Improvement of selectivity and sensitivity of odor sensor using stochastic resonance and subspace method

本研究旨在探讨如何改善气味感测系统的选择性与灵敏度。随机共振可以提高气体传感器的灵敏度，然后我们考虑了系统参数出现随机共振的数值模拟，特别是，在系统中的非线性组件。提出了基于子空间的Prony方法，我们已经表明，该方法可以提取多组分气体的传感器响应的特征，这些特征可以用于提高选择性，并且还可以给出输入气味中包含的气体组分的数量的信息。

In this study, an improvement of selectivity and sensitivity of odor sensor system are investigated. The stochastic resonance can be expected to improve a sensitivity of a gas sensor, then we have considered the system parameter to appear the stochastic resonance by numerical simulations, especially, in terms of the nonlinear component in the system. The subspace-based Prony's method is proposed and we have shown that the method can extract features of a sensor response to multicomponent gasses which can be used to improve a selectivity and also give an information of the number of the gas components contained in an input odor.