Layout Knowledge Multi Granularity Reconstruction Method for Complex Product

摘要

To analyze the complex product layout design knowledge incompleteness and attributes' associated correlation ship at the product scheme design stage, a novel multi-granularity reconstruction methodology for complex product layout knowledge was proposed. First of all, the complex product layout knowledge incomplete information system was established, on the basis of the hierarchy multi granularity decomposition of the obtained layout knowledge. Then, the layout knowledge reconstruction process was divided into two stages: layout knowledge attributes reduction and trading off the attributes' degree. At the layout knowledge attributes reduction stage, on the layout knowledge single granularity analysis of viewpoints, the layout knowledge attributes reduction was carried out through the rough set theory for the incomplete information system. Via taking the layout types as decision attribute, the attribute reduction algorithm, which was based on the relative knowledge granularity and attributes' significance, was provided. Subsequently, on the multiple granularity knowledge viewpoints, the attributes grey relational degree was given on the grey theory. The conditional attributes were regarded as parameter value and the decision attribute as basic value sequence. Afterwards, at the trading off the attributes' degree stage, the layout knowledge multiple source information system was modeled upon the attributes' significance and grey relational degree. The layout knowledge attribute degree was verified by Demspter-Shafer theory. Finally, machining center layout knowledge modeling, reconstruction, and attributes identifying were taken as an example to verify the effectiveness and validity of this method.