The Optimization of Number of Sensors in One-Chip Electronic Nose Microarrays with the Help of 3-Layered Neural Network

We have tested the gas response of multisensor microarrays of "electronic nose" type based on monolithic SnO
₂
film segmented by metal electrodes. With 3-layered neural network we show that not all the sensor segments of the film are necessary to be involved for successful gas identification. By proper choosing the sensors it is possible to significantly decrease the number of sensors without a loss of recognition power.