Electrical Resistance Tomographic Measurment Based On Injecting Adjacent Symmetrical Dual Current Sources

Electrical Resistance Tomography(ERT) is a new kind of Process Tomography(PT) which is suitable for the two-phase flow measurement. Two-phase parameter acquired from ERT reconstructed image, so the quality of reconstructed image is very important. This thesis presents a deeply research work of dual-current source image reconstruct algorithms in ERT research aimed at increasing the quality of reconstruction image, and proposes improving plans.The main contributions and works are as follows:1. Summarized the developing process and present status of the research on ERT, and summed up the ERT system structure, technique character, future development prospect and so on. Analyzed the established of ERT model, the division of finite element mesh and the solution of forward problem, and introduced several ERT reconstruction algorithms.2. Reconstructed images of several different flows with the measuring voltage data under symmetrical dual-current source using the Sensitivity Theory Method (STM) and the Modified Newton Raphson (MNR) method. Introduced the concept of learning-factor and used the method of adaptively calculated regional mean resistivity filter performed on some regional elements to improve the MNR algorithm, and improved the quality of reconstruction images effectively.3. Applied a new approximation method in the problem of MNR iterate approximation, sine vector approximation algorithm to used the effective information bring by different phase of dual-current source adequately, and obtained preliminary results.This thesis researched the image reconstruction algorithm under dual-current source deeply, improved old algorithm, formed a useful foundation for the future research work on dual-current source ERT development.