电阻抗断层成像的MPSO-MNR算法研究

计算机工程与应用 ›› 2013, Vol. 49 ›› Issue (9): 29-32.

电阻抗断层成像的MPSO-MNR算法研究

张辉1，李颖2，王西明1，张小娣1

1.咸阳师范学院物理与电子工程学院，陕西咸阳 712000
2.伦敦大学学院土木、环境与测绘工程系，英国伦敦 WCIE 6BT

出版日期:2013-05-01 发布日期:2016-03-28

MPSO-MNR algorithm study of 2-D electrical impedance tomography

ZHANG Hui1, LI Ying2, WANG Ximing1, ZHANG Xiaodi1

1.College of Physics and Electronic Engineering, Xianyang Normal University, Xianyang, Shaanxi 712000, China
2.Department of Civil, Environmental and Geomatic Engineering, University College London, London WCIE 6BT, UK

Online:2013-05-01 Published:2016-03-28

摘要/Abstract

摘要： 基于修正粒子群算法（MPSO）和修正的牛顿-拉夫逊（MNR）算法的优点和局限，提出MPSO-MNR算法，通过对研究的平面圆形求解域采用有限元法进行剖分，电流注入采用三角电流法的园域内单个、两个仿真目标采用该算法进行电阻抗断层静态重构。采用定义的适应值函数和误差总和作为评价重构质量的物理量。数值仿真结果表明，在一定迭代次数内，提出的MPSO-MNR算法对求解域内目标位置定位准确，能够较准确反映场域内电阻率的分布。

关键词: 修正的粒子群算法, 电阻抗断层成像, 修正的牛顿-拉夫逊算法

Abstract: This paper which is based on the advantages and limits of the Modified Particle Swarm Optimization（MPSO）algorithm and Modified Newton-Raphson（MNR）algorithm proposes the MPSO-MNR algorithm, and presents the numerical simulation of 2-D Electrical Impedance Tomography（EIT）. In this paper, the MPSO-MNR algorithm is used to study the imaging of the EIT. The circle-domain solved is discretized into irregular polygons by the Finite Element Method（FEM）. Trigonometric current pattern and the MPSO-MNR are adopted to reconstruct the impedance distribution, and the reconstruction results are evaluated by the fitness value function and the total error. The numerical simulation results show that the MPSO-MNR algorithm can locate the targets accurately and reflect the impedance distribution within certain iterations.

Key words: Modified Particle Swarm Optimization（MPSO）algorithm, Electrical Impedance Tomography（EIT）, Modified Newton-Raphson（MNR）algorithm

张辉1，李颖2，王西明1，张小娣1. 电阻抗断层成像的MPSO-MNR算法研究[J]. 计算机工程与应用, 2013, 49(9): 29-32.

ZHANG Hui1, LI Ying2, WANG Ximing1, ZHANG Xiaodi1. MPSO-MNR algorithm study of 2-D electrical impedance tomography[J]. Computer Engineering and Applications, 2013, 49(9): 29-32.