Generate QR code
Open Access
Subscription Access
Comparison of criteria for identification of mathematical models in solving measurement problems
https://doi.org/10.32446/0368-1025it.2022-8-41-45
Abstract
The problem of instability of the results of structural-parametric identification of metrological characteristics of a functional type during calibration and calibration of measuring instruments using various identification criteria with a small number of measurements is described. These characteristics include calibration diagrams, conversion functions, calibration characteristics, error distribution functions, etc. It is shown that the results of solving the problem of identifying metrological characteristics of a functional type are affected by the order of splitting the data of joint measurements into blocks in the cross-observation scheme of the maximum compactness method. The comparison of the criteria of structural identification – the minimum of the average modulus of the inadequacy error and the maximum of the Kappa criterion (probability of agreement) is carried out on the example of the thermometer calibration problem. It is established that if it is impossible to divide the data of joint measurements into equal parts, the kappa criterion has a higher stability of the results. The results can be applied when writing specialized automatic data processing programs for calibration and calibration of measuring instruments, unambiguous identification of metrological characteristics of a functional type according to measurement protocols, for example, when performing certification of measurement methods.
References
1. Levin S. F., Blinov A. P., Measurement Techniques, 1988, vol. 31, no. 12, pp. 1145–1150. https://doi.org/10.1007/BF00862607
2. Levin S. F., Lisenkov A. N., Senko O. V., Kharatyan E. I., System metrological support of static measurement tasks “MMKstat M”, User’s guide, Moscow, Gosstandart, AN SSSR Publ., 1998. (In Russ.)
3. Gogin S. S., Measurement Techniques, 2006, vol. no. 7, pp. 657–659. https://doi.org/10.1007/s11018-006-0165-x
4. Suleiman I. A., Measurement Techniques, 2012, vol. 55, no. 1, pp. 37–40. https://doi.org/10.1007/s11018-012-9913-2
5. Nevskaya E. E., Measurement Techniques, 2017, vol. 60, no. 1, pp. 15–18. https://doi.org/10.1007/s11018-017-1142-2
6. Levin S. F., Measurement Techniques, 2018, vol. 61, no. 6, pp. 528–539. https://doi.org/10.1007/s11018-018-1462-x
7. Chikmarev A. D., Measurement Techniques, 2020, vol. 63, no. 9, pp. 686–689. https://doi.org/10.1007/s11018-021-01840-5
8. Levin S. F., Mathematical theory of measurement problems, Test and Measuring Instruments and Systems, 2006, no. 3, pp. 23–34. (In Russ.)
9. Davidova T. V, Bogdanova U. U., Using the statistical “folding knife” method to assess reliability indicators in a small sample, Proc. Сonference, Military Academy of Military Air Defense of the Armed Forces of the Russian Federation named after Marshal of the Soviet Union A. M. Vasilevsky, 2017, pp. 327–333. (In Russ.)
10. Feckovich I. V., Popova V. B., Jacknife method in statistical analysis of agricultural activity, Economic analysis: theory and practice, 2012, vol. 11, no. 20, pp. 32–36. (In Russ.)
11. Jian Shi, Yiwei Zhang, Haiwei Luo, Jijun Tang, BMC Bioinformatics, 2010, vol. 11, 168. https://doi.org/10.1186/1471-2105-11-168
12. Levin S. F., Measurement Techniques, 2021, vol. 64, no. 4, pp. 273–281. https://doi.org/10.1007/s11018-021-01929-x
Review
For citations:
Chikmarev A.D. Comparison of criteria for identification of mathematical models in solving measurement problems. Izmeritel`naya Tekhnika. 2022;(8):41-45. (In Russ.) https://doi.org/10.32446/0368-1025it.2022-8-41-45
Views:
185
Email this article 