FIELD OF INVENTION

The present invention relates to a fluid measurement device. More particularly, the present invention is an improved flow meter for determining fluid characteristics.

BACKGROUND OF THE INVENTION

Flow meter is a device for quantifying fluid movement. For instance, volumetric flow rate of a fluid can be measured using target flow meter and vortex flow meter. Target flow meter comprises a target for determining the force of fluid impinging thereon in which the force is measured by mechanical stress which is later converted into velocity. However, the mechanical stress is subjective to temperature variation and the force of fluid is density dependent. Therefore, these parameters need to be corrected. On the other hand, vortex flow meter comprises of a bluff and a vortex sensor in which the fluid creates vortices as it impinges the bluff. The vortex sensor measures the frequency of the vortices and converts it into velocity. These velocities are then converted into volumetric flow rate based on mathematical algorithm. These types of flow meter may be used to determine the mass flow rate of a fluid using a pre-calibrated density. However, they face difficulties in the presence of temperature and pressure fluctuation and multiphase flow application. This is due to the non-linear relationship between the density, temperature, and pressure of multiphase flow. Alternatively, Coriolis flow meter may be used to determine mass flow rate of a fluid. It is based on the principle of motion mechanics. However, flow meter tube mechanic motion is temperature and pressure dependent. Therefore, Coriolis flow meter is highly sensitive to fluid temperature and pressure and ambient environment in which external vibration and magnetic field will substantially affect the accuracy thereof. Further, vapour phase in a multiphase flow may result in substantial measurement error due to density variation.

Machine learning may be used to overcome the drawbacks of conventional flow meter. Instead of relying on complicated mathematical algorithms, a machine learning module uses correlations to determine the outputs based on the inputs. For instance, Ahmadi et al. discloses an artificial neural network (ANN) for predicting oil flow rate of a reservoir, wherein the ANN comprises three layers with temperature and pressure as the inputs. In addition, China Patent Number CN106918377B discloses a flow meter for use in a production system using temperatures and pressure drops as inputs.

Existing flow meters have limited inputs and therefore, limit the accuracy and reliability of the outputs. Additionally, the inputs are obtained using sensors which may malfunction during operation. Furthermore, the machine learning module may not be able to notice such malfunction of sensors. Consequently, these aspects adversely affect the accuracy and reliability of the flow meter.

Therefore, it is essential to provide an improved flow meter that is capable of obtaining a plurality of inputs in order to improve the accuracy and reliability of the output fluid characteristics while notifying a user of a potential malfunction within the improved flow meter. The present invention provides a solution to the aforementioned problems.

SUMMARY OF INVENTION

One aspect of the present invention is to provide an improved flow meter for determining fluid characteristics with improved accuracy and reliability by using a sensing module, processing module, and neural network module. In particular, the sensing module determines inputs such as mechanical stresses, a vibration, a temperature, a first pressure, and a second pressure. Subsequently, the processing module uses the inputs to determine a voltage, a frequency, a resistance, current, and a pressure difference. Thereafter, the neural network module determines fluid characteristics such as a first mass flow rate, fluid density, and liquid fraction. Additionally, the neural network module determines a second mass flow rate using the pressure difference to further determine the fluid characteristics with improved accuracy and reliability while serving as a redundancy measurement.

Another aspect of the present invention is to provide an improved flow meter for determining potential malfunction within the improved flow meter and notifying a user via an alert module.

At least one of the preceding aspects is met, in whole or in part, in which the embodiment of the present invention describes an improved flow meter (100) for determining fluid characteristics comprising (a) means (101) for holding a probe (102); (b) a sensing module (103) operatively coupled to the means (101) for holding the probe (102), wherein the sensing module (103) comprises a mechanical stress sensor to determine mechanical stresses produced when a fluid impinges on the probe (102), a vibration sensor to determine vibration induced by vortices when the fluid impinges on the probe (102), a heat sensor to determine temperature of the fluid, and a first fluid pressure sensor (104) to determine a first pressure of the fluid; (c) a processing module (105) operatively coupled to the sensing module (103) to convert the mechanical stresses into voltage, the vibration into frequency, the temperature into resistance, and the first pressure into current; and (d) a neural network module (106) operatively coupled to the processing module (105) in which the neural network module (106) uses the voltage, frequency, resistance, and current to determine fluid characteristics including a first mass flow rate, fluid density, and liquid fraction, characterised in that the sensing module (103) further comprises a second fluid pressure sensor (107) operatively coupled to the sensing module (103) such that the processing module (105) is able to determine a pressure difference between the first fluid pressure sensor (104) and the second fluid pressure sensor (107), thereby allowing the neural network module (106) to determine a second mass flow rate using the pressure difference to further determine the fluid characteristics with improved accuracy and reliability while serving as a redundancy measurement.

