Robust Automated Stopping Criterion for Semi-Convergent Image and Velocity Reconstruction in Electrical Capacitance Volume Tomography

Electrical Capacitance Volume Tomography (ECVT) is a low-cost and high-speed sensing modality with great potential for industrial multiphase flow monitoring. In this work, we examine the use of the slope of the capacitance vector residual curve, as directly provided by the measurement data, to formulate a robust stopping criterion for the iterative image reconstruction in ECVT. The methodology is illustrated based on experimental data from a gas-solid fluidized bed. We show that the proposed stopping criterion can improve the performance of both the image reconstruction and the flow velocity profiling in ECVT applications. For concreteness, we focus on the popular Landweber iterative reconstruction algorithm although other reconstruction algorithms exhibiting semi-convergent behavior might benefit from the present analysis as well.