Forward simulation for electromagnetic tomography and optimized design for sensor parameters
电磁成像正演仿真与传感器参数优化设计

Firstly, FEM (Finite Element Method) is used to analyze forward problem and compute real potential and imaginary potential at different manifold distribution. Secondly, study the effect of the number of sensors, the angle and electrode thickness. Thirdly, the optimized parameters of EMT are gotten by neural network method. The parameters of sensor design are optimized by using neural network method. It can be found that the effect of potential is mainly affected by medium distribution, especially the imaginary potential distribution. The different number of sensors affects the measurement largely and the potential is affected by different angle and electrode thickness to some extent.