Klimova K. Improvement of the model method for measurement uncertainty estimation

Research object - the process of measurement uncertainty evaluation for linear and nonlinear model function with arbitrary distribution laws of the input values in the presence or absence of correlation between them. Research objective - improvement of measurement uncertainty estimate reliability. Theoretical and practical outcomes - the methods for estimating of uncertainty in measurement during metrological works were improved, a budget of uncertainty in measurement that takes into account the distribution of the excesses of the input values was developed. Research methods - probability theory and mathematical statistics; measurement theory; statistical modeling techniques. Scientific novelty - the interpolating formula that based on the Student coefficient values for fractional number for degrees of freedom is developed first. The date transpositions method for calculation of unbiased estimate of the measurement results and reliable estimate of the summary "A"-type standard measurement uncertainty for uncorrelated estimations input values with free number of measurements is improved. Also the excess method for calculation the reliable estimates of expanded measurement uncertainty that includes the statistical laws of input values is improved first. The outcomes of dissertational work are implemented in: Scientific and Practical Center of diseases of ear, nose and throat (Kharkov), National Scientific Center "Institute of Metrology", "Nikolaevstandartmetrologiya" State Enterprise, Kharkov National University of Radioelectronics. Scope of application - research organizations, testing and calibration laboratories, institutions of higher education.