Preferably, the mechanical stress sensor is a strain gauge sensor.

Preferably, the vibration sensor is a piezo electric sensor.

Preferably, the heat sensor is a resistance temperature detector sensor.

Preferably, the first fluid pressure sensor (104) and the second fluid pressure sensor (107) are pressure transducers.

In another preferred embodiment of the present invention, the improved flow meter (100) further comprises an output module (108) operatively coupled to the neural network module (106) for displaying the determined fluid characteristics thereon.

Still, in another preferred embodiment of the present invention, the improved flow meter (100) further comprises an alert module (109) operatively coupled to the neural network module (106) to notify a user in event where potential malfunction occurs within the improved flow meter.

BRIEF DESCRIPTION OF THE DRAWING

For the purpose of facilitating an understanding of the present invention, there is illustrated in the accompanying drawings the preferred embodiments from an inspection of which when considered in connection with the following description, the present invention, its construction and operation and many of its advantages would be readily understood and appreciated.

Figure 1 shows an improved flow meter (100) according to a preferred embodiment of the present invention.

Figure 2 illustrates a deviation between the first mass flow rate and the second mass flow rate determined by the neural network module (106).

DETAILED DESCRIPTION OF THE INVENTION

Hereinafter, the present invention shall be described according to the preferred embodiments of the present invention and by referring to the accompanying description and drawings. However, it is to be understood that limiting the description to the preferred embodiments of the invention is merely to facilitate discussion of the present invention and it is envisioned that those skilled in the art may devise various modifications without departing from the scope of the appended claim.

The present invention relates to an improved flow meter (100) for determining fluid characteristics with improved accuracy and reliability. The following description is explained based on a preferred embodiment of the present invention as exemplified in Figure 1.

According to the preferred embodiment of the present invention, the improved flow meter (100) comprises means (101) for holding a probe (102). Preferably, the means (101) for holding the probe (102) is a semi-rigid cantilever arm. The probe (102) may be positioned substantially at the centre of a conduit as fluid velocity is generally highest at the center. Preferably, the probe is a target having a shape of plate or sphere.

Still, according to the preferred embodiment of the present invention, the improved flow meter (100) comprises a sensing module (103) operatively coupled to the means (101) for holding the probe (102) for collecting inputs. Preferably, the sensing module (103) comprises a mechanical stress sensor to determine mechanical stresses produced when a fluid impinges on the probe (102), a vibration sensor to determine vibration induced by vortices when the fluid impinges on the probe (102), a heat sensor to determine temperature of the fluid, and a first fluid pressure sensor (104) to determine a first pressure of the fluid. More preferably, the mechanical stress sensor is a strain gauge sensor, the vibration sensor is a piezo electric sensor, the heat sensor is a resistance temperature detector sensor, and the first fluid pressure sensor (104) is a pressure transducer.

Yet, according to the preferred embodiment of the present invention, the improved flow meter (100) comprises a processing module (105) operatively coupled to the sensing module (103) for processing the inputs collected by the sensing module (103). Preferably, the processing module (105) converts the mechanical stresses into voltage, the vibration into frequency, the temperature into resistance, and the first pressure into current. In an exemplary embodiment of the present invention, the processing module (105) utilizes power spectral density in which the vibration data provided thereto may be converted into a plurality of frequency bin in which the desired frequency bin may be selected as the desired inputs. Advantageously, this may filter the noises present in the vibration input.

Further to the preferred embodiment of the present invention, the improved flow meter (100) comprises a neural network module (106) operatively coupled to the processing module (105). Preferably, the neural network module (106) uses the voltage, frequency, resistance, and current provided by the processing module (105) to determine fluid characteristics including a first mass flow rate, fluid density, and liquid fraction.

