Correcting for two-phase flow in a digital flowmeter
View Patent ↗A flowmeter is disclosed. The flowmeter includes a vibratable conduit, and a driver connected to the conduit that is operable to impart motion to the conduit. A sensor is connected to the conduit and is operable to sense the motion of the conduit and generate a sensor signal. A controller is connected to receive the sensor signal. The controller is operable to detect a single-phase flow condition and process the sensor signal using a first process during the single-phase flow condition to generate a validated mass-flow measurement. The controller is also operable to detect a two-phase flow condition and process the sensor signal using a second process during the two-phase flow condition to generate the validated mass-flow measurement.
1. A flowmeter comprising:
a vibratable conduit to receive a two-phase mixture that includes a liquid phase and a gas phase;
a driver connected to the conduit and operable to impart motion to the conduit;
a sensor connected to the conduit and operable to sense the motion of the conduit and generate a sensor signal; and
a controller connected to receive the sensor signal, the controller configured to:
determine a first property of the two-phase mixture from the sensor signal; and
determine a mass flow rate measurement of the liquid phase of the two-phase mixture based on a non-parametric mapping function and the first property.
2. The flowmeter of claim 1 wherein the non-parametric mapping function comprises a neural network.
3. The flowmeter of claim 2 wherein the first property comprises a raw mass flow rate of the mixture and, to determine the mass flow rate measurement of the liquid phase, the controller is configured to input the raw mass flow rate into the neural network to generate an error correction factor and to correct the raw mass flow rate based on the error correction factor.
4. The flowmeter of claim 3 wherein the neural network receives at least one input parameter in addition to the raw mass flow rate and applies a set of predetermined coefficients to the additional input parameter and the raw mass flow rate.
5. The flowmeter of claim 4 wherein the neural network is a multi-layer perceptron neural network comprising:
an input layer for receiving the additional input parameter and the raw mass flow rate;
a hidden layer having processing nodes for applying the set of predetermined coefficients to the additional input parameter and the raw mass flow rate; and
an output layer that generates an output parameter.
6. The flowmeter of claim 4 wherein the additional input parameter includes one of a temperature parameter, a damping parameter, or a density parameter.
7. The flowmeter of claim 2 wherein the neural network is a radial basis function network.
8. The flowmeter of claim 1 wherein the mass flow rate measurement is a validated mass flow rate measurement comprising a corrected mass flow rate generated from a raw mass flow of the mixture and an uncertainty parameter calculated by the controller.
9. The flowmeter of claim 1 wherein the controller is configured to detect a two-phase flow condition.
10. A flowmeter controller for use with a vibratable conduit configured to receive a two-phase mixture that includes a liquid phase and a gas phase, a driver connected to the conduit and operable to impart motion to the conduit, and a sensor connected to the conduit and operable to sense the motion of the conduit and generate a sensor signal, the controller comprising:
at least one processing device configured to:
receive the sensor signal;
determine a first property of the two-phase mixture from the sensor signal; and
determine a mass flow rate measurement of the liquid phase of the two-phase mixture based on a non-parametric mapping function and the first property.
11. The controller of claim 10 wherein the mapping function is a neural network.
12. The controller of claim 11 wherein the first property comprises a raw mass flow rate of the mixture and, to determine the mass flow rate measurement of the liquid phase, the processing device is configured to input the raw mass flow rate into the neural network to generate an error correction factor and to correct the raw mass flow rate based on the error correction factor.
13. The controller of claim 12 wherein the neural network receives at least one input parameter in addition to the raw mass flow rate and applies a set of predetermined coefficients to the additional input parameter and the raw mass flow rate.
14. The controller of claim 13 wherein the neural network is a multi-layer perceptron neural network comprising:
an input layer for receiving the additional input parameter and the raw mass flow rate;
a hidden layer having processing nodes for applying the set of predetermined coefficients to the additional input parameter and the raw mass flow rate; and
an output layer that generates an output parameter.
15. The controller of claim 13 wherein the additional input parameter includes one of a temperature parameter, a damping parameter, or a density parameter.
16. The controller of claim 11 wherein the neural network is a radial basis function network.
17. The controller of claim 10 wherein the mass flow rate measurement is a validated mass flow rate measurement comprising a corrected mass flow rate generated from a raw mass flow of the fluid and an uncertainty parameter calculated by the controller.
18. The controller of claim 10 wherein the processing device is configured to detect a two-phase flow condition.
19. A method of generating a measurement of a property of a two-phase mixture flowing through a conduit, wherein the two-phase fluid includes a liquid phase and a gas phase, the method comprising:
imparting motion to the conduit
sensing motion of the conduit;
determining a first property of the two-phase mixture from the sensor signal; and
determining a mass flow rate measurement of the liquid phase of the two-phase mixture based on a non-parametric mapping function and the first property.
20. The method of claim 19 wherein the mapping function is a neural network.
21. The method of claim 20 wherein the first property comprises a raw mass flow rate of the mixture and determining a mass flow rate measurement comprises inputting the raw mass flow rate into the neural network to generate an error correction factor and correcting the raw mass flow rate based on the error correction factor.
22. The method of claim 21 wherein inputting the raw mass flow rate into the neural network to generate an error correction factor comprises inputting at least one input parameter in addition to the raw mass flow rate into the neural network such that the neural network applies a set of predetermined coefficients to the additional input parameter and the raw mass flow rate.
23. The method of claim 22 wherein the additional input parameter includes one of a temperature parameter, a damping parameter, or a density parameter.
24. The method of claim 19 wherein the mass flow rate measurement is a validated mass flow rate measurement comprising a corrected mass flow rate generated from a raw mass flow of the mixture and an uncertainty parameter calculated by the controller.
25. The method of claim 19 further comprising detecting a two-phase flow condition.