TY - JOUR AU - Malaichuk, V. AU - Klymenko, S. AU - Astakhov, D. PY - 2023/02/23 Y2 - 2024/03/28 TI - COMPUTER PROCESSING OF MEASUREMENTS IN PROBLEMS OF OBSERVATION OF THE CONDITION OF TECHNICAL OBJECTS JF - Journal of Rocket-Space Technology JA - Rocket-Sp. Tech. VL - 30 IS - 4 SE - Control systems, telecommunications, navigation and automation DO - 10.15421/452213 UR - http://rocketspace.dp.ua/index.php/rst/article/view/147 SP - 99-106 AB - Nonparametric statistics allow us to draw statistical conclusions, in particular, to evaluate the characteristics of the distribution and test statistical hypotheses, without, as a rule, poorly substantiated assumptions that the distribution function of the sample elements is part of a parametric family. For example, there is a widespread belief that statistics are often subject to normal distribution. Meanwhile, the analysis of specific results of observations, in particular, measurement errors, always leads to the same conclusion – in the vast majority of cases, the real distributions are significantly different from normal. In the tasks of non-destructive testing -non-parametric statistics are the basis for research and decision-making on the suitability or quality of products. Despite the large number of publications on specific issues of non-parametric statistics, the internal structure of this scientific field remains a pressing issue. Uncritical use of the normality hypothesis often leads to significant errors, for example, in the rejection of sharply observed results of observations (emissions), in statistical quality control and in other cases. Investigation of classical methods of nonparametric statistics in the problems of monitoring the condition of quality control objects on the basis of measured values. Methods. The most frequently used in practice criteria of nonparametric statistics, such as the Anderson and Smirnov criteria, are considered. And also in article their analogs for use in problems of reliability and quality control at the analysis of the measured sizes are resulted. The theoretical bases of statistical processing of experimental samples of independent random variables are considered in the article. A study of the scientific and technical problem of monitoring and control over the condition and quality of technical objects by comparing experimental samples of measurements in the processes of their development, testing and operation. New criteria for processingexperimental samples of independent random variables have been developed, the use of which, according to experimental results, has shown greater informativeness than the use of the already well-known Anderson and Smirnov criteria. ER -