Study On ?-conditioned Problems In Image Reconstruction Of Oil-air Lubrication ECT System

Image reconstruction is an important step to monitor oil-air two-phase flow parameters online in oil-air lubrication pipelines.But the ill-conditioned problems in image reconstruction of oil-air lubrication ECT system is further exacerbated,because the oil-air lubrication system has the characteristics of small pipe diameter,thin oil film and air-liquid two-phase flow with less liquid and more air.Therefore,ill-co nditio ned proble ms in image reconstructio n of oil-air lubrication ECT system are studied in-depth in this thesis,and the main research work and conclusions are as follows,(1)A mathe matical model for image reconstruction of the oil-air lubrication ECT system is established,and the main causes and influencing factors of the ill-conditioned problems are analyzed.The ill-conditioned degree of the sensitivity matrix in the image reconstruction of oil-air lubrication ECT system is diagnosed employed the condition index—variance decomposition ratio(CIVDP)method.The results show that there is strong ill-conditioning between the sensitivity matrix column vectors in image reconstruction of oil-air lubrication ECT system.The ill-conditioning is related to the characteristics of oil-air lubrication EC T system,such as oil film thickness and number of capacitor plates.The degree of ill-conditioning can be weakened effectively when the regression analysis and residual processing of the sensitivity matrix column vectors are done.(2)The regularization parameters in image reconstruction of oil-air lubrication ECT system are determined by U-curve method,the ill-conditioning of the sensitivity matrix after regularization is analyzed.The first image reconstruction of the pipeline section of the oil-air lubrication ECT system is completed by the LBP algorithm and the TIKHONOV regularization algorithm respectively.The secondary image reconstruction of the pipeline section is completed after the threshold filtering threshold is optimized for the gray distribution matrix of the first reconstructed image.The results show that the regularization parameters selected by the U-curve method have a significant effect on weakening the ill-conditioning of the sensitivity matrix compared with the L-curve method.Threshold filter threshold optimization can improve the image reconstruction quality of oil-air lubrication ECT system.(3)A three-dimensional physical model and finite element model of oil-air lubrication annular flowing pipe and ECT system,employed COMSOL software,is established and the capacitance value of the sensor is obtained by combining electromagnetic field theory after simulation.A RBF mapping model with one-to-one correspondence from the capacitance value to image pixels grayscale is constructed;the RBF algorithm is applied to the image reconstruction process of oil-air lubrication two-phase flow;and the algorithm is verified by the relative error of image(IME)and image correlation coefficient(CORK).The results show that the ill-conditioned problems caused by sensitivity matrix can be avoided by RBF algorithm,which is suitable for ECT image reconstruction of oil-air lubrication in terms of accuracy and relevance of imaging.The image reconstruction quality has been greatly improved comparing with other image reconstruction algorithms.It will contribute to the subsequent study on the internal flow characteristics of oil-air lubrication pipelines.