CIESC Journal

• 多相流 • 上一篇    下一篇

基于遗传算法的快速X射线计算机层析成像多相流测试技术(Ⅱ)静态实验验证

丁宇龙 吴昌宁 程易 金涌   

  1. 清华大学化工系
  • 出版日期:2007-02-05 发布日期:2007-02-05

Genetic algorithm based fast X-ray computed tomographic method for multiphase flow measurement (Ⅱ)Validation by real experiments

  

  • Online:2007-02-05 Published:2007-02-05

摘要: 采用静态实验模型和一套X射线测试系统模拟了二值多相体系多角度同时投影的X射线CT测量过程。静态实验模型模拟空气-水两相体系,截面相浓度分布满足0-1特征。X射线测试系统由一台工作电压为150 kV的闪光X射线机、一台X射线平板检测器和一套数据采集单元组成。通过旋转静态实验模型,分步实现不同角度的投影数据采集。基于模拟的多角度“同时”采集的投影数据,使用本系列研究所构建的基于遗传算法的快速X射线CT多相流测试技术(GA-XCT)对截面图像进行重建,测试结果表明:在3~24个有限角度实测投影数据的情况下,GA-XCT表现出了明显优于传统CT图像重构算法(滤波反投影算法)的图像重构能力,且具有良好的抗噪声能力。

Abstract: Simultaneous measurements of X-ray projection data from several angles around a static experimental model were implemented by rotating the model between an X-ray tube and a flat plate detectorThe static model simulated a two-phase system with water and air inside, forming four bubbles at the cross-sectionThe measurement system consisted of a Model 150-I Mobile X-ray Unit operated at the potential of 150 kV, a MiniX-1 X-ray flat plate detector, and a data acquisition systemBy using the measured instantaneous projection data from different angles, the cross-sectional image was reconstructed with the established genetic algorithm based fast X-ray computed tomographic method(GA-XCT)A series of real experiments clearly demonstrated the feasibility and capability of the GA-XCT in the application of “bubbles” reconstruction compared with the conventional CT reconstruction algorithm using limited projection dataMeanwhile, the GA based image reconstruction is robust to the noise.