A Sequential Ultrasound and Electrical Capacitance Process Tomography Sys- Tem

Abstract A novel concept for the fusion of Ultrasound Reflection Tomography (URT) and Electrical Capacitance Tomography (ECT) for industrial process applications is presented. The method is intended to combine the strengths of both principles while reducing their respective disadvantages. ECT is sensitive to bulk rather than to phase boundaries and suffers from blurred images. URT is able to detect phase boundaries with good resolution but is not able to reconstruct absolute material parameters. In our approach URT and nonlinear Gauss-Newton based ECT are sequentially coupled by using the URT results as a natural regularization for the ECT reconstruction algorithm. Two different methods are investigated that can also be simultaneously used. Adaptive mesh grouping uses data from the ultrasonic measurements to merge regions without phase boundaries before performing the ECT inversion, thus reducing the number of unknowns. For the second method a weighted regularization matrix is composed based on the URT data, leading to a physically meaningful regularization. The methods are tested with simulated gas-liquid two-phase flow profiles. The achieved results show a significant improvement compared to stand-alone ECT.