电阻层析成像系统阵列电极宽度优化
Width Optimization of Array Electrode for Electrical Resistance Tomography System

在电阻层析成像技术中, 其他条件相同的情况下, 不同宽度阵列电极对应的灵敏度矩阵不同. 为了提高算法反演精度, 以敏感场均匀分布时灵敏度矩阵条件数的倒数为适应度函数, 在优化有限元模型拓扑结构的同时, 利用改进粒子群算法优化电阻层析成像系统阵列电极宽度, 并将优化结果应用于改进牛顿-拉夫逊算法. 仿真实验结果表明, 相比其他两种不同宽度阵列电极与拓扑结构的有限元模型, 优化结果对应的灵敏度矩阵条件数分别降低了36.444 3% 和24.345 6% , 有效改善了灵敏度矩阵的病态性, 从而提高了算法反演精度.
In electrical resistance tomography(ERT)，array electrodes of different widths lead to different sensitivity matrices，other things being equal. In order to enhance the inversion accuracy of the image reconstruction algorithms，the reciprocal of the condition number of the sensitivity matrix with homogeneous distribution of resistivity in the sensitive field，was designed as the fitness function，based on which the width of array electrode for electrical resistance tomography system and the topology of the finite element mesh were optimized using the improved PSO algorithm，and the best result obtained by the improved PSO algorithm was used to generate the sensitivity matrix，which was thereafter applied to image reconstruction using the improved Newton-Raphson algorithm. Simulation results demonstrate that，compared to the other two finite element meshes of different topologies and width，the condition number is reduced by 36.444 3% and 24.345 6%，the ill-posedness is improved effectively，and thus it is propitious to enhance the inversion accuracy of the algorithm