A method and system for numerical simulation of flow characteristics inside a regulating valve

Abstract

The application provides a kind of adjusting valve internal flow characteristic numerical simulation method and system, it is related to adjusting valve technical field, including: in adjusting valve multiple flow sensors and pressure sensors are arranged;Flow characteristic curve is obtained based on the actuator adjusting valve of positioner control is executed;Deep learning model is established, and deep learning model is trained, and the flow characteristics of adjusting valve are simulated using deep learning model, the application is used to solve the problem that the flow data of adjusting valve is not accurate and comprehensive enough in the process of analyzing the flow characteristics of adjusting valve in prior art, only by relative flow to obtain flow characteristics, it will lack the basis for correction when flow characteristics deviate, at the same time, the analysis method used after obtaining the flow data of adjusting valve in prior art is not intelligent enough, so that the flow characteristics obtained after analysis always remain unchanged, which cannot be consistent with the actual situation of adjusting valve.

Description

Technical Field

[0001] This invention relates to the field of control valve technology, and in particular to a numerical simulation method and system for the internal flow characteristics of a control valve. Background Technology

[0002] Control valves are used to regulate process parameters such as flow rate, pressure, temperature, and liquid level in industrial automation process control. Based on the control signals in the automation system, the valve opening is automatically adjusted to regulate the flow rate, pressure, temperature, and liquid level of the medium. The flow characteristics of a control valve refer to the relationship between the relative flow rate of the medium through the control valve and its opening degree under the condition that the pressure difference across the control valve remains constant.

[0003] Existing improvements to the flow characteristics of control valves typically involve simulating the flow characteristics of the valve under different operating conditions and then calculating the flow characteristics based on the simulated parameters. For example, Chinese patent CN113420514A discloses a numerical simulation method for the internal flow characteristics of a multi-stage pressure-reducing control valve. This method establishes an effective numerical calculation method based on the flow characteristics of the multi-stage pressure-reducing control valve, and the calculation results can reflect the changing laws of the internal flow of the valve, providing a theoretical basis and reference for the rational design of the pressure-reducing valve. However, the above methods and existing technologies have several drawbacks in analyzing the flow characteristics of control valves. First, the acquisition of flow data is not accurate or comprehensive enough. Obtaining flow characteristics only through relative flow rates lacks a basis for correction when deviations occur. Furthermore, the analysis methods used after acquiring the flow data are not intelligent enough, resulting in the obtained flow characteristics remaining unchanged and failing to match the actual situation of the control valve. Therefore, it is necessary to improve the existing methods for analyzing the flow characteristics of control valves. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to provide a numerical simulation method and system for the internal flow characteristics of control valves. This method solves the problems in existing technologies where the flow data of control valves is not accurate or comprehensive enough, and the flow characteristics are obtained only through relative flow. This results in a lack of basis for correction when deviations occur in the flow characteristics. Furthermore, the analysis methods used in existing technologies after obtaining the flow data of control valves are not intelligent enough, leading to the flow characteristics obtained after analysis remaining unchanged and failing to match the actual situation of the control valve.

[0005] To achieve the above objectives, this application provides a numerical simulation method for the internal flow characteristics of a control valve, comprising:

[0006] The flow characteristic data of the control valve is simulated using the rheological extraction method, and the flow characteristic curve is obtained by simulation based on the flow characteristic data;

[0007] A deep learning model is established and trained based on the flow characteristic curve. The trained deep learning model is then used to simulate the flow characteristics of the control valve.

[0008] Multiple flow sensors and pressure sensors are installed in the regulating valve;

[0009] The liquid flowing through the control valve is denoted as the flowing medium; the port through which the flowing medium enters the valve body is denoted as the inlet, and the port through which the flowing medium exits the valve body is denoted as the outlet.

[0010] Flow sensors are placed in the pipes corresponding to the inlet and outlet, and are referred to as the input sensor and the output sensor, respectively. The flow sensors are used to collect the velocity of the flowing medium.

