I.V. Denisov – Dr.Sc.(Eng.), Associate Professor

N.A. Sedova – Ph.D.(Eng.), Associate Professor

V.A. Sedov – Ph.D.(Phys.-Math.), Associate Professor

A.E. Sonin – Post-graduate Student

D.A. Grintsevich – Post-graduate Student

The method for identify of the parameters of fiber-optical measuring networks is proposed. This method performs line-by-line identification from fiber-optical measuring lines experimental data. The solutions of traditional reconstructive tomography methods are investigated in relation to the problem of reconstructing data on local effects on fiber-optical measuring networks sensitive areas using flexural modulation of lines. The solution of the fiber-optical tomographic problem for extended fields encounters to the problem of reverse reconstruction of existing physical fields. In direct formulation, such problem is incorrect and requires the use of special computational methods. However, for fiber-optical measuring networks of the amplitude detection principle, it is possible to introduce additional conditions. They allow to identify the parameters of the measuring networks and significantly simplify and improve the quality of reconstruction of external physical fields. The linearized expression is obtained for the power of optical radiation based on quasidistributed energy. When building of the mathematical model of fiber-optical measuring network from the fiber-optical measuring lines, mathematically allows to move from the nonlinear equations system to the linear algebraic equations system. The resulting linear algebraic equations system is linearized problem of reconstructive tomography in fiber-optical systems with amplitude modulation. The calculations show that such linearization is admissible under certain conditions. It is valid at the shallow modulation depth of the optical signal. Moreover, individual characteristics of the group of similar elements are taken into account. This is due to the design and installation features of the fiber-optical measuring lines. The used algorithmic implementation of the procedure for determining constant coefficients by the proposed method requires the formation of an experimental sample. It consists of external influences on the sensitive areas of the selected measuring line and the corresponding responses of this measuring line. The experiments performed on the developed of measuring systems layout showed that the range of variation of the coefficient’s values of projection matrix, which are assumed identical in the generalized approach in the proposed parametric identification scheme, can reach 18.2%. Analytical expressions and quantitative estimates are obtained for the parameters of measuring networks by measuring systems models. It is shown that there is clear dependence of the values of the projection matrix coefficients not only on the individual technical features of the measuring transducers, but also on the features of their integration into the measuring lines, which confirms the effectiveness of the proposed line-by-line identification method.

1. Kulchin Yu.N. Raspredelennye volokonno-opticheskie izmeritelnye sistemy. M.: Fizmatlit. 2001.
2. Khelgason S. Preobrazovanie Radona. M.: Mir. 1983.
3. Kulchin Yu.N., Denisov I.V. i dr. Optoelektronnaya raspredelennaya signalnaya sistema. Izmeritelnaya tekhnika. 2005. № 7. S. 28−32.
4. Kulchin Yu.N., Denisov I.V. i dr. Maket optoelektronnoi neiropodobnoi izmeritelnoi sistemy. Mikrosistemnaya tekhnika. 2003. № 10. S. 40−42.