System of standardless diagnostic of cell panels based on Fuzzy-ART neural network

Abstract

Distinctive line of composite materials is indivisibility of their creation process and formation of properties from reception of ready details and knots. According to this, achievement of high reliability of products from composites it is necessary to improve, first of all, manufacturing techniques and to provide the control of observance of all parameters of technological process. Not less important is increasing reliability of an estimation of a technical condition of a product at the moment of manufacturing and while process of exploitation. Composite materials are characterized by considerable disorder of physical parameters, materials which enter into composite, low density, considerable disorder of properties both from a product to a product, and in the product itself, their anisotropy, unavailability of the sufficient information of the basic types of defects. Therefore to this, it is impossible to apply traditional methods of nondestructing control, which based on knowledge of physical parameters of materials, such as elastic modulus, density, propagation speed of ultrasonic waves, dielectric permittivity.