EXPERT TREND IDENTIFICATION OF STRUCTURAL STABILITY

. Formulation of the problem. Luck of a unified concept, which identified integral criteria – structural stability of the static system, caused by the existing incompleteness of his formal representations, what justifies and initiate research of this system. Identification of the trend of structural stability. Quantification values of this function using two-digit logic, representing either the integrity of the structure of the object or its destruction is impossible, because it leads the task to conditionally correct. Relatively small changes, for example, of some technological parameter, changing structural stability of the identification object, what is not fixing by two-digit logic. In this connection, regularization of the named task is permissible through the use of an expert system that includes a specialized knowledge base. For the practical substantiation of the approach to determining structural stability, a metal was chosen (rolled from low carbon low alloy steel Ст3пс steel), whose reference points were assigned in the range of characteristics qualities limited by normative documents: ultimate strength   в = 370...490 MPa; yield strength   Т = 205...245 MPa; hardness  HRB = 62...70. Based on the analysis of the influence of synergistically interacting variables and the resulting equation, the trend of structural stability is determined. The significance of the work lies in establishing the trend of the structural stability of the object of identification, which allows predicting the values of the parameters that determine it. Conclusions and recommendations . An algorithm for determining the trend of parameters, according to which the structural stability of the object of identification is changed, is given: 1. Establishing his expert identification; 2. Determination of the working area of probabilistic assessments that establish the trend of structural stability and its quantification; 3. Establishing a trend of structural stability. отсутствие


Formulation of the problem
Saving the integrity of objects of different nature is associated with their structural stability.From positions of the dynamic system theory image f is C k -structural stable, if any C k close to it Image g is topologically conjugated to it by some homeomorphism h close to the identity [4]: where the dynamics of g differs from the dynamics of f only by a (continuous) change of coordinates.
Static systems remain unchanged composition, field properties and structure.It becomes obvious that the state of the structure of an object of identification of a static system cannot be described by a single criterion, and from these positions structural stability is an integral criterion, acting as a function of a number of its key characteristics (strength, ductility, etc.).For example, the quality of many materials is assessed by structurally sensitive characteristics [58], which initiates the appointment of structural stability as an integral criterion.
Traditionally, such a function should be determined using two-digit logic that reflects the ability of an identification object to maintain the integrity of the structure (the identification object is structurally stable  1; the identification object is structurally unstable  0).Such an approach makes the problem conditionally correct [9,10], because according to Adamar [11]:  it does not have a single solution (in the class of interest);  it can have many solutions (from two or more);  if the procedure for finding the solution is unstable (that is, with the slightest measurement error or small perturbations of the original data [12], the resulting solution may differ significantly from the exact one).
When determining the structural stability based on the analysis of theoretical regularities and statistically confirmed experiments, changes in parameters affecting it would be recorded.The current lack of formal representations of structural stability [13], initiates a search for those that include this task in the category of conditionally correct.
In this regard, we accept that regularization this conditionally correct task is possible by creating an expert system [1417], which should contain a specialized knowledge base, including: 1. Admissible formalization of structural stability.
2. The purpose of the working area of interrelated parameters, within which the structural stability is determined.
3. Quantification of structural stability.Determination of the trend of structural stability.

Identification of the trend of structural stability 2.1 Permissible formalization of structural stability
Accepted, the integral criterion structural stability can be perceived as a result of the synergistic interaction of its defining parameters.
The absence of the classical definition of structural stability can probably be explained by the difficulty of its representations as an n-dimensional vector, where n is the number of its determining particular indicators.Such an interpretation is mentally perceived as a smooth manifold homeomorphic to a sphere.
In this regard, the formalization of structural stability is difficult to deterministic description, which initiates an indirect, for example, expert interpretation of it.

Assignment of the working area of parameters defining structural stability and its quantification
Their synergistic effects on the object of identification.In this regard, it is necessary to establish reference points of the workspace of the identification object, in the range of which the degree of influence of parameters on its structural stability is determined.
For the practical substantiation of the approach to determining structural stability, a metal was chosen (rolled from Ст3пс steel), whose reference points were assigned in the range of characteristics qualities limited by normative documents: ultimate strength   в = 370...490 MPa; yield strength   Т = 205...245 MPa; hardness  HRB = 62...70 Ст3пс low carbon low alloy steel had a ferritic-pearlite structure and did not respond to heat treatment (Fig. 1.).The degree of influence of synergistically interacting variables predicted by experts on the structural stability of the metal Y exp is given in the planning matrix.
In this matrix (see table), the UL and LL are the upper and lower levels of the variables X 1  X 6 (the percentage of elements of the chemical composition of the object of identification); MY  medium level; VV  is the variable variation interval.
Matrix rows  situations in which the expert evaluates the numerical values of structural stability within fixed points (from 0 to 1).
Based on the analysis of the influence of synergistically interacting variables and the resulting equation, the trend of structural stability is determined (see Fig. 2).

Fig. 2. Structural stability trend
The significance of the work lies in establishing the trend of the structural stability of the object of identification, which allows predicting the values of the parameters that determine it.

Conclusions and recommendations
A regularization of the conditionally correct problem of establishing the trend of structural stability is proposed, including: 1. Establishing his expert identification.
2. Determination of the working area of probabilistic assessments that establish the trend of structural stability and its quantification.
3. Establishing a trend of structural stability.The use of a synergistic approach in determining the influence of interacting parameters on structural stability is due to the self-organization of the expert system, as an open system, exchanging information with the external environment.

Fig. 1 .
Fig. 1.Steel structure St3ps, × 500Predicting the degree of influence of interrelated parameters on structural stability in the range of values from 0 to 1, its trend was expertly established.To facilitate the work of experts, specialists in materials science, who determine the probabilities of the influence of interacting parameters on the structural stability of the object of identification, the values of probabilities were established taking into account the data [1825, etc.].