A feature based solution to Forward Problem in Electrical Capacitance Tomography

A new feature-based technique is introduced to solve the nonlinear Forward Problem (FP) of the Electrical Capacitance Tomography (ECT) with the target application of monitoring the metal-fill profile in Lost Foam Casting (LFC) process. The new technique to solve the FP is based on key features extracted from the metal distributions and the Correction Factor (CF). The CF is predicted by an Artificial Neural Network (ANN) based on key distribution features. The CF adjusts the linear solution of the FP for nonlinear effects. The data for the ANN training was generated through ANSYS finite element analysis and the codes written in MATLAB. The ANN was implemented using MATLAB Neural Network Toolbox. This approach shows promising results. The ANN was able to learn the effect of these features on the CF with the % RMS error of 2.21 for training data. For the previously unseen test metal distributions, the average RMS error was 2.2%.