SYNTHESIS METHOD FOR THE FUNCTIONAL STRUCTURE OF NEW TECHNICAL DECISIONS BASED ON PATENT MASSIFT DATA
D.M. Korobkin, S.A. Fomenkov, S.G. Kolesnikov
The authors of the paper present a developed method for synthesizing the functional structure of new technical systems based on the use of a morphological matrix formed as a result of semantic processing of patent corpus. One of the goals of the synthesis of new technical solutions is to improve the characteristics of individual components of a technical object (TO). For this, it is necessary to detail the TO constructive functional structure, i.e. define the TO components and the functions of the TO components. On the basis of the extracted SAO from the claims, the relationships between the technical object and its components are determined. The extracted attribute SAOs are used to define the functions of a technical object. Based on the morphological matrix, we configure new technical solutions and rank them using the method of conjoint analysis (CA). The CA’s criterion is calculated based on the number of patents from which the selected technical function was extracted. Another task of the synthesis of new technical solutions is the realization of technical function. In this method, we use the information from the developed matrix linking physical effects (PE) and the technical functions (TF) performed by them. The required TF can be implemented directly by the PE, otherwise the task for the synthesis of the physical operating principle (POP) is formed. Among the POP structures constructed, an optimal one is selected based on the criterion estimates of all the PE included in it. Based on the optimal POP, the flow functional structure of the synthesized technical system is formed.
Keywords: :morphological analysis, technical functions, physical effects, patents