Modeling of diagnostics of engineering objects to optimize their life cycle

Abstract

The aim of the study is to create a methodology for diagnosing the state of objects on the basis of improving the transformation of state coordinates and their classification. The vector space transformation allows to perform the recognition of images of multidimensional objects represented by the vector space by the decision function based on the rotation of coordinates with the subsequent reflection of one of the coordinate axes. In real production, this approach allows on simplified databases of precedents of small volume with the choice of the presence of the parameter \"yes/no\" on the basis of the same mathematical model and the formulation of optimization problems to carry out not only the study of the influence of individual production factors on the quality and durability of the product within its overall life cycle, but also to obtain data on the basis of which general recommendations can be determined by changing the manifestation of features relative to the target function. This will allow, in addition to the actual diagnosis of the existing production, to conduct research on future production, as well as to analyze the development strategies of both individual stages and their combination.