Finite element mesh optimisation for improvement of the sensitivity matrix in electrical resistance tomography

作者：Xiao Liqing; Xue Qian^*; Wang Huaxiang

来源：IET Science Measurement & Technology, 2015, 9(7): 792-799.

DOI：10.1049/iet-smt.2014.0319

• flow visualisation
• finite element analysis
• genetic algorithms
• image reconstruction
• Newton-Raphson method
• finite element mesh optimisation
• sensitivity matrix
• electrical resistance tomography
• genetic algorithm
• image reconstruction
• modified Newton-Raphson algorithm

In electrical resistance tomography (ERT), finite element (FE) meshes of different topologies lead to different sensitivity matrices, which have great influence on the imaging quality of the ERT system. To improve the ill-posedness of the traditional sensitivity matrix and thus to improve the image quality, the reciprocal of the sensitivity matrix condition number is designed as the fitness function, based on which the modified genetic algorithm is utilised to optimise the topology of the FE mesh offline, and the optimised FE mesh is used to generate the sensitivity matrix, which is thereafter applied to image reconstruction using the modified Newton-Raphson algorithm. The feasibility of the proposed method is demonstrated in both simulation and prototype experiments. Comparisons among the mesh the authors optimised, the traditional mesh and meshes modified with other methods (all the meshes have the same number of nodes and elements) show that the proposed method has obviously reduced the condition number of the sensitivity matrix, and thus enhanced the imaging quality.

• 出版日期2015-10
• 单位徐州工程学院; 中国民航大学; 天津大学