Hydromechanics-based flow velocity estimation using single ERT sensor

Abstract

Multi-phase flow detection is an important task in flow velocity estimation. Owing to its fast response, no radiation, and low cost, the electrical resistance tomography (ERT) technique with a pair of sensors is typically used to estimate the flow velocity based on the cross-correlation (CC) measurement principle. However, due to the unreasonable a priori assumption, uncertain parameter, and limitation of ERT, this remains inaccurate and unsteady under complex and multi-flow conditions. We propose a novel flow velocity computation method that uses a single ERT sensor instead of a pair. Different from existing methods that are based on the CC principle, both the Churchill and Bernoulli equations in hydromechanics are used to compute flow velocity, based on which an inherent relation among the flow velocity, pressure difference, and density value in a detected pipeline is derived. Since the pressure difference and density value can be accurately computed from measurements from a single ERT sensor, the estimated flow velocity can greatly overcome the limitations when a pair of ERT sensors is employed. The proposed method is validated via a set of experiments, showing that it is more accurate and steadier than existing methods.