Bouldakova T.I., Gridnev V.I., Kirillov K.I., Lantcberg A.V., Suyatinov S.I.

The paper presented a method for model processing biosignals and especially its implementation in program-analytical complex. At present, the mathematical models in the form of differential equations, obtained by the algorithms of design (reconstruction) of complex systems on based of Taken-s method, are used for research the internal dynamics of biosystem. In methods of reconstruction (inverse problem of dynamics) the output signals generated by biosystem are the background information. If a few signals are available for the registration, the accuracy of reconstruction increases. Moreover, the reconstruction of biosystem models using several realizations of biosignals of different functional origin provides an effective method of coprocessing biosignals, fundamentally different from statistical methods. The basis of the model processing is structural and parametric identification of the researched subsystems on the basis of registered biosignals of system «heart-lung-vessels». Evaluation of intercouplings metrics of the researched subsystem models allow to determine the characteristics or signs of a functional state in the early stages. The practical implementation of the proposed algorithm of the model processing biosignals, its approbation and perfecting implement with the help of the complex that allows to record of biosignals synchronous, to execute of coprocessing of biosignals based on the model analysis, to store input data and processing results for future usage. In the work the features of schematics for implementation of model processing of biosignals are reviewed. They are the provision of synchronous registration of biosignals with the necessary accuracy and the dropping of the noise signal level. Therefore the modular-hierarchical structure of complex consisting of mobile device of registration and workstation is offered. The mobile device is intended for registration of biosignals and primary processing of information, it also realized the self-testing and analysis of electromagnetic noise levels of different origins. It consists of two functional blocks: measuring and computing. Workstation based on the personal computer, which implemented such procedures as selfdiagnostics, interactive user interaction, management of computing process, coprocessing of biosignals etc. The example of implementation of coprocessing of biosignals in the «heart - vessel» system is adduced. For an estimation of model parameters, the method of invariant imbedding was used. The proposed approach allows framing the structural - parametric model, objectively reflecting the state of the body at operation of the different factors, to store this information in a convenient format and easily to handle its various mathematical methods.