PRACTICAL APPLICATION OF REGRESSION ANALYSIS FOR INFLUENCE ASSESSMENT OF FACTORS ON THE SENSITIVITY OF THE CAPACITIVE HUMIDITY SENSOR OF THE TWO-LAYER STRUCTURE

Abstract

In the work, the method of planning a multifactorial experiment with regression analysis of the influence of the factor was used to optimize the manufacturing parameters of the capacitive humidity sensor of the two-layer structure. Capacitive humidity sensors of a two-layer structure, made on a sitall substrate with a size of 0.7×0.9 mm, served as experimental samples. A copper film is applied to the surface of the substrate, which forms covers of capacitive elements in the form of a meander. The moisture-sensitive layer in such a structure is a hygroscopic NaCl salt, which performs the function of a dielectric. To create a moisture-sensitive layer, solutions of hygroscopic salt NaCl with concentrations of 0.89 mol/l and 5.33 mol/l were used, which were applied by spraying on the surface of capacitive humidity sensors with thicknesses of 5.0 μm and 10.0 μm. A moisture-absorbing polymethylmethacrylate film with thicknesses of 40 μm and 80 μm served as a protective layer. According to the Cochrane criterion, on the basis of statistical processing of the results of the factorial experiment, it is proved that the factorial experiment is reproducible. A regression equation was developed based on which it was established that the concentration of the salt solution used to create the moisture-sensitive layer significantly affects the sensitivity of the capacitive moisture sensor of the two-layer structure. According to the Student's criterion, significant coefficients of the regression equation were determined. According to Fisher's criterion, it is proved that the adjusted regression equation is adequate to the results of the factorial experiment. The regression equation on the scale of real factors contributes to the optimization of the manufacturing parameters of the capacitive humidity sensor of the two-layer structure with the maximum value of the model's response function − "sensitivity". It was established that the highest sensitivity of 2.36 nF/% is provided with the following optimal parameters of the manufacturing process of a capacitive humidity sensor of a two-layer structure: the concentration of the adsorbing material NaCl is 5.33 mol/l, the thickness of the protective layer is 80 μm. Considering the fact that the thickness of the moisture-sensitive layer does not affect the sensitivity of the capacitive humidity sensor, it is advisable to take the thickness of this layer in the range from 5.0 μm to 10.0 μm.