Kruglenko I. The mechanisms for increase of identification ability of multichannel systems of gas analysis in the dynamic mode.

The thesis deals with investigation of mechanisms for increase of identification ability of multisensor arrays in the dynamic mode. A comprehensive study of the main mechanisms of adsorption of multicomponent gas substances on surfaces of organic molecular crystals enabled to determine interrelation between the classification ability of an array, the features of proceeding of adsorption-desorption processes at interfaces and the nature of sensitive layers. This made it possible to exert purposeful control over chemical functionality of a sensor array by optimizing both type and number of sensors in an array using the most informative part of sensor response. It is shown that the procedure of optimization on the basis of the cluster analysis in the fuzzy logic concept enables one to estimate expedience of use of various sensors in an array and optimize it for specific applications. Application of nanostructure materials (zeolites) was shown to be expedient for increase of classification ability when analyzing multicomponent media with overbalance of any of the active mixture components. It is found that application of physical and physico-chemical techniques for restoration of sensitive elements (treatment of a sensor array with alcohol vapors) considerably improves system reproducibility, thus providing formation of stable chemical images. Thus the work determined and substantiated the mechanisms for kinetic discrimination in multichannel systems of gas analysis. This made it possible to determine both physical and physico-chemical conditions for minimization of the effect of external factors on the functional properties of sensor array.