Algorithmisation of design features selection of ordered construction families

Purpose: The main reason for starting research concerning new methods of creating ordered families of construction was a need to substitute traditional method used to realize those processes by methods oriented on computer aiding. Design/methodology/approach: Paper treats about development of algorithmic methods realizing gamma assignation (which determines relations between parameters of the future technical feature and its design features) in the process of ordered families creating. The main tool of algorithmization is developed theory of constructional similarity, supported by neural networks application method. Findings: The main achievements presented in this paper are developed theory of constructional similarity oriented on computer aiding and neural network application method, both used in designing process of new technical features. Moreover, both methods are highly susceptible to parametrisation. Research limitations/implications: Analyzed methods develop algorthmisation of designing environment and suport integration with the process of preparating the production. Intense research on the theory of technological similarity is being made nowadays. Practical implications: Described methods were being developed on practical examples of creating the module systems of hydraulic cylinders used in mining, slag cars used in metallurgy and gears series of types. Nowadays the gripping devices series of types is being made. Originality/value: Presented in the paper method of the constructional similarity, the algorithmic method and neural networks application method are basis of selection of design features in the process of ordered construction families (series of types and module systems of constructions) creating. All of these methods support intensive development of the types of technical features and affect on their competitive on the ready market.