主催: The Japan Society of Mechanical Engineers
会議名: 第30回 原子力工学国際会議（ICONE30）
開催日: 2023/05/21 - 2023/05/26
Industrial processes involving gas-liquid two-phase flow are one of the most common phenomena in the field of nuclear energy. How to determine the flow pattern is a basic problem in two-phase flow analysis, and an effective two-phase flow pattern prediction model is very important for the research of multiphase flow. In many traditional studies, people still cannot get accurate enough prediction results. In this paper, machine learning tools and data-based algorithms will be used to determine the flow pattern, and the algorithm with the highest accuracy will be selected to establish a flow pattern prediction model that meets the requirements. By using the flow pattern prediction model in this paper, the accuracy of flow pattern prediction can reach more than 99%, which is higher than the traditional methods.
