Карпова, Л.В.Горошко, А.В.Пирожок, В.В.Karpova, L.Goroshko, A.V.Pyrozhok, V.V.2020-12-262020-12-262020Карпова Л. В. Статистична обробка результатів вимірювань характеристик міцності керамічних резисторів з полімодальною щільністю розподілу / Л. В. Карпова, А. В. Горошко, В. В. Пирожок // Вимірювальна та обчислювальна техніка в технологічних процесах. – 2020. – № 1. – С. 34-40.https://elar.khmnu.edu.ua/handle/123456789/9724В статті представлені результати досліджень щодо підвищення міцності герметичних мікромодулів, направлених на виявлення причин розкиду значень характеристик міцності керамічних резисторів, виготовлених різними виробниками. Для статистичної обробки результатів вимірювань міцнісних характеристик резисторів запропоновано метод обробки полімодальних сумішей щільності розподілу ймовірностей.Much of the electronic equipment (REA) is operated in conditions of changes in atmospheric pressure, temperature, humidity, vibration and other destabilizing factors. The elements of such REA are subjected to static and dynamic mechanical load. According to various data, about 40% of such radio elements fail due to mechanical failure. In contrast to the field of mechanical engineering, in instrument making for some structures of CEA elements there are no strength standards and in general insufficient attention is paid to ensuring the strength of structures. One of the reasons for this state of affairs is the use of new materials with insufficiently studied physical and mechanical characteristics in the production of CEA. Due to objective and subjective reasons, the parameters measured on a real object, which characterize the quality of the electronic system, strength characteristics of materials of CEA elements, properties of materials, etc., usually have a scatter of values, ie can acquire arbitrary values in some numerical intervals. . Studies show that the variance of the values of such characteristics can be hundreds of percent, which often does not allow to provide a given strength of the structures of the elements of CEA [2-6]. One of the urgent tasks on the way to solving the problem of increasing the reliability of REA is the development and improvement of experimental methods for testing REA, in particular methods of statistical processing of measurement results. Having data on the implementation of these random variables, you can accurately estimate their true values, such as the method of confidence intervals. The article presents the results of studies on increasing the strength of sealed micromodules aimed at identifying the causes of the spread in the strength characteristics of ceramic resistors manufactured by various manufacturers. For statistical processing of the results of measuring the strength characteristics of resistors, a method for processing polymodal mixtures of the probability distribution is proposed. The method consists in representing the mixture in the form of a superposition of Gaussian laws, followed by the decomposition of the mixture. Proposed effective methods of decomposition. The method was applied to determine the strength characteristics of resistors manufactured by three different factories. It has been substantiated that the features of production significantly affect the spread of values in the characteristics. The method has been tested and shown to be satisfactory. The developed method of statistical processing is more versatile than the existing parametric methods. In the case when empirical data formed under the influence of several dominant reasons are subject to assessment, and it is impossible to identify these reasons and divide the sample into corresponding subsamples, the method is more accurate than the existing ones.ukстатистична обробкащільність розподілусуміш розподілівзакон Гауссаstatistical processingdistribution densitymixture of distributionsGauss's lawСтаття519.25