Optimization Design And Algorithm Investigation For Multiphase Flow CT

Multiphase flow measurement widely exists in industrial process, such as oil, chemical engineering. Multiphase flow measurement with high precision, real-time and on-line performance is of significant role for safe and efficient operation, control and management in these industrial processes. After introducing the basic theory of multiphase flow measurement and radiation CT, system optimizing and image reconstruction are investigated in following aspects for radiation CT system of multiphase flow measurement in this paper.1) Base on Monte Carlo method, a simulation software named Gris is developed for multiphase flow radiation CT (MFR CT) using the high energy particle transport simulation toolkit Geant4. Accurate geometric & physical model are constructed with the physical processes existing in MFR CT, such as Compton Effect, Photoelectric Effect and Pair Effect. The process of radiation particle transport in MFR CT is simulated to get the trace of particle and the sensor response of radiation particle and its secondary particles.2) The different architectures and parameters of the radiation particle transport model are simulated using Gris. Through simulations the information of particle distribution and sensor response can be obtained. Architectures and parameters of the system are optimized, and the system noise and measurement precision can be improved.3) Several image reconstruction algorithms based on the theory of backprojection and iterative algorithm have been investigated, such as linear backprojection, R-L filtered backprojection, S-L filtered backprojection, ART iterative etc. Using these algorithms, data obtained through Gris simulation are employed to reconstruct the cross section image in pipe. Comparing these reconstructed images the most suitable algorithm for MFR CT can be obtained.Finally, by means of the optimized system modle, several multiphase flow patterns have been simulated, and the reconstruction images have been obtained. To counter the problems existed in MFR CT, some improvement suggestions are presented and discussed.