INCREASING THE PRECISION OF APPROXIMATION MODELS

Petrjanin Dmitrij L'vovich, postgraduate student, Penza State University (440026, 40 Krasnaya street, Penza, Russia), ra4cbh@mail.ru
Yurkov Nikolay Kondrat'evich, doctor of technical sciences, professor, head of sub-department of radio equipment design and production, Penza State University (440026, 40 Krasnaya street, Penza, Russia), yurkov_NK@mail.ru

004.021, 519.651, 519.654

Background. The challenge now is a hot topic of approximation of solutions to various technical studies. The aim is to find approximations of the models with the greatest possible precision and low error. The necessity of the development of effective methods for the analysis of experimental curves and constructing approximation models.
Materials and methods. Dana model increasing accuracy of approximation. A technique for reducing the error of approximation by multipass calculations and finding random approximation error, which allows you to assess the results of the study of technical systems and get the results more close to the original data. An algorithm for improving the accuracy of approximation of the model on which developed software that allows you to download experimental data and make calculations to find the approximation models. The technique, which allows to specify the approximation of the model, and to identify the best in terms of the approximation error.
Results. It is shown that the developed mathematical model improve accuracy and reduceerror of approximation technique improved the accuracy of the output model of the experimental data up to 10 %.
Conclusions. The use of the approximation model allows to evaluate the results of research and technical systems to obtain results more close to the original data.

approximation, the data, model, the calculation, method.