Adaptive Mesh Grouping in Electrical Impedance Tomography for Bubble Visualization

The bubble visualization in two-phase flow using the E1T (Electrical Impedance Tomography) technique requires an image reconstruction process. When the conventional iterative image reconstruction algorithms are used, the processing time increases rapidly and the convergence characteristics become very poor as the spatial resolution increases. In order to overcome this problem, this study proposes an adaptive mesh grouping method utilizing the genetic algorithm and the fuzzy set theory. Computer simulations using the improved Newton-Raphson method combined with the proposed method show promising results that mesh grouping may become a useful way to mitigate the ill-conditioning phenomenon which makes theEIT inverse problem difficult.