Study On The Inverse Problem And Optimal Design Of Three-dimensional ERT System Based On Double Exciting Layers

ERT(Electrical Resistance Tomography),which characterized by lower costs,non-invasive and visualization,is one of the most promising research subjects in two-phase flow detection fields.At present,research has mainly been focusing on two-dimensional space,with little research on three-dimensional space.However,the three-dimensional information such as the geometrical shape,relative location and space distribution of dicrete phase is needed to get in most cases.In order to realize the three-dimensional observation of the measured field,the study of three-dimensional ERT technology and optimization design of array electrodes is necessary.The thesis was focus on the ERT technology,based on the simulations in the finite element analysis software ANSYS and fluid analysis software FLUENT and also experiments in laboratory.Based on the way of positive problem,take an explorative research on the multi-layers electrodes combinational exciting mode and RSM optimization methods.A modified image reconstruction for the direct three-dimensional tomography algorithm was made based on the traditional two-dimensional tomography algorithm,the simulative and physical experiments were proceeded to verify the optimized results.The main work of this thesis is showed as follow:(1)The mathematical model of ERT system was established based on the assumption of the quasi-static current field,the methods of deducing the system equation using finite element method and calculating the voltage distribution was discussed,the solution of forward problem was completed.(2)According to the finite element analysis software ANSYS soving the positive problem of three-dimensional ERT system,make sure the structure of ERT sensor:three-layers and eight-electrodes.Based on this structure,we provide a new exciting method:multi-layers combinational exciting method,and the influence of multi-layers electrodes exciting mode and the structural parameters were anslysised.(3)Considering the limitation of orthogonal test in three-dimensional ERT senor,multi-objects optimization algorithms and response surface method were proposed.Focusing on the unmatched problem in traditional RSM method,the model was optimized and tested to be feasibility and accuracy.The three optimal results were compired and analysised.(4)A district three-dimensional ERT imaging reconstruction algorithm was proposed for taking away the disadvantages of two-dimensional interpolation algorithm.The algorithm of sensitivity coefficient matrix was improved account of the problems such as dimension increasing,calculation complexity;the speed was increased by more than 10 times afer algorithm imorovement.An adaptive threshold value was added into the Tikhonov regularization algorithm,so it can canculate the threshold value between its iteration,and the quality of image reconstruction was improved about 10%.(5)Verify the experimental results.In order to simulating the lively process results,the gas-liquid two phases flow in vertical pipe was simulated in fluid analysis software FLUENT by MHD and UDS model.On the other way,the physical experiment was proceeded by three-dimentional ERT system under the laboratory condition,the output of dynamic and static simulation were consistent with physical results,which can proved the feasibility of three-dimensional ERT in actual measurement situation.Three-dimensional ERT system was proposed,the simulative and physical experiment was proceeded using optimized multi-layers electrodes sensor,the results revealed that the correlation coeddicient achieve above 0.7 after optimization.