Continuous-wave Ultrasonic Tomography for Oil/water Two-Phase Flow Imaging Using Regularized Weighted Least Square Framework.

Hao Liu,Chao Tan,Feng Dong
DOI: https://doi.org/10.1177/0142331219853073
IF: 2.146
2019-01-01
Transactions of the Institute of Measurement and Control
Abstract:Two-phase flow is a regime that commonly existed in many applications of process industry, on which the technique of ultrasonic tomography (UT) could provide reliable measurement with benefits of non-invasion, non-radiation and low-cost. In order to provide considerable visualization of low acoustic impedance contrast configurations in two-phase flow, a regular weighted simultaneous algebraic reconstruction technique (RW-SART) using continuous-wave excitation is proposed and presented in detail. With measurement strategy and formulized reconstruction model introduced, the measurement system is introduced with modification on continuous-wave excitation module and fan-beam angle transducers. A regularized weighted least square (RWLS) framework combining total variation (TV) regularization and SART reconstruction is derived and presented in detail to address the problem that under-determined and ill-conditioned situation couldn’t be resolved simultaneously. Both numerical simulation and experimental investigation are conducted to validate the feasibility of continuous-wave excitation and the improvement of proposed reconstruction strategy, from which the distribution of oil-water two-phase flow is effectively reconstructed. The implementation of measurement system provides an option of continuous high-voltage excitation and measurement in UT. By utilizing both the geometric information and penalty constraint, the proposed reconstruction approach provides a framework that introduces regularization information to the algebraic iteration reconstruction. Consequently, it has better reconstruction result than TV, Landweber, L1, filtered back projection (FBP) and SART method.
What problem does this paper attempt to address?