Research On Image Reconstruction Algorithms For Electrical Capacitance Tomography

In Electrical Capacitance Tomography(ECT), the sensing electrodes attached around the measured field is excited by voltage with alternating frequency and amplitude to obtain the corresponding capacitances, which are used to obtain the visual distribution of the permittivity within the measured field using reconstruction algorithms. This method has many advantages, such as non invasion, fast response, no radiation, low cost, which has a broad prospect in industrial multi-phase flow detection.The accuracy of image reconstruction algorithm is one of the problems, which determines the successful applications of ECT in the multi-phase flow detection, and the research on the reconstruction algorithm has been carried out. In this paper, the Kalman filtering is used in the image reconstruction of ECT. Kalman filter obtains the minimum variance estimate of the gray value of image by measurement capacitances, it makes full use of the statistical characteristics of noise. The main work and conclusions are as follows:(1) Kalman Filter is used in the image reconstruction of ECT, and the stability of the Kalman Filter equation for ECT is studied. The Kalman filter equation of ECT is proved to be unstable in theory, and the finial estimation values are affected by the initial values of the filter, which is studied in this paper. The optimal initial value combination of Kalman filter is obtained by simulation experiment. It can be seen in the simulation results that the quality of the reconstructed image by Kalman filter is better than Landweber, Tikhonov and Sensitivity coefficient method.(2) The ECT image reconstruction based on adaptive Kalman filtering is studied in this paper because of the time varying of the multi-phase flow pattern. A new ECT state model with system noise is proposed, and the system noise is used to reflect the variation of the flow pattern of the multiphase flow. The adaptive Kalman filtering method based on the maximum likelihood criterion is used to modify the system noise, and the adaptive Kalman filtering equations are established, which are tested in simulation and experiment. The results show that this method can effectively improve the quality of reconstructed images.