ABSTRACTS OF ARTICLES OF THE JOURNAL "INFORMATION TECHNOLOGIES".
No. 8. Vol. 24. 2018

DOI: 10.17587/it.24.529-537

A. E. Aleksandrov, Ph. D., Professor, e-mail: femsystem@yandex.ru, Moscow Technological University (MIREA) Moscow, 107996, Russian Federation

Evaluation of the Reliability of Metal Testing Results Based on an Alternative Algorithm

The problem was set to reconstruct the characteristics of the instrument for non-destructive testing and justify the initial defectiveness for using it in construction of mathematical models for calculating and predicting the level of safety of energy facilities. A description of the solution of the direct problem for correcting the distribution of detected defects and constructing their actual distribution is given. A method for solving the inverse problem is developed to reconstruct the characteristics of the instrument used to measure the defectiveness of the objects under investigation on the basis of an alternative algorithm. A technique is proposed for processing the results of defectiveness in order to justify its use in calculating and predicting the level of safety of energy facilities. An alternative algorithm for reconstructing the actual size of the defects and the characteristics of the measuring instruments (the probability function for detecting a defect of a given size) on the results of the performed metal testing is developed on the basis of the revealed feature of the selected distribution functions of the detected and actual defect sizes. For the developed original and alternative algorithms, a numerical study of the effect of input disturbances on the results of the reconstructed parameters is conducted. The alternative algorithm for reconstructing the actual sizes of defects and characteristics of measuring instruments presented in this paper has a lower sensitivity to perturbing input errors than the original algorithm, is more resistant to input errors, and allows obtaining more reliable data for constructing mathematical models for calculating and predicting the safety level of energy objects.
Keywords: probabilistic safety analysis, initial defectiveness, probabilistic methods of fracture mechanics, solution of inverse problems

P. 529–537

To the contents