[0011] A pressure sensor is placed in the valve core of the valve body. When the valve body contains multiple valve cores, the pressure sensor is placed in the valve core closest to the input sensor.

[0012] Furthermore, the flow characteristic data of the control valve are simulated using the rheological extraction method, and the flow characteristic curve of the control valve is obtained, including: control valve based on positioner control actuator;

[0013] The positioner controls the actuator to adjust the valve opening to 0 and allows the flowing medium to enter the valve body's inlet.

[0014] The valve opening is gradually increased, and the data collected from the input sensor, output sensor, and pressure sensor are obtained.

[0015] The flow characteristic curve of the control valve is obtained using rheological extraction based on the data collected from the input sensor, output sensor, and pressure sensor.

[0016] Furthermore, the rheological extraction method includes:

[0017] During the process of increasing the valve opening from 0 to 100%, the valve opening is stopped every unit percentage increase from 0, and the valve opening at this time is recorded as the intermittent opening. When the valve opening is in the intermittent opening and continues for the standard sampling time, the valve opening is increased again until the valve opening is increased to 100%.

[0018] All intermittent openings obtained during the process of increasing the valve opening are recorded in ascending order as intermittent opening 1 to intermittent opening T;

[0019] For any intermittent opening T1 from intermittent opening 1 to intermittent opening T, the data collected by the input sensor within the standard acquisition time corresponding to the valve opening at intermittent opening T1 is recorded as intermittent input data T1SR, the data collected by the output sensor within the standard acquisition time corresponding to the valve opening at intermittent opening T1 is recorded as intermittent output data T1SC, and the data collected by the pressure sensor within the standard acquisition time corresponding to the valve opening at intermittent opening T1 is recorded as intermittent pressure data T1YL, where T1 is a positive integer greater than or equal to 1 and less than or equal to T.

[0020] Furthermore, the rheological extraction method also includes:

[0021] For intermittent input data T1SR: Establish a Cartesian coordinate system, denoted as the input analysis coordinate system, where the X-axis of the input analysis coordinate system is in units of time, and the Y-axis is in units of flow velocity; Based on the flow velocity of the circulating medium collected within the standard acquisition time in the intermittent input data T1SR, plot the corresponding curve in the input analysis coordinate system, denoted as the intermittent input curve T1RQ;

[0022] The leftmost and rightmost points of the intermittent input curve T1RQ are denoted as points L to T1RQ and R to T1RQ, respectively. The standard acquisition duration is the X-axis coordinate of point R to T1RQ minus the X-axis coordinate of point L to T1RQ. The line segment obtained by connecting points L to T1RQ and R to T1RQ is denoted as the intermittent input line segment T1RD. The Y-axis coordinate corresponding to the midpoint of the intermittent input line segment T1RD is denoted as the intermittent input velocity T1RS.

[0023] Obtain the intermittent input velocity corresponding to the intermittent input data of all intermittent openings, and record them as intermittent input velocity 1RS to intermittent input velocity TRS respectively.

[0024] Furthermore, the rheological extraction method also includes:

[0025] For intermittent output data T1SC: Establish a Cartesian coordinate system, denoted as the output analysis coordinate system, where the unit of the X-axis of the output analysis coordinate system is time, and the unit of the Y-axis of the output analysis coordinate system is flow velocity; Based on the flow velocity of the circulating medium collected within the standard acquisition time in the intermittent output data T1SC, plot the corresponding curve in the output analysis coordinate system, denoted as the intermittent output curve T1CQ;

[0026] The leftmost and rightmost points of the intermittent output curve T1CQ are denoted as L~T1CQ and R~T1CQ, respectively. The standard acquisition duration is the X-axis coordinate of point R~T1CQ minus the X-axis coordinate of point L~T1CQ. The line segment obtained by connecting points L~T1CQ and R~T1CQ is denoted as the intermittent output line segment T1CD. The Y-axis coordinate corresponding to the midpoint of the intermittent output line segment T1CD is denoted as the intermittent output velocity T1CS.

