rates, and indoor environmental parameters in 63 California single-family houses built between 2011 and 2017. The resulting model provides time-varying formaldehyde emission rates, normalized by floor area, for each study home, assuming a well-mixed mass balance transport model of the home, and a well-mixed layer transport model of indoor surfaces. The surface layer model asserts an equilibrium concentration within the surface layer of the emitted materials that is a function of temperature and RH; the dwelling ventilation rate serves as a surrogate for indoor concentration. We also developed a more generic emission model that is suitable for broad prediction of emission for a population of buildings. This model is also based on measurements aggregated from 27 homes from the same study. We showed that errors in predicting household formaldehyde concentrations using this approach were substantially less than those using a traditional constant emission rate model, despite requiring less unique building information.

show abstract

Section: Introductionmentioning

confidence: 99%

A Time-Varying Model for Predicting Formaldehyde Emission Rates in Homes

Zhao

Walker

Sohn

et al. 2022

IJERPH

View full text Add to dashboard Cite

show abstract

“…These studies report mixed levels of success and may not be generalizable to other environments. In particular, Mohammadshirazi et al 22 reported poor performance when predicting concentrations of PM 2.5 in an office space and further uncertainty would be likely in less predictable residential environments. Additional concerns for predictive automation include the time resolution that can be used given the present performance of low‐cost particle sensors, particularly issues around short‐term fluctuations in measured concentrations, 22,24 and possible reliance on external data (e.g., outdoor PM 2.5 concentrations) that requires internet connectivity and may suffer from data quality and availability issues in some locations.…”

Section: Introductionmentioning

confidence: 99%

“…Additional concerns for predictive automation include the time resolution that can be used given the present performance of low-cost particle sensors, particularly issues around short-term fluctuations in measured concentrations, 22,24 and possible reliance on external data (e.g., outdoor PM 2.5 concentrations) that requires internet connectivity and may suffer from data quality and availability issues in some locations.…”

Section: Introductionmentioning

confidence: 99%

“…Recently, studies have been published which use machine learning techniques to predict future concentrations of indoor air pollutants 21,22 and temperatures 23 for the purpose of automating building systems. These studies report mixed levels of success and may not be generalizable to other environments.…”

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Evaluation of fixed and adaptive concentration thresholds for particle filter systems

2022

View full text Add to dashboard Cite

Particle filtration can effectively reduce indoor concentrations of particulate matter (PM) but may incur high energy use. This study evaluates fixed and adaptive concentration thresholds to automate the operation of filtration systems. Simulated environments were derived from week-long continuous PM measurements from Dylos DC1700 (N = 104) and Alphasense OPC-N2 (N = 100) particle counters deployed in apartments in Toronto. A fixed threshold of 4.0 μg•m −3 resulted in a mean air cleaner runtime of 6.9%-21.0% depending on clean air delivery rate (CADR) and particle sensor, while providing mean concentration reductions of 67%-71% compared to operating the air cleaner constantly (runtime = 100%). In most environments, runtime could be further reduced by raising the fixed threshold while resulting in only a modest decrease in absolute and normalized mean exposure reduction. Using an adaptive threshold derived from a k-means clustering approach generally provided substantial exposure reduction while preventing high runtimes. These results were generally insensitive to cleaning power and the monitor used to measure particle concentrations.Reducing the energy usage of particle filter systems will make them a more viable and sustainable means of improving occupant health.

show abstract