Uncertainty of estimating the characteristic parameter of the two-component mixture

Abstract

The issues of estimating the quality of two-component mixtures, which is determined on the basis of a characteristic parameter, are considered. The value of the characteristic parameter is determined based on the measurement results of component parameters, followed by the calculation according to the model equation.
 Since the estimation is carried out in so-called industrial conditions, a parallel-serial organization of the measurement procedure was chosen, in which a common measurement channel (MC) is used. With such an organization, the accuracy of the measurement is mainly determined by the characteristic bias of the conversion characteristic of the MC. Performing computational operations on the measurement results leads to instrumental correlation, which makes a certain contribution to the combined standard uncertainty when estimating a characteristic quantity, which, for example, in the confectionery industry and construction, is moisture.
 Research of the reproduction accuracy of a two-component mixture was carried out.
 To reduce the influence of moisture deviation of the components, the calibration was carried out to compare the calculated value of the characteristic quantity with the model one.
 The uncertainty of estimating the characteristic parameter during the control measurements when monitoring the stability of the technological process is analyzed. It was established that the presence of instrumental correlation increases the uncertainty of estimating the characteristic value by two times. To increase the accuracy of the estimation, a method and its structural implementation are proposed, which allows significant reducing the combined standard uncertainty.