Research On Image Reconstruction Algorithm And Flow Pattern Identification For Electrical Capacitance Tomography

Electrical Capacitance Tomography(ECT) is an important branch of Process Tomography(PT),which is mainly used for real-time imaging of multiphase flow inside pipelines and vessels.As an emerging technology,ECT has the advantages of low cost,non-intrusive,real-time and simple operation.Moreover,due to the rapid development of computer technology,its imaging speed and accuracy are greatly improved.Therefore,ECT technology has become a kind of An important means in the field of multiphase flow detection.This topic will focus on the ECT system image reconstruction algorithm and flow pattern identification research,the main work and results are as follows:1.The ECT image reconstruction algorithm based on sparsity is studied.ECT image reconstruction has nonlinearity and morbidity,which directly affects the accuracy of image reconstruction.In industrial production,some ECT images are sparse.The sparse reconstruction algorithm of Compressed Sensing(CS) theory is directly applied to ECT image reconstruction,which can effectively overcome the nonlinearity and morbidity of ECT image reconstruction.This paper takes Barzilai-Borwein Gradient Projection for Sparse Reconstruction(GPSR-BB) as an example.It can be seen from simulation and static experiments that the algorithm can effectively distinguish multiple objects and the fidelity of edge information is high.2.The ECT stream identification algorithm based on sparsity is studied.The normalized measured capacitance value signal is sparsely represented by the training sample set,which makes it sparse and satisfies the basic requirements of the compressed sensing theory application,and uses the Orthogonal Matching Pursuit(OMP) algorithm.The standard samples correspond to the sparse solution of the training sample set;finally,the attribution type of the flow pattern to be measured is determined according to the linear correlation between the sample to be tested and the standard sample sparse solution.It can be seen from simulation and static experiments that the method has strong anti-interference ability,and the accuracy of flow pattern identification can basically meet the requirements of industrial production.