摘要
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 prediction errors for ethane, propane and propylene were 5. 11%, 8.28%, 2. 64%, respectively, in the ANN prediction.
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 prediction errors for ethane, propane and propylene were 5. 11%, 8.28%, 2. 64%, respectively, in the ANN prediction.
出处
《高等学校化学学报》
SCIE
EI
CAS
CSCD
北大核心
1997年第6期886-888,共3页
Chemical Journal of Chinese Universities
基金
国家自然科学基金
福建省自然科学资助
