Edinburgh virtual multiphase flow imaging dataset

This dataset is for publication "Digital twin enables quantitative multiphase flow imaging with low-cost tomography". ABSTRACT: Multiphase flow is ubiquitous in nature, industry, and research. Quantitative multiphase flow imaging is the key to understanding this complex phenomenon. However, there are as yet no low-cost, scalable imaging techniques to achieve quantitative multiphase flow imaging for a wide range of scenarios. Here we report a digital twin (DT) framework that unlocks the real-time quantitative multiphase flow imaging capability of electrical tomography (ET) that is non-invasive, non-radioactive, scalable, and low-cost. The DT framework, building upon a synergistic integration of the 3D field coupling model, deep learning, and edge computing, allows ET to dynamically learn the flow features in the virtual space and implement the learned model in the physical device, thus providing unprecedented resolution and accuracy in the real world. The DT framework is demonstrated on dynamic gas-liquid two-phase flows, showing step-change performance compared to the state of the art. It can be readily extended to various tomography modalities and multiphase flows at different scales for precise flow visualization and characterization.本数据集对应发表论文《Digital twin enables quantitative multiphase flow imaging with low-cost tomography》。 摘要：多相流在自然界、工业生产与科研场景中无处不在。定量多相流成像是解析这一复杂物理现象的核心手段。然而，当前仍缺乏可适配多类应用场景、低成本且可规模化推广的定量多相流成像技术。本文提出一种数字孪生（Digital Twin, DT）框架，可使非侵入式、无放射性、可规模化且低成本的电学层析成像（Electrical Tomography, ET）具备实时定量多相流成像能力。该DT框架依托三维场耦合模型、深度学习与边缘计算的协同集成，可使ET在虚拟空间中动态学习流场特征，并将习得模型部署于物理设备中，从而在现实场景中实现前所未有的成像分辨率与精度。本框架已在动态气液两相流场景中得到验证，相较于当前最优技术，其性能实现了跨越式提升。该框架可便捷拓展至多种层析成像模态与不同尺度的多相流场景，以实现精准的流场可视化与特性表征。

搜集汇总

背景与挑战

背景概述

该数据集是用于支持'Digital twin enables quantitative multiphase flow imaging with low-cost tomography'研究的虚拟多相流成像数据，发布于2022年，由爱丁堡大学工程学院数字通信研究所提供。它基于数字孪生框架，整合3D场耦合模型和深度学习技术，旨在实现低成本、高精度的多相流（特别是气液两相流）定量成像，适用于工业和科研场景中的流动可视化和表征。