the radioelectronic module
hybrid intellectual system
Now processes of designing and preparation of manufacture of modern radioelectronic modules (REM), being constructive base practically all types of radio engineering systems and devices (RESD), represent a complex of versatile problems of high level of the complexity, demanding considerable time and material inputs. The rigid market competition, high rates of development of element base and materials, modernisation of manufacturing techniques, increase of functionality and reliability RESD in a combination to constant improvement of their designs and ergonomics lead to reduction of the period of a demand of any electronic product in the market of radioelectronic equipment.
The structure of a route of optimum designing has been developed for the decision put above problems competitive REM which includes hybrid intellectual system of optimisation acceptance of predesign technical decisions (HISOAPTD) and a programmno-methodical complex (PMC) konstruktorsko-technological optimisation (KTO) REM. Methods and algorithms of optimum designing competitive REM which raise efficiency of application of modules in various RESD are developed. Difference of the offered methods and algorithms is use of a statistical method in a combination with gradiant methods and use of a method of research of space of parametres at a stage of optimisation of element base, and also use of genetic algorithms at the decision of optimising problems at stages of konstruktorsko-technological optimisation of manufacturing of printed-circuit boards (PCB) and REM. The structure of corresponding design procedures and the software, differing by the complex approach to the decision of problems of konstruktorsko-technological optimisation at various stages of creation REM are considered structure of a route of optimum designing competitive REM. Also the structure and principles of work HISOAPTD based on a combination of an expert subsystem, using indistinct logic models of knowledge, an expert subsystem on the basis of artificial neural networks and an expert subsystem in which are put in pawn productional models of knowledge are considered. On the basis of considered above mathematical models and algorithms KTO is developed PMC REM with which help optimisation of designs of modules is carried out multicriteria, efficiency of processes of manufacturing of software and REM, and also process of a covering by a varnish modules and control operations raises. Therefore competitiveness of modules raises, at decrease in their cost price and reduction of terms of an exit of products on the market.