[0027] Obtain the intermittent output velocity corresponding to the intermittent output data of all intermittent openings, and record them as intermittent output velocity 1CS to intermittent output velocity TCS respectively.

[0028] Furthermore, the rheological extraction method also includes:

[0029] For intermittent pressure data T1YL: the maximum pressure received by the pressure sensor in the intermittent pressure data T1YL is recorded as the intermittent pressure peak value T1FZ;

[0030] Obtain the peak intermittent pressure corresponding to all intermittent pressure data, and record them as intermittent pressure peak value 1FZ to intermittent pressure peak value TFZ respectively.

[0031] Furthermore, the rheological extraction method also includes:

[0032] Establish a Cartesian coordinate system, denoted as the flow analysis coordinate system, where the unit of the X-axis of the flow analysis coordinate system is the valve opening degree, and the unit of the Y-axis is the flow velocity;

[0033] Based on the intermittent input velocity corresponding to each intermittent opening, point markings are made in the flow analysis coordinate system, and the marked points are sequentially recorded as input point 1SRD to input point TSRD based on the abscissa from smallest to largest; based on the intermittent output velocity corresponding to each intermittent opening, point markings are made in the flow analysis coordinate system, and the marked points are sequentially recorded as output point 1SCD to output point TSCD based on the abscissa from smallest to largest.

[0034] For any input point T1SRD from input point 1SRD to input point TSRD, the average of the ordinate of input point T1SRD and the ordinate of output point T1SCD is denoted as input-output average T1, and the point corresponding to (intermittent opening T1, input-output average T1) is denoted as input-output average point T1JD, where intermittent opening T1 is the abscissa of input point T1SRD and output point T1SCD.

[0035] Obtain the average inflow and outflow points corresponding to all input points, and denote the curve formed by all the average inflow and outflow points as the inflow and outflow flow curve.

[0036] Furthermore, the rheological extraction method also includes:

[0037] After obtaining the inlet and outlet flow curves, the valve opening is readjusted to 0, and the inlet and outlet flow curves are obtained multiple times while keeping the flow medium unchanged, based on the method of obtaining the inlet and outlet flow curves. All inlet and outlet flow curves are recorded as inlet and outlet flow curve 1 to inlet and outlet flow curve C respectively.

[0038] For any inflow / outflow curve C1 from inflow / outflow curve 2 to inflow / outflow curve C-1, and for any average inflow / outflow point α within inflow / outflow curve C1, the average inflow / outflow point α is processed using the mean judgment formula, and the ordinate of the average inflow / outflow point α is reassigned. The mean judgment formula is as follows: Where, α i Let α be the ordinate of the average point α. i-1 Let α be the ordinate of the point whose x-coordinate is equal to the average inlet and outlet point α in the inlet and outlet flow curve C1-1. Based on the processing method of the average inlet and outlet point α, process all the average inlet and outlet points in the inlet and outlet flow curve C1. Then, based on the processing method of the inlet and outlet flow curve C1, process the inlet and outlet flow curve C1+1. Continue in this way until the inlet and outlet flow curve C is processed. The resulting inlet and outlet flow curve C is recorded as the flow characteristic curve.

[0039] Furthermore, a deep learning model is established and trained based on the flow characteristic curve. The trained deep learning model is then used to simulate the flow characteristics of the control valve, including:

[0040] Build deep learning models;

[0041] Deep learning models are trained based on traffic characteristic curves;

[0042] The flow of the medium in the control valve is simulated using a trained deep learning model, and the flow characteristic curve is obtained using a rheological extraction method. The obtained flow characteristic curve is recorded as the flow characteristic of the control valve.

[0043] Furthermore, training the deep learning model based on the traffic characteristic curve includes:

[0044] The rheological extraction method and the flow characteristic curve are input into the deep learning model, and the flow characteristic curve is re-obtained based on the deep learning model using the rheological extraction method, which is denoted as the new simulated curve.

