Deep learning models for improved accuracy of a multiphase flowmeter

Measurement of oil and gas two-phase flow with variable regimes relies to a large extent on patterns their transitions. Using multiphase flowmeters in flows high volume fractions is therefore usually associated uncertainties. This work presents dynamic neural network method measure the rate using nonlinear autoregressive exogenous inputs (NARX). Total temperature total pressure are used as obtained results compared multilayer perceptron (MLP). Comparison between modeling experimental data shows that NARX can predict less error MLP model, e.g. an absolute average percentage deviation (AAPD) 0.68% instead 1.02%. The present hence be seen proof-of-concept study should motivate further applications deep learning models facilitate enhanced accuracy metering.