Dual Chamber Bionic Olfactory System Research Based On Nasal Mucous Chromatography

With the rapid development of sensor technology, electronics technology, computer technology and pattern recognition technology, artificial olfaction technology has been widely used in the field of food inspection, environmental monitoring, disease diagnostics and fire alarm. However, comparing to biological olfactory system, the electronic instruments have a small number of gas sensors which are instable (susceptible to environmental factors), poor cross-selective and easy poisoned, seriously influent the recognition performance, leading to the instruments of the detection accuracy and real-time far less to that of artificial. Furthermore, what sets biological olfactory apart from artificial olfactory system is the active biological olfactory perception process (Active Olfactory), i.e., by means of dynamically adjusting the gas velocity and the gas flow through the cellular sensors array, so that the odor presented complex spatiotemporal patterns at the cell surface of the sensor array, providing pretreatment for subsequent processing in olfactory cortex of brain. In the 1960’s, Mozell conducted a series of studies for olfactory and proposed the hypothesis of "nasal chromatographic" effects, that is above the olfactory perception of dampers similar to GC effect, and the purpose is to separate different odor molecules in different regions of the olfactory mucosa, making all kinds of odors to compose different spatial patterns.In this paper, just inspired by the olfactory hypothesis mentioned above, and characterizes a bionic nasal olfactory system based on chromatographic effect with the idea of bionic electronic mucosa. The sensor arrays are distributed in two chambers which connected by the delay channel to simulate the nasal mucosa in two different partitions. The system includes odor sampling configuration, pneumatic controlled, sensor array driver module and data acquisition subsystem, and the system control and data acquisition are completed by NI virtual instrument configuration and software LabView. This novel detection instrument generates a kind of transient response information which called "spatio-temporal" data, and it is the gas chromatographic effect simulation of the model of biological nasal mucosa which called "nasal chromatography" effect. Then, coupling with an appropriate pattern recognition algorithm, the odor discrimination ability of instrument will be highly enhanced. And the experiments emphatically focus on the result of the changing in reaction chambers, retentive channel, odor flow and other parameters. The spatio-temporal information is extracted and analyzed in the system validation experiments, which shows a successful discrimination between four simple odors by PCA and other classical recognition algorithms. Meanwhile, Fluent of the fluid simulation software is also been used to simulate the airflow and odor distribution inside the chambers. The experiment results demonstrate that the system experimental results are same to the simulations, which means the system has successfully simulated the process of the "nose chromatography". Comparing to the single chamber system, the double-chamber system has a better performance, which has be seen as a novel construction method of gas chamber and sensor array of electronic nose.