Regularization parameter optimization based on the constraint of Landweber algorithm for electrical capacitance tomography
| Autor: | Jin Liu, Kai Zhao, Jun Li, Ji Ou, Yuan-shen Huang, Mengtao Xie, Haima Yang, Yuan Baolong, Jie Ying |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Correlation coefficient
Computer science 0207 environmental engineering Inverse 02 engineering and technology Electrical capacitance tomography Iterative reconstruction 01 natural sciences Regularization (mathematics) Landweber iteration Computer Science Applications 010309 optics Tikhonov regularization Image reconstruction algorithm Modeling and Simulation 0103 physical sciences Electrical and Electronic Engineering 020701 environmental engineering Instrumentation Algorithm |
| Zdroj: | Flow Measurement and Instrumentation. 69:101620 |
| ISSN: | 0955-5986 |
| DOI: | 10.1016/j.flowmeasinst.2019.101620 |
| Popis: | Image reconstruction algorithms play an important role in practical applications of electrical capacitance tomography. In the present paper, a combined image reconstruction method is proposed, which takes the results of Landweber algorithm as the constraint condition of Tikhonov algorithm's regularization parameter, calculates the regular parameter, inverts the inverse matrix of sensitivity matrix, and finally obtains the dielectric constant distribution; thus, reconstructed images with improved clarity were obtained. Simulation test are carried out to evaluate and analyze the proposed method from image error, correlation coefficient, image reconstruction time, and anti-noise ability. The results revealed that the Tikhonov regularization algorithm had excellent anti-noise ability; thus, it significantly improved the clarity of reconstructed images and clearly distinguished the multi-phase flow pattern and distribution. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect