Abstract:
The aim of this thesis is to enhance process flow tomography image reconstruction,performance, speed and robustness while preserving the safety of the tomography system. This aim is realized through: (1) Fusion of some complementary tomography systems and development of appropriate image reconstruction algorithms based on the different boundary data measurements. This is to enhance the quality and the resolution of the reconstructed images and thus an improvement in real-time flow measurement and estimation. (2) Exploiting the temporal correlation, that usually exists between consecutive reconstructed multiphase flow
images, to improve reconstruction quality. (3) Developing a hierarchical reconstruction method for process flow tomography to speed-up the reconstruction process and making it more stable by reducing the number of reconstruction parameters. This thesis focuses on addressing the three aforementioned objectives with results that illustrate enhanced imaging qualities speed and error/disturbance robustness and thus good estimation of the different components of a multiphase flow in a given cross section of a pipeline.
