Efficient Image Reconstruction Algorithm for ECT System Using Local Ensemble Transform Kalman Filter

Abstract

One of the vital processes that should be monitored and analyzed continuously in the oil-gas and petroleum-related industries is the multi-phase flow inside pipes. Multi-phase flow means flowing two or more phases of gas, liquid, or solid inside a pipe. Electrical Capacitance Tomography (ECT) is a feasible and economical solution for monitoring dynamic applications. The ECT system offers the benefits of no radiation, non-intrusive, and non-invasive. Despite its potential, ECT systems deployment's major limitation is the crucial need to develop rapid image reconstruction algorithms. In this paper, a Local Ensemble Transform Kalman Filter (LETKF) is developed as a non-linear system estimator for reconstructing images in the ECT system. This method manages each node of the model independently by assimilating only the observations at a predefined distance. The localized approach of the LETKF gives it high computational efficiency allowing it to be applied to large dynamic systems. A quantitative analysis using Image Error (IE) and Coefficient Correlation (CC) measures has been applied to prove the effectiveness of the proposed algorithm. Indeed, the IE has been significantly decreased (around 62%), and the CC greatly increased (around 58%). Then, the influence of the noise was discussed. The results are promising and prove the algorithm feasibility.

Keywords

• ECT
• image reconstruction
• Kalman filter
• multi-phase flow