【摘要】 电容层析成像(ECT)是监测多相流动的一种新技术,其主要技术难点在于克服反演过程中的不适定性,快速而又精确的重建图像。本文提出了一种利用有限元(FEM)以及多元线性回归的方法(MLR)建立ECT系统正向模型,用正则化方法解决ECT系统图像重建这一不适定性问题的图像重建算法,即通过先验知识的引入来降低反演方程的病态程度,改善求解条件以达到克服不适定性的目的,精确的重建图像。仿真结果表明,该方法能快速而且高精度的获得图像重建,具有很强的实用性。更多还原

【Abstract】 Electronic Capacitance Tomography(ECT) is a new technique for monitoring multiphase flow,whose tough point focuses on conquering ill-posed characteristic in inversing process so as to reconstruct image with fast and accurate efficiency.This paper proposes a scheme using Finite Element Method and Multiple Linear Regression to model ECT forward problem,using regularization to solve the ill-posed inverse problem in image reconstruction.In order to solve the ill-posed inverse problem,this method introduces the transcendental information to reduce infection of ill-pose property.Simulation results illustrate that this method is capable of reconstructing the two-phase flow cross-section fast and accurately,and is practical.更多还原