Function Analysis Model based on Granular Computing
In actual design process, design problems are generally ill-defined problem. The initial design
constraints information is often incomplete. These information needs to be subdivided. So in design
process, the granules of the problem domain and solution domain are changing continually. From the
view point of granules, the design process is a dynamic process. According to above analysis, a function analysis model based on granular computing is proposed as shown in Fig. 1. This model contains two
domains: problem domain and solution domain. The horizontal arrows represent the mapping between
problem domain and solution domain. The vertical arrows represent decomposition process of whole into
parts in problem domain and solution domain, which is the granular subdivision process from larger and
higher level granule to smaller and lower level sub-granule between the different granularity levels. The
gradient arrows represent the solution of higher level in solution domain influencing the granular
subdivision of next low level in problem domain. The arc arrow represents the construction process which
is a bottom-up process, which can validate the correctness of each decomposition result.
Figure