Visualized Measurement Of Multiphase Pipe Flow Using γ-CT/ECT

As two relatively mature process tomography (PT) techniques, computed tomography (CT) and electrical capacitance tomography (ECT) are endowed with broad prospects of industrial application. However, the dynamical complexity and random variability of multiphase flow have put forward a severe chanllenge to PT techniques, to improve the accuracy and real-time performance of visualized measurement of multiphase flow, the application of y-CT and ECT in multiphase flow measurement has been investigated systematically in this dissertation.Firstly, the design of y-CT and ECT systems is discussed, and imaging systems suitable for multiphase flow detection are bulit to provide quality hardware platforms for algorithm and application research.Secondly, image reconstruction algorithms based on convex optimization are studied, and both simulation and experimental data measured with our γ-CT and ECT systems are utilized to testify the validity of the algorithms,l1-norm regulization and nuclear norm regulization are adopted respectively to address the ill-posed inverse problem, and alternating direction method (ADM) and accelerated proximal gradient singular value thresholding (APG) algorithm are introduced into image reconstruction based on the dual problem and matrix reconstruction problem respectively, which obviously improved the imaging qulity in acceptable reconstruction time.Thirdly, to improve the imaging quality of the conventional iterative reconstruction algorthms for the low energy y-CT system, effective constraint and projection operators are constructed based on the fixed point model of the iterative algorithms; a local reconstruction technique is also proposed, which reconstructs image of the object locally and improves the space resolution without increasing algorithm complexity. For the dual-plane ECT system, commonly used time delay estimation algorithms are discussed and modern signal processing methods are introduce into the flow velocity measuring module of the ECT system. In order to improve the reliability of the conventional cross-correlation method, three new methods of estimating flow velocity are proposed:change point detection based method, cross-correlation method with adaptive time window and dynamical lag correlation exponent based method. Experimental results of coal ash measurement in a pneumatic pipeline demonstrate the effectiveness of the proposed methods.Eventually, dual-modality fusion is studied for combining the merits of y-CT and ECT. Results of both static and dynamic experiments demonstrate that the dual-modality system is capable of increasing the accuracy and scope of multiphase flow measurement.