Home >
Tools >
Algorithm Development and Model Testing
Algorithm Development and Model Testing
Often, the research of the Airflow and Pollutant Transport group can be advanced by using existing computational tools or by creating
new tools that are based on known computational algorithms. Sometimes, though, our work requires the development of entirely new methods
for problem solving. Algorithms developed in the group include new approaches for:
computed tomography of pollutant concentrations in air ("computed tomography" means mapping spatial distributions of pollutants
by using information about the path-integrated amount along different paths);
computational methods for marrying computational fluid dynamics models to multi-zone building models;
optimizing the solution of nonlinear systems of equations;
and interpreting pollutant sensors in buildings so as to determine the airflow conditions, pollutant release location, etc.
The ability to develop new algorithms as needed is due to the wide range of fields of expertise among group members: the group includes
specialists in computational fluid dynamics, Bayesian data analysis, computational optimization, and other areas.
