Wet Gas Metering Using Multi-Hole Orifice and V-Cone in series with Neural Network

Due to the complexity and variability of the working conditions, the high precision metering of wet gas is a challenge. This paper innovatively proposes a heterogeneous double differential pressure wet gas metering method with a multi-hole orifice plate and a V-cone in series, utilizing a BP neural network. The two types of heterogeneous throttling devices exhibit self-rectifying and are sensitive to the fluctuating features of wet gas, making them structurally suitable for wet gas metering. This method extracts the mean values of differential pressure, pressure, and temperature signals from two throttle devices as static characteristics, and the probability density functions and variational mode decomposition of differential pressure and pressure from the two throttle devices as dynamic features. Because of The neural network possesses powerful nonlinear fitting ability, by combining static and dynamic features as inputs to the neural network, with gas-phase and liquid-phase flow rates as outputs, a double differential pressure wet gas metering model based on a BP neural network is established, thereby achieving wet gas metering without gas-liquid separation. This achievement provides a new solution for wet gas flow metering and is expected to play an important role in the field of international multiphase flow metering.