主催: 一般社団法人 日本機械学会
会議名: 第24回 動力・エネルギー技術シンポジウム
開催日: 2019/06/20 - 2019/06/21
Two-phase flow regime identification in internal flow system is crucial for various energy systems involving phase-change. Closure of several conservation equations often times require proper selection of interfacial transfer models, which are highly dependent on flow regime. In the current study, non-intrusive methodology to identify two-phase flow regime is proposed using vibration signals acquired from the force transducers attached on the external piping structure. For the objective flow regime identification, machine learning techniques are adopted for the flow regime clustering.
