V.B. Kulikov - Post-graduate Student, Nizhniy Novgorod State Technical University n.a. R.E. Alekseev. E-mail: vb.kulikov@yandex.ru V.P. Khranilov - Dr.Sc. (Eng.), Professor, Nizhniy Novgorod State Technical University n.a. R.E. Alekseev. E-mail: hranilov@nntu.nnov.ru

In the simulation process monitoring and diagnosis of complex systems of different origin is necessary to correct the extraction, identification and interpretation of measuring data and experimental data stochastic processes (SP). A decisive role in the choice of stochastic processes isledovanie prediction and filtering algorithms plays an analysis of the statistical characteristics of random variables (SV). At the same time, in order to avoid methodological errors need to take into account the effects of smaller, compared with the main order: local fashion, small extremes, or Cauchy Pareto distribution with a «heavy-tailed». For correct resolution of these problems is proposed as the main approach to probabilistic analysis with CB said specifics use methods correctly identify the distribution laws ST. They, in particular, based on the methods of solution of inverse problems of mathematical physics, proposed and developed in the scientific school of Academician A.N. Tikhonov. Basis of presentation The approach is based on correct methods of reconstruction of the density distribution of SW and implementation methods of the joint venture with approbation on extensive factual material in the areas of natural proskhozhdeniya prikladnh such as study of the processes and haemo- lymph formation, analysis of indicators of the level of white blood cells, B-lymphocytes, immunoglobulins, phagocytic numbers and other indicators blood and lymph elektrogastroenterograficheskie (EGEG) studies in the diagnosis of gastrointestinal tract. To describe such a model is proposed to use the joint venture Fredholm integral equation of one kind. The integrand probability density function of the magnitude of the problem is unknown. The right side of keeping up the empirical distribution functions for each study ST. Thus, the identification of joint finite-dimensional distributions taking into account the factor of possible multimodal and other specific features of this joint venture, the restoration of the density distribution of the immune parameters was performed on the set of trigonometric functions with a limited number of members of decomposition of N depending on the size L of the observed data. The proposed approach does not require histograms, select the number and width of the intervals of grouping data, requires approximation of the histogram and the application of the criteria for the approval testing of hypotheses concerning the distribution law. This paper presents the results of the stochastic analysis of concrete realization of the random process of clinical medicine - EGEG signal. As a result, studies have found, and in fact proved multimodal nature kvazifraktalnogo stochastic process on all temporary site. Applied value of the proposed stochastic approach is based on the correlation and comparative analysis EGEG-, or other biosignals natural origin for groups of patients in different degrees of norms and pathology. This polymodal stochastic characteristics carry information about the nonequilibrium dynamics of multi-state organs and body systems. In this case, the density of local areas calculated moment functions, entropy characteristics are monitored and investigated the process of therapy.

 

1. Kulikov V.B. Identifikacija odnomernykh mnogomodalnykh plotnostejj raspredelenija verojatnosti pri ogranichennom obeme dannykh metodom reguljarizacii // Vestnik Saratovskogo gosudarstvennogo tekhnicheskogo universiteta. 2013.  № 3(72). S. 7-11.
2. Kulikov V.B. Vosstanovlenie polimodalnykh plotnostejj verojatnosti po ehksperimentalnym dannym v strukturakh so stokhasticheskimi svojjstvami // Vestnik Nizhegorodskogo universiteta im. N.I. Lobachevskogo. 2014. № 1(1). S. 248-256.
3. Tikhonov A.N., Arsenin V.JA. Metody reshenija nekorrektnykh zadach. M.: Nauka. 1986. 288 s.
4. Alekseev V.G. Neparametricheskoe ocenivanie plotnosti verojatnosti i ee proizvodnykh.L2-podkhod // Avtometrija. 2007. № 6. S. 39-47.
5. Vapnik V.N., Stefanjuk A.R. Neparametricheskie metody vosstanovlenija plotnosti verojatnosti // Avtomatika i telemekhanika. 1978. № 8. S. 38-52.
6. Antipov O.I., Nagornaja M.JU. Fraktalnyjj analiz ehlektrogastroehnterograficheskogo signala // Biomedicinskaja radioehlektronika. 2010. № 10. S. 40-44.
7. Kuljas M.O., Nagornaja M.JU. Apparatno-programmnyjj kompleks dlja zapisi i obrabotki ehlektrogastroehnterografiches-kikh signalov // Fizika volnovykh processov i radiotekhnicheskie sistemy. 2011. № 1. S.93-99.
8. Aratskijj D.B., Leontev E.A., Morozov O.A.i dr. Informacionno-optimalnye metody v fizike i obrabotke ehksperimentalnykh dannykh. Monografija. N. Novgorod: Nizhegorodskijj un-t. 1992. 146 s.