Identification Of Flow Regimes And Image Reconstruction For Electrical Capacitance Tomography System

Electrical capacitance tomography is a branch of process imaging technologies,which is applied on the measurement and analysis of the multi-phase flow detection and precision measurement in the industrial sealed pipelines.ECT has the advantages of low price,non radiation,fast response,simple structure and non intrusive,which is very significant for the development of industry.In recent years,the technology of software and hardware has been improved rapidly.Accordingly it is important theoretical significance and practical application value to study the visualization technology of fluid in the pipeline.This paper focuses on the flow regime identification and image reconstruction in ECT system.The main work of this paper includes the following contents:Firstly,the three major components of ECT system are analyzed.The mathematical model and the distribution function of sensitivity are given for ECT system,and as the theoretical basis the principle of image reconstruction is analyzed.Secondly,considering the symmetric structure of the sensor,the collected capacitance values are extracted to reduce the dimension of the input parameters.After the study of multi-feed forward and topology structure of BP neural network,the traditional BP neural network is improved in two aspects: adaptive learning rate is considered for the problem of slow learning speed,and additional momentum is adopted for the prone to local minimum problem.Applying the improved BP neural network algorithm on flow pattern identification is effective.Finally,for the low precise of image reconstruction,in this paper,combinedthe advantages of the Approximate Conjugate Gradient Pursuit(ACGP)algorithm to track the direction,Stage wise Weak Conjugate Gradient Pursuit(SWCGP)tracking algorithm can reduce the number of iterations,Spectral Projected Gradient Pursuit(SPGP)tracking algorithm is adopted to reconstruct images.Compared with Linear Back Projection algorithm,according to the error and iteration of experiment results,the algorithm has a better reconstruction images.