Designing Of Electrical Capacitance Tomography Oil-water Two-phase Flow Measurement System Development Platform

Electrical Capacitance Tomography (ECT) is a new technology for monitoring two-phase flows. By using an array of capacitance measurement electrodes mounted on a circular pipe wall, it is possible to measure and study the two-phase flow by means of reconstructed images. Comparing to other flow imaging technology, it's low-cost, fast, non-invasive, applicable to a wide range, operation safe, and so on. In order to obtain a thorough understanding of ECT measurement sensitivity distribution within a flowing pipe and ultimately a quality image of the flow, this paper presents a research work, based on the ECT measurement principle, on the following issues:1. Using MATLAB to establish a two-dimensional finite element method (FEM) ECT model for oil-water two-phase flow measurement; using ANSYS to establish a three-dimensional model of capacitance array sensors. An effective method of calculating capacitance is proposed, and the inter-electrode sensitivity distribution is described in this thesis.2. Analyzing the impacts of seven structural parameters of capacitance sensor on the distribution of sensitivity. The structural parameters include the size of electrodes, the thickness of pipe wall, the thickness of shielding layer, and the depth of radial shielding, and so on. In order to maximize the magnitude of the overall sensitivity and other system performances, a radial basis function (RBF) neural network and a particle swarm optimization (PSO) algorithm are used to optimize sensor structure. The quality of reconstructed images is improved as a result of the optimization.3. Designing an ECT image reconstruction algorithm package that includes the Linear Back Projection (LBP) algorithm, the Tikhonov regularization algorithm, and the Landweber iteration algorithm. In accordance with this objective, the impacts of the control parameters of algorithms (such as the relaxation factor, regularization factor etc.) on the quality of images are studied and discussed in the thesis. The accuracy of image reconstruction and the anti-noise performance of three algorithms are analyzed on the basis of the image quality evaluation criteria of both image correlation coefficient and variance. A method of adaptively adjusting the regularization factor based on the capacitance measurement by means of using a neural network is proposed, that extends the application scope of Tikhonov algorithm.4. Setting up a comprehensive ECT simulation platform that can be used for oil-water two-phase flow measurement. The platform includes the modeling of two-phase flow, the modeling of sensor, a forward solver based on the finite element method, and an ECT image reconstruction algorithm package. The platform also integrates some of the special functions that are developed for various optimization purposes.The work described in this thesis provides a technical foundation for the ultimate designing of an ECT two-phase flow measurement system in China University of Petroleum. The simulation platform, the method of ECT sensor structure optimization, and the image reconstruction algorithm package can all be used for the actual development of ECT system.