PSI - Issue 20

Gusev E.L. et.al. / Structural Integrity Procedia 00 (2019) 000–000

5

Gusev E.L. et al. / Procedia Structural Integrity 20 (2019) 294–299

298

the flashbacks interval [0, Tmin], where Tmax=20. Forecast error the residual resource based on the constructed prediction model of optimal complexity with the optimal number of parameters on the prediction interval [Tmin, Tmax] is δ = 0,5% (curve 2 ). The prediction error based on the application of the prediction model corresponding to a fixed number of parameters n=4 is δ = 4.0% (curve 3). Thus, the prediction error when using forecasting models with a small number of parameters, such as the model of Bulmanis V.N. (the number of parameters does not exceed four) , can be 800% or more compared to the use of forecasting models of optimal complexity with the optimal number of parameters.

Fig. 1. Comparative analysis of the results of predicting the residual life on the basis of the optimal prediction model of optimal complexity corresponding to the optimal number of parameters n* =7, with a prediction error of 0,5% (curve 2 ), and on the basis of the prediction model corresponding to a fixed number of parameters n=4, with a prediction error of 4.0% (curve 3). Curve 1 – real time dependence of residual life. The vertical line separates the time interval on the left, which precedes the prediction, and the time interval on the right, on which the prediction is made. 6. Summary Within the framework of the refined variational formulations of inverse forecasting problems, a study of the problem of developing effective methods for predicting the residual resource of composite materials and structures under the influence of extreme environmental factors characteristic of the Arctic zone's sharp-continental climate is carried out. An integral part of this problem is the solution of the problem of effective construction of solutions that deliver a global minimum to multiparametric efficiency criteria that determine the variational formulation of the problem. For the first time, a new generalized model for predicting residual life, reliability, durability was developed, describing at the physical level the processes occurring in composite materials and structures under the simultaneous influence of several destabilizing physical factors. The principle of multiplicity of forecasting models, which was the basis of the developed approach, was formulated for the first time. Comparative qualitative studies have been carried out and methods have been