[0045] The novel simulation curve and the flow characteristic curve are placed in the same flow analysis coordinate system. For any intermittent opening T1 from intermittent opening 1 to intermittent opening T in the flow analysis coordinate system, when the ordinate of the novel simulation curve with the intermittent opening T1 is not equal to the ordinate of the flow characteristic curve with the intermittent opening T1, the intermittent opening T1 is recorded as the calibration opening. Based on the calibration opening, the calibration method is used to adjust the ordinate of the flow characteristic curve with the calibrated opening.

[0046] The calibration method includes: recording the intermittent pressure peak value of the intermittent pressure data corresponding to the calibration opening in the rheological extraction method corresponding to the obtained flow characteristic curve as the standard calibration peak value;

[0047] The peak value of the intermittent pressure corresponding to the calibration opening in the rheological extraction method corresponding to the new simulation curve is recorded as the new simulation peak value.

[0048] When the standard calibration peak value equals the new simulated peak value, the ordinate of the flow characteristic curve with the horizontal axis corresponding to the standard opening degree is replaced with the ordinate of the new simulated curve with the horizontal axis corresponding to the standard opening degree.

[0049] When the standard calibration peak value is not equal to the new simulated peak value, the ordinate corresponding to the standard opening degree in the flow characteristic curve is not replaced;

[0050] All calibrated openings from intermittent opening 1 to intermittent opening T in the flow analysis coordinate system are obtained, and the calibration method is used to process all calibrated openings. Based on the flow characteristic curve obtained after processing and the new simulation curve obtained by rheological extraction method using the deep learning model, the calibration method is used to adjust the vertical axis in the flow characteristic curve.

[0051] When the flow characteristic curve processed by the calibration method and the new simulated curve re-obtained by the deep learning model using the rheological extraction method are placed into the same flow analysis coordinate system, and all intermittent openings from intermittent opening 1 to intermittent opening T in the flow analysis coordinate system are not calibrated openings, the deep learning model is recorded as a deep learning model that has been trained.

[0052] The beneficial effects of this invention are as follows: First, multiple flow sensors and pressure sensors are installed in the regulating valve. Then, the valve is regulated by an actuator controlled by a positioner, and a flow characteristic curve is obtained based on valve changes and the data collected by the flow and pressure sensors. Finally, a deep learning model is established and trained based on the flow characteristic curve. The trained deep learning model is then used to simulate the flow characteristics of the regulating valve. The advantage of this is that by using flow and pressure sensors for data acquisition, data collected by the pressure sensor related to the flow characteristics can be obtained simultaneously with the flow characteristics acquired by the flow sensor. This helps to correct the flow characteristics based on the data collected by the pressure sensor in subsequent adjustments. Furthermore, by acquiring the flow characteristic curve and learning from it using the deep learning model, the learned deep learning model can simulate flow characteristics that better match the actual situation of the regulating valve, thereby assisting operators in performing more accurate work.

[0053] Advantages of additional aspects of the invention will be set forth in part in the detailed description of the invention below, and in part will be obvious from the description or may be learned by practice of the invention. Attached Figure Description

[0054] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0055] Figure 1 This is a schematic diagram of a numerical simulation system for the internal flow characteristics of a control valve according to the present invention.

[0056] Figure 2 This is a flowchart illustrating the steps of a numerical simulation method for the internal flow characteristics of a control valve according to the present invention.

[0057] Figure 3 This is a schematic diagram illustrating the acquisition of intermittent input flow rate according to the present invention;

[0058] Figure 4 This is a schematic diagram illustrating the acquisition of the inflow and outflow curves according to the present invention. Detailed Implementation

[0059] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0060] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0061] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0062] Example 1, First Aspect, Please refer to Figure 1 As shown, the present invention provides a numerical simulation system for the internal flow characteristics of a control valve, including a sensor deployment module, a characteristic curve acquisition module, and a model training module;

[0063] The sensor deployment module is used to install multiple flow sensors and pressure sensors in the control valve;

[0064] The sensor deployment module is configured with a sensor deployment strategy, which includes:

[0065] The liquid flowing through the control valve is denoted as the flowing medium; the port through which the flowing medium enters the valve body is denoted as the inlet, and the port through which the flowing medium exits the valve body is denoted as the outlet.

