Evaluation of the accuracy of processing algorithms for a posteriori measurement results obtained from tripled measuring channels

The problem of assessing the conditional (refined) confidence limits of the measurement error using a posteriori information about the measurement results is considered. A comparative analysis of the accuracy of various algorithms for processing a posteriori measurement results obtained using three measuring channels of the same type is carried out. Algorithms for processing the results of three equally accurate measurements by the arithmetic mean and median values, as well as by the arithmetic mean of the maximum and minimum values are considered. Three parameters were used as a posteriori information: the ratio of the difference between the maximum and minimum values of the results of three measurements to the accuracy indicator of the measuring channel; the ratio of the minimum to the maximum value of the two differences – the maximum and median or median and the minimum value of the three measurement results; product of the above two parameters. The first parameter characterizes the relative spread, the second – the uniformity of the scattering of the results of three measurements, the third – the density and uniformity of the scattering of the results of three measurements. For three parameters, boundary values are found, relative to which the conditional confidence limits of the error of the algorithms are greater or less than the confidence limits of the unconditional error of these algorithms. From the point of view of accuracy indicators, the use of one or another algorithm for processing three equally accurate measurements is justified, depending on the distribution law of their error. Relationships are proposed for assessing the accuracy of the results of processing these measurements.

1. Gamm A. Z., Kolosok I. N., Obnaruzhenie grubyh oshibok teleizmerenij v jelektrojenergeticheskih sistemah [Detection of gross measurement errors in electric power systems], Novosibirsk, Nauka Publ., 2000, 149 p. (In Russ.)

2. Hrapov F. I., Vlasov P. V., Povyshenie dostovernosti rezul’tatov izmerenij v raspredelennyh mnogokanal’nyh informacionno-izmeritel’nyh sistemah, Vestnik metrologa, 2017, no. 4, pp. 9–15. (In Russ.)

3. Anishchenko V. A., Energetika. Proceedings of CIS higher education institutions and power engineering associations, 2017, vol. 60, no. 2, pp 108–117. (In Russ.) https://doi.org/10.21122/1029-7448-2017-60-2-108-117

4. Vlasov P. V., Khrapov F. I., Subota L. S., Cheremina E. M., Measurement Techniques, 2019, vol. 62, no. 9, pp. 754–761. https://doi.org/10.1007/s11018-019-01691-1

5. Mironovskij L. A., Slaev V. A., Algoritmy ocenivanija rezul’tata treh izmerenij [Algorithms for evaluating the result of three measurements], Saint Petersburg, Professional Publ., 2010, 192 p. (In Russ.)

6. Ventcel’ E. S., Ovcharov L. A., Teorija verojatnostej i ejo inzhenernye prilozhenija [Probability theory and its engineering applications], Moscow, Vysch. Shkola Publ., 2007, 298 p. (In Russ.)

7. Shipunov A. B., Baldin E. M., Volkova P. A., Korobejnikov A. I., Nazarova S. A., Petrov S. V., Sufi janov V. G., Nagljadnaja statistika. Ispol’zuem R! [Visual statistics. Use R!], Moscow, DMK Press, 2012, 298 p. (In Russ.)

8. Robert I. Kabakoff , R in Action. Data analysis and graphics with R, N. Y., Manning Publications Co, 2012, 450 p.