Modeling and measurement accuracy enhancement of flue gas flow using neural networks

This paper discusses the modeling of the hue gas how in industrial ducts an d stacks using artificial neural networks (ANN's), Based upon the individua l velocity and other operating conditions, an ANN model has been developed for the measurement of the volume flow rate. The model has been validated b y the experiment using a case-study power plant. The results have shown tha t the model can largely compensate for the nonrepresentativeness of a sampl ing location and, as a result, the measurement accuracy of the flue gas flo w can be significantly improved.