A HYPERGRAPH FRAMEWORK FOR OPTIMAL MODEL-BASED DECOMPOSITION OF DESIGN-PROBLEMS

Decomposition of large engineering system models is desirable since in creased model size reduces reliability and speed of numerical solution algorithms. The article presents a methodology for optimal model-base d decomposition (OMBD) of design problems, whether or not initially ca st as optimization problems. The overall model is represented by a hyp ergraph and is optimally partitioned into weakly connected subgraphs t hat satisfy decomposition constraints. Spectral graph-partitioning met hods together with iterative improvement techniques are proposed for h ypergraph partitioning. A known spectral K-partitioning formulation, w hich accounts for partition sizes and edge weights, is extended to gra phs with also vertex weights. The OMBD formulation is robust enough to account for computational demands and resources and strength of inter dependencies between the computational modules contained in the model.