Graph theory applied to noise and vibration control instatistical energy analysis models

摘要

A fundamental aspect of noise and vibration control in statistical energy analysis (SEA) modelsconsists in first identifying and then reducing the energy flow paths between subsystems. In thiswork, it is proposed to make use of some results from graph theory to address both issues. On theone hand, linear and path algebras applied to adjacency matrices of SEA graphs are used todetermine the existence of any order paths between subsystems, counting and labeling them, findingextremal paths, or determining the power flow contributions from groups of paths. On the otherhand, a strategy is presented that makes use of graph cut algorithms to reduce the energy flow froma source subsystem to a receiver one, modifying as few internal and coupling loss factors aspossible.