Application of artificial neural networks for hydrocarbon gas mixture analysis
- 化学化工－已发表论文 
An array composed of sixtorganiceen metal oxide semiconductor gas sensors was constructed to analyze gas mixtures quantitatively. The responses of the sensor array to ethane, propane and propylene were treated by three-layer artificial neural networks (ANN) with the method of error back-propagation and partial least-squares (PLS), The pattern recognition results indicated that the concentration predicted with ANN is better than that with. PLS, The average predict-ion errors for ethane, propane and propylene were 5.11%, 8.28%, 2.64%, respectively, in the ANN prediction.