The paper presents the first attempt to combine methods of stochastics (mathematical statistics) and methods of theory of chaos and self-organization for studying such complex (chaotic) processes as postural tremor. It was established that when re-registering tremor in each subject by n=15 or n=30 obtained tremorograms do not exhibit normal distribution, and non-parametric distributions show distinctions at pairwise comparison on Wilcoxon test (only 2 or 3 pairs from 210 may belong to the same general population). Static physical load sharply changes this picture and the number of such ("similar") pairs increases. The estimation method for effect of a load on tremor is proposed. Simultaneously, within calculating quasi-attractors there is a clear picture of division of chaotic dynamics of tremor parameters with load and without load. Prospects of a new method application in physiological measurements are discussed. Limited method of stochas- tics in description of complexity is underlined, and necessity of calculation quasi-attractor´s parameters in phase space of state is proved.

In the framework of the compartmentae-cluster approach there is possibility of constructing adequate mathematical models that may be of several types supposedly stationary modes of biomechanical systems: in the traditional deterministic approach, when the state vector оf the biomechanical system have equal value and in the framework of the new theory of chaos and self-organization, when system state vector x=x(t)=const. The vector can occur within a bounded volume of the phase space of states. The message signals presented arbitrary human motion under the influence of an alcoholic beverage and the simulated signals for a given external exposure control (Ud=60 у.е.) was compared. Different values ​​of the damping coefficient (b=1,1; b ´= 3,4) present the normal and unnormal state of human body. A comparison was made, the resulting figures and draw conclusions about the impact of damping coefficient on the size of the area of ​​quasi-attractor. Present the state of the biological dynamical system (the human body) under alcohol effect and in normal state.

Process modeling is one of the main directions in science and engineering. When there is a model of simple system it is possible to obtain the parameters of its state vector before a process starts. However, the modeling of complex processes (systems) applying standard methods of computing and simulation software is confronted with difficulties of chaotic dynamics of such systems. A number of models that allow to describing the dynamics of complex biological dynamic systems is extremely small, as to repeat the same results of experiments it is impossible on the basis of deterministic or stochastic models. In additional, the authors propose a model for description of human state vector through three-compartment two-cluster control system. This model may be implemented using an application package exhibiting an operation of each cluster separately. The output signals of the model seen can be compared with the actual experimental data observed as output signals. The signals obtained at the output of the simulation model show different states of the system under external control actions which cause the change in the internal properties and states due to external (disturbance) signal. Control external signal from the first cluster and accordingly the output signal from the second cluster were divided into four main components that had the same analogues in the dynamics of complex biological behavior of dynamical systems [2,4].

So far, the fact that statistically significance differences between samples (sets of parameters of the body xi) before the treatment and after the treatment led to the conclusion of non effective treatment. However, in the framework of the theory of chaos and self-organization the assessment of the significant differences can be carried out without going through statistical methods, based on the analysis parameters of quasi-attractors or using neural emulators. In this report the authors present examples of uncertainties of the 1st kind in regenerative medicine and introduce new technologies to resolve these uncertainties. A procedure for finding differences between samples and determining the parameters of the order (the most important diagnostic features) based methods neurocomputing. The effectiveness of this approach in the evaluation of the effectiveness treatment effects of abnormalities in the body during acute stroke in a kinesotherapy is presented. The authors proved ineffective stochastics and opportunities neurocomputing in the task of system synthesis

Processes simulation is one of the main directions in science and technology. In the presence of a simple model of the process, the results of the process before it starts can be obtained. The simulation of complex processes (systems) within the standard methods of computing and simulation software runs into difficulties because of the chaotic dynamics of such systems. The number of models that allow to describe the complex biological processes of dynamic systems is extremely small, because it is impossible to repeat the same results of experiments based on deterministic or stochastic models. The authors propose a model that allows the description of the state vector of a person within tree compartmental two-cluster systems management. The model is implemented using the package of applied programs that demonstrate the performance of each cluster separately. On the model output signals are obtained, they are compared with real experiments and the resulting signals. Signals obtained at the output of the simulation model signals show different values of external control actions in which the change in the properties of the output signal. Control signal and the output signals respectively were divided into four main components that have the same name counterparts with complex biological dynamical systems.