Neural network model improvements for identification of contaminant source position inside of buildings

9th International Conference and Exhibition - Healthy Buildings 2009, HB 2009

Volumn , Issue , 2009, Pages

  Vukovic, Vladimir ^a   Srebric, Jelena ^a  

^a The Pennsylvania State University   (United States)

Author keywords

Contaminant source position; Multi zone modeling; Neural networks; Validation

Indexed keywords

BACKGROUND CONCENTRATION; CONTAMINANT CONCENTRATIONS; CONTAMINANT DISTRIBUTIONS; CONTAMINANT SOURCES; DISCRETE VALUES; MULTIPLE INPUTS; NEURAL NETWORK MODEL; NON-LINEAR TRANSFORMATIONS; OUTPUT DOMAIN; POLLUTANT CONCENTRATION; POLLUTANT SPECIES; SOURCE POSITION; SOURCE STRENGTH; VALIDATION;

EXHIBITIONS; INVERSE PROBLEMS; MATHEMATICAL TRANSFORMATIONS; NEURAL NETWORKS; POLLUTION;

CONTAMINATION;

EID: 84872648175     PISSN: None     EISSN: None     Source Type: Conference Proceeding    
DOI: None     Document Type: Conference Paper

References (2)

1. Vukovic V., and Srebric J. 2007. Application of Neural Networks Trained with Multi-Zone Models for Fast and Accurate Detection of Contaminant Source Position in Buildings. ASHRAE Transactions, 113(2), 154-162.
2. Yoshino H., Yun Z., Kobayashi H., and Utsumi Y. 1995. Simulation and Measurement of Air Infiltration and Pollutant Transport Using a Passive Solar Test House. ASHRAE Transactions, 101(1), 1091-1099.