In the present invention, the sensing module (103) further comprises a second fluid pressure sensor (107) operatively coupled to the sensing module (103) to determine a second pressure of the fluid. Preferably, the second fluid pressure sensor (107) is another pressure transducer. The second fluid pressure sensor (107) can be positioned at upstream or downstream of the first fluid pressure sensor (104) such that two different pressures of the fluid can be determined. Subsequently, the processing module (105) determines a pressure difference between the first pressure and the second pressure. Thereafter, the neural network module (106) determines a second mass flow rate using the pressure difference to further determine the fluid characteristics with improved accuracy and reliability such as an updated mass flow rate. Additionally, the second mass flow rate can serve as a redundancy feature of the improved flow meter (100) such as a backup measurement.

In an exemplary embodiment of the present invention, the neural network module (106) firstly determines a first mass flow rate based on the voltage, frequency, resistance, and current provided by the processing module (105). Simultaneously, the neural network module (106) determines a first fluid density and a first liquid fraction. Subsequently, the neural network module (106) uses the pressure difference received from the processing module (105) to determine a second mass flow rate using the pressure difference based on Equation 1.

Q = C _d EEA _d ]2p _gas AP (Equation 1) where

Q = second mass flow rate

C _d = discharge coefficient of the improved flow meter determined by calibration

(dimensionless)

E = velocity of approach factor, E = where ft is cross sectional of the probe divided by cross sectional area of tube internal diameter

E = Expansibility factor A _d = area of the improved flow meter throat at operating conditions

Pgas = gas density

AP = pressure difference

Then, the neural network module (106) determines a deviation between the first mass flow rate and the second mass flow rate. The deviation is due to the presence of liquid in a liquid-gas two-phase flow. In particular, the presence of liquid changes the density of the fluid, thereby adversely affecting the accuracy of the first mass flow rate and the second mass flow rate in a different magnitude, as illustrated in Figure 2. The second mass flow rate, which is based on the pressure difference, has a higher error as compared to the first mass flow rate. This is because the second mass flow rate is determined using a set of gas density data that is provided during calibration of the improved flow meter (100). The presence of liquid in the liquid-gas two-phase flow changes the fluid density. Consequently, the change in the fluid density may not be accurately accounted by the processing module (105) when determining the pressure difference. Nonetheless, the deviation between the first mass flow rate and the second mass flow rate is proportional to the liquid fraction of the liquid-gas two-phase flow. Therefore, the neural network module (106) computes another liquid fraction based on this deviation. Subsequently, the neural network module (106) uses the computed liquid fraction to determine an updated mass flow rate with improved accuracy. Additionally, the computed liquid fraction can be used to determine an updated fluid density. Preferably, the aforementioned operations can be repeated in the form of iteration to further improve the accuracy of the determined fluid characteristics. Advantageously, the improved flow meter (100) described herein is beneficial for determining fluid characteristics in a conduit having a liquid-gas two-phase flow.

According to another preferred embodiment of the present invention, the improved flow meter (100) further comprises an output module (108) operatively coupled to the neural network module (106) for displaying the determined fluid characteristics thereon. Still, according to another preferred embodiment of the present invention, the improved flow meter (100) further comprises an alert module (109) operatively coupled to the neural network module (106) to notify a user in event where potential malfunction occurs within the improved flow meter (100). The abnormal deviation may be due to a malfunction of one or more sensors in the sensing module (103) which results in an abnormal determination of fluid characteristics by the neural network module (106). In view of this, the neural network module (106) is configured to determine an abnormal deviation between the first mass flow rate and the second mass flow rate based on other information such as the first liquid fraction, the computed liquid fraction, or a combination thereof. When the neural network module () detects such abnormality, it will trigger the alert module () to notify the user that a potential malfunction has occurred within the improved flow meter (100). In event of such malfunction, the second mass flow rate determined using the pressure drop can be used as a backup or redundancy measurement.

One skilled in the art will readily appreciate that the present invention is well adapted to carry out the objects and obtain the ends and advantages mentioned, as well as those inherent therein. The embodiment described herein is not intended as limitations on the scope of the present invention.