Abstract

The use of an artificial neural network (ANN) as an automatic parameter estimation tool, to infer optimal Kv values in urban air quality, is discussed. An ANN computer system is implemented that has embedded a state-of-the-science air quality model (CAM). The ANN is trained with the meteorological information to infer Kv values. The application of the ANN system to data sets showed the distinctive property of neural network, and provide functional approximations capable of being generalized to other cases.