Analysis of chemical exposure through inhalation using hybrid neural network

摘要

In this analysis, human health risk through inhalation due to exposure to Benzene from vehicular emissions in New Zealand is assessed as an example of the application of a hybrid neural network. Exposure factors affecting the inhalation are inhaled contaminant, age, body weight, health status and activity patterns of humans. There are four major variables affecting the inhaled contaminant viz., gas emissions from motor vehicles on the road, wind speed, temperature and atmospheric stability. The topic of uncertainty applies equally to all variables involved in exposure analysis. Neural network and fuzzy theory is implemented to solve the uncertainty, which exists to a greater extent. The architecture of hybrid neural network that is used to estimate the exposure of carcinogens through inhalation is explained in detail in this paper.