两相流参数测量中,电阻层析成像技术的图像重建算法可归为两大类,即直接法和迭代法。对常用的几种图像重建算法进行了比较性研究,结果表明:直接法速度快,但重建图像质量不理想,仅可作定性分析;迭代法总体上说重建图像质量较高,但速度较慢。当增加像素数以提高重建图像分辨率时,直接法重建图像耗时增量较小,迭代型算法中改进的Newton-Raphson算法、BFGS算法由于目标函数的Hessian矩阵维数增加,计算时间大大增加,而改进的BFGS算法因简化了Hessian矩阵及其逆矩阵的计算,时间增量则较小。
Abstract
The electrical resistance tomography reconstruction algorithms of two-phase flow can be categorized into two kinds,direct algorithms and iterative algorithms.These reconstruction algorithms are compared and studied,the results show that:the direct algorithms are fast,but the reconstruction images are blurry and can only be used as qualitative analysis.The iterative algorithms are slower,but the reconstruction images quality is better. As the pixels rise to enhance resolution of reconstruction images,the increment time of the direct algorithms is very small.The reconstruction time of the Newton-Raphson algorithm and BFGS algorithm is largely increased because of the increased dimension of the Hessian matrix.The Hessian matrix and its inverse matrix are simplified in the modified BFGS algorithm,therefore the reconstruction time is increased a little.
关键词
计量学 /
电阻层析成像 /
图像重建算法 /
两相流
Key words
Metrology /
Electrical resisitance tomography /
Image reconstruction algorithm /
Two-phase flow
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
［1］邵嘉庆,徐苓安.电阻层析成象系统的研究［J］.自动化与仪表,1999,14（2）:1-3.
［2］Barber D C,Brown B H. Applied potential tomography［J］. J Phys E: Sci Instrum,1984,17（8）: 723.
［3］Xie C G,Huang S M, Hoyle B S,et al. Transputer-based electrical capacitance tomography for real-time imaging of oilfield flow pipeline［M］//Beck M S,Campogrande E,Morris M,et al.Tomography Techniques for Process Design and Operation.Southamphton:Computational Mechanics Publications,1993,333-346.
［4］Yorkey T J,Webster J G,Tompkins W J.Comparing Reconstruction Algorithms for Electrical Impedance Tomography［J］.IEEE Transaction on Biomedical Engineering,1987,34（11）: 843-852.
［5］Yu J H,Huang Z Y,Ji H F,et al.Image Reconstruction Algorithm of Electrical Resistance Tomography for the Measurement of Two-phase Flow［C］//Proceedings of IEEE Sensors.Toronto,Canada,2003.
［6］Kin M C,Kim S,Kim K Y,et al.Regularization methods in electrical impedance tomography technique for the two-phase flow visualization［J］.International Communications in Heat and Mass Transfer,2001,28（6）:773-782.
［7］Klose A D,Hielscher A H.Quasi-Newton methods in optical tomographic image reconstruction［J］.Inverse Problems,2003,19（2）: 387-409.
PDF(441 KB)
