Study On Image Reconstruction Algorithms For Electrical Capacitance Tomography System

Electrical Capacitance Tomography(ECT) technology is one of process the tomography technologies which is based on the principle of capacitance-sensitive. It is a detection technology that expresses processes characteristics in real-time online. Aero engine is the core component of the aircraft. In order to obtain the real-time images and monitor the fluid in the tail pipe, this paper uses electrical capacitance tomography technology to monitor aero-engine tail pipe.In recent years, ECT technology has been developed in a number of areas, but there are many problems, such as hardware and image reconstructions.In order to analyze the ECT system in depth, the main studies of this paper are on the following aspects:Firstly, 16-electrodes ECT sensor system is simulated by the COMSOL Multiphysics which is an electromagnetic finite element simulation software. Through solving the forward problem of the ECT, the equivalents cloud maps of the node potential and vector of electric force for the capacitance sensor model can be evaluated and then the simulation results are analyzed.Secondly, MATLAB based software is programmed for imaging processing. Multiple algorithms are achieved in this software, such as, linear back projection algorithm, SIRT algorithms, Landweber iterative algorithm, conjugate gradient algorithm, etc. The image reconstruction results of these imaging algorithms are compared by using the correlation coefficient and imaging time.Thirdly, due to a preliminary study, the quality of the image get from those existed algorithm is still improvable. Considering the ill-conditioned problem of the Electrical Capacitance Tomography image reconstruction, a novel regularization method is proposed based on Tikhonov regularization method. This method introduces a mathematical transformation, which overcome the shortcomings of the disruption of conventional regularization.Fourthly, using Tikhonov regularization method, Conjugate Gradient method and the modified Conjugate Gradient method reconstruct the image which is about flow patterns existing in the aero-engine tail pipe, and analyze the results.Finally, the wavelet analysis technique is applied in image post-processing to improve the image quality by denoising, enhancement and image fusion.