RIVS Research and Production, Saint-Petersburg, Russia:
A. V. Bondarenko, Head of Analytical Center, Candidate of Engineering Sciences, A_Bondarenko@rivs.ru
N. I. Karamyshev, Head of Computer Programming Sector of Analytical Center
Ya. M. Katsman, Senior Researcher of Analytical Center, Candidate of Engineering Sciences

The article addresses the importance of credibility of mining and processing express analysis and control, which is the kernel of the integrated system of product quality control. The authors point at the main determinants of the analytical control reliability: the representativity of sampling of products under control, which is ensured by automatic pulp product assay system, and the value of error of X-ray fluorescence analyzer and further mathematical processing of the test data. Major attention is devoted to methods to minimize overall error of the analytical unit and to numerical and program implementation of these methods to support the proprietary automatic analytical control system ASAK-RIVS. Mathematical processing of X-ray phase test data includes the stages of: selection of maximum reliable spectrum line intensities, considering spectral background and deconvolution; correction of the intensities for instrumental eff ect; calculation of content of elements in the samples from pre-constructed calibration constraint equations based on linear and nonlinear regression models. The methods to enhance mathematical processing data reliability are discussed for each stage, and the applied mathematical models are described. The authors give brief presentation of capabilities of APM Analitika ASAK-RIVS program within ASAK-RIVS, with characterization of screen images of operation in the modes of calibration, control and correction of X-ray phase analysis results. It is concluded that the analytical control reliability is enhanced with the use of the developed technical, mathematical and program support ASAK-RIVS.

1. Karpov Yu. A., Savostin A. P. Methods of sample cutting and sample preparation. Third edition. Moscow : BINOM. Laboratoriya znaniy, 2015. 243 p. ISBN 978-5-9963-2584-9
2. Deconvolution With Applications in Spectroscopy. Ed. P. A. Jansson. Academic Press, Orlando, Florida, 1984. 342 p.
3. Monakhov V. V., Naumenko P. A., Chashinskaya O. A. An envelope method for removing background from X-ray fluorescence spectra. Instruments and Experimental Techniques. 2006. Vol. 49, Iss. 1. pp. 56–60.
4. Brunetti A. Removal of the continuum of X-ray spectra using morphological operators. IEEE Transactions on Nuclear Science. 1998. Vol. 45, Iss. 5. pp. 2281–2287.
5. Clayton E., Cohen D. D., Duerden P. Thick target PIXE analysis and yield curve calculations. Nuclear Instruments and Methods. 1981. Vol. 180, Iss. 2. pp. 541–548.
6. Mikhaylenko V. I., Mikhalchuk V. V. Methods of undefined spectra decomposition. Zhurnal prikladnoy spektroskopii. 1987. Vol. 46, No. 4. pp. 535–543.
7. Wazwaz A.-M. The regularization method for Fredholm integral equations of the first kind. Computers and Mathematics with Applications. 2011. Vol. 61. pp. 2981–2986.
8. Blokhin M. A. X-ray transmission physics. Moscow : Gostekhizdat, 1957. 518 p.
9. Kahoul A., Aylikci N., Aylikci V., Deghfel B., Kasri Y., Nekkab M. New procedure calculation of photoninduced Kβ/Kα intensity ratios for elements S to U. Journal of Radiation Research and Applied Sciences. 2014. Vol. 7, Iss. 3. pp. 346–362.
10. Liu H., Zhang Z., Sun J., Liu S. Blind spectral deconvolution algoritm for Raman spectrum with Poisson noise. Photonic Research. 2014. Vol. 2, No. 6. pp. 168-171.
11. Manoylov V. V., Novikov L. V. Assesment of parameters of mass-spectrometer peak in doublet. Nauchnoe priborostroenie. 2012. Vol. 22, No. 3. pp. 30–35
12. Tomilov A. V., Kalinin B. A., Aleksandrov O. E., Seleznev V. D. Mathematical processing of massspectrum that contains incomplete separated peaks. Analitika i kontrol. 2008. Vol. 12, No. 3-4. pp. 107–112.
13. Lawson C. L., Hanson R. J. Solving Least Squares Problems. Prentice-Hall, 1974. 337 p.
14. Fan Jinyan. The modified Levenberg-Marquardt method for nonlinear equations with cubic convergence. Mathematics of Сomputation. 2012. Vol. 81, No. 277. pp. 447–466.
15. Shelestov M. S., Insepov B. K., Bondarenko A. V. et al. Automated X-ray spectral analysis on Mongol-Soviet enterprise «Erdenet». Apparatura i metody rentgenovskogo analiza. 1987. Iss. 36. pp. 31–38.
16. Tyrsin A. N., Maksimov K. E. Assessment of linear regression equations by least absolute deviation method. Zavodskaya laboratoriya. Diagnostika materialov. 2012. Vol. 78, No. 7. pp. 65–71.
17. Radchenko S .G. Regression analysis methodology. Kiev : Korniychuk, 2011. 376 p.
18. Bondarenko A. V., Karamyshev N. I., Katsman Ya. M. Non-linear assessment of parameters in the engineering practice of the Scientific Production Association “RIVS”. Gornyi Zhurnal. 2011. No. 10. pp. 65–70.