Simulation study of Electrical Capacitance Tomography

By using ANSYS for formulation and solution of the finite element, a model of Electrical Capacitance Tomography electrodes array was constructed. Based on the model, the sensitivity matrix of the sensor was calculated and sensitivity response was analyzed. The flow pattern of two phase flow was identified by using RBF neural network at first, and then the image was reconstructed by using BP neural network. A method was introduced that take the characteristic parameters of stratified flow, annular flow and core flow as the output of BP Neural Network for image reconstruction. Simulation results showed that it can rapidly reconstruct satisfactory images.