Design Of ERT Sensor And Image Reconstruction Based On COMSOL

Electrical Resistance Tomography is a mode of Electrical Tomography.Its imaging object is the distribution of electrical conductivity.By arranging the electrode arrays on the boundary of the measured field,a sensitive field from different directions is formed.Through the measurement of potential difference between the electrodes,the projection data from different scanning directions are obtained.Using the appropriate image reconstruction algorithm,the distribution of electrical conductivity in the field is calculated by the projection data vector,and the distribution of the medium in the measured field is visualized.Because of its characteristics of non invasion,no radiation and fast response,the technology has broad application prospects in multi-phase flow measurement and medical clinical monitoring.It is currently a research hot-spot in tomography technology.In this thesis,multiple physics software Comsol Multiphysics is used to build finite element models of ERT system.The sensitive field of the ERT sensor is distributed in a three-dimensional space,but the modeling of the ERT sensor is usually carried out in two dimensions,and image reconstruction of the ERT system is also given in the form of two-dimensional graph.In this thesis,a two-dimensional finite element model of ERT sensor model is established,and the influence of electrode material and electrode width on the performance of the sensor is analyzed.Then the three-dimensional finite element model of ERT sensor is established,and the influence of electrode length on the performance of the sensor is analyzed.By analyzing the influence of sensor structure parameters on voltage response and sensitivity field,the structural parameters of ERT sensor are determined.The axial sensing space of the sensor is studied in this thesis.The Landweber iterative algorithm is used as the image reconstruction algorithm,and the initial value of the iteration is the result of the linear back projection algorithm.The sensitivity matrix used for reconstruction is improved sensitivity matrix.According to the evaluation method of ERT reconstruction image,two kinds of sensitivity evaluation indexes are defined.Image reconstruction is achieved by distributing the state of different media in different axial positions.The axial sensing space of ERT sensor is studied by comparing the visual effect and sensitivity evaluation index of the reconstructed image.This work has practical significance for the interpretation and application of ERT reconstruction images.The sensitivity of the center of ERT sensor is much lower than that of the edge region,which is difficult for high quality image reconstruction.To solve this problem,a new image reconstruction method based on improved sensitivity matrix is proposed in this paper.Starting with the positive problem,the range of sensitivity of each projection direction is limited by the threshold of filtering of the sensitivity matrix element.The relative sensitivity of the low sensitivity region is enhanced and the relative sensitivity value of the central region of the sensor is enhanced,thus reducing the morbid state of the sensitivity matrix.Through simulation and experiment,the image reconstruction of various electrical conductivity distribution is carried out,the error of image reconstruction and the model error of positive problem is compared.The validity and applicability of the proposed method are verified.This paper attempted to obtain sensitivity matrix by using 3D ERT sensor model.The sensitivity matrix has a narrow sensitivity band,and it has an advantage in distinguishing multiple target distribution.