Research On Image Reconstruction Algorithm For Electrical Capacitance Tomography

Electrical capacitance tomography(called ECT),is a kind of tomography method based on low frequency capacitance and resistance measurement,which has been widely used in industrial process imaging and monitoring in recent years.Based on the principle of capacitance sensing and resistance measurement,ECT has the advantages of simple structure,low cost,no radiation and good safety.It is one of the hot spots in the research of process tomography.Firstly,based on the 12-electrode ECT imaging system,the composition,measurement and reconstruction principle of the ECT system are analyzed.The mathematical model of the distribution information of the two-phase ECT system is constructed and a 12-electrode two-phase flow simulation environment is set up.In addition,the traditional ECT flow pattern recognition and image reconstruction algorithms are studied,and the imaging characteristics of different algorithms are analyzed and summarized.In addition,the flow pattern identification of ECT system is studied for the low accuracy of current ECT flow pattern identification.Firstly,the capacitance data obtained from the capacitance sensor are processed by different flow patterns,and the capacitance characteristic values of various flow patterns are extracted.Secondly,the rough neural network is trained by using the characteristic parameters of the capacitance,and then the identification of the convective type is done by the rough neural network.Through the analysis of the experimental results of Matlab simulation environment,it is found that the imaging accuracy of this method is improved with respect to BP neural network.Finally,in response to the problem of image reconstruction in ECT technology,the feasibility of applying convolutional neural network(called CNN)to ECT image reconstruction is studied.On the basis of in-depth research for convolution neural network for the more time-consuming process of deepstructure and training issues,and a fast convergence convolution neural network(called FCCNN)image reconstruction method is proposed.Then,the simulation results are compared and analyzed with the experimental results of the traditional algorithm.The experimental results show that the improved algorithm has definitely improved on the image reconstruction efficiency and quality of the common flow pattern.