[0066] Flow sensors are placed in the pipes corresponding to the inlet and outlet, and are referred to as the input sensor and the output sensor, respectively. The flow sensors are used to collect the velocity of the flowing medium.

[0067] A pressure sensor is placed in the valve core of the valve body. When the valve body contains multiple valve cores, the pressure sensor is placed in the valve core closest to the input sensor.

[0068] In the specific implementation process, when the regulating valve can flow in both directions, the port corresponding to the valve core where the pressure sensor is placed is the input port, ensuring that the pressure sensor can receive the pressure of the flowing medium acting on the valve core when the flowing medium enters the regulating valve.

[0069] The control valve includes a positioner and an actuator. The positioner receives the stroke control signal that the valve needs to open or close, and then controls the actuator to output the corresponding stroke. The characteristic curve acquisition module is used to control the actuator to regulate the valve based on the positioner, and to acquire the flow characteristic curve based on the valve changes and the acquisition results of the flow sensor and pressure sensor.

[0070] The feature curve acquisition module is configured with a feature curve acquisition strategy, which includes:

[0071] The positioner controls the actuator to adjust the valve opening to 0 and allows the flowing medium to enter the valve body's inlet.

[0072] The valve opening is gradually increased, and the data collected from the input sensor, output sensor, and pressure sensor are obtained.

[0073] The flow characteristic curve of the control valve is obtained using rheological extraction based on the data collected from the input sensor, output sensor, and pressure sensor.

[0074] Rheological extraction methods include:

[0075] During the process of increasing the valve opening from 0 to 100%, the valve opening is stopped every unit percentage increase from 0, and the valve opening at this time is recorded as the intermittent opening. When the valve opening is in the intermittent opening and continues for the standard sampling time, the valve opening is increased again until the valve opening is increased to 100%.

[0076] In the specific implementation process, the unit percentage and standard acquisition time can be adjusted according to the actual analysis needs. In this embodiment, the unit percentage is set to 1% and the standard acquisition time is set to 5s. That is, when the valve opening starts from 0, for every 1% increase, the valve opening is stopped for 5s. After 5s, the valve opening is increased again until the valve opening is increased to 100%. And the number of intermittent openings can be calculated to be 100.

[0077] All intermittent openings obtained during the process of increasing the valve opening are recorded in ascending order as intermittent opening 1 to intermittent opening T;

[0078] For any intermittent opening T1 from intermittent opening 1 to intermittent opening T, the data collected by the input sensor within the standard acquisition time corresponding to the valve opening at intermittent opening T1 is recorded as intermittent input data T1SR, the data collected by the output sensor within the standard acquisition time corresponding to the valve opening at intermittent opening T1 is recorded as intermittent output data T1SC, and the data collected by the pressure sensor within the standard acquisition time corresponding to the valve opening at intermittent opening T1 is recorded as intermittent pressure data T1YL, where T1 is a positive integer greater than or equal to 1 and less than or equal to T.

[0079] Rheological extraction methods also include:

[0080] For intermittent input data T1SR: Establish a Cartesian coordinate system, denoted as the input analysis coordinate system, where the X-axis of the input analysis coordinate system is in units of time, and the Y-axis is in units of flow velocity; Based on the flow velocity of the circulating medium collected within the standard acquisition time in the intermittent input data T1SR, plot the corresponding curve in the input analysis coordinate system, denoted as the intermittent input curve T1RQ;

[0081] The leftmost and rightmost points of the intermittent input curve T1RQ are denoted as points L to T1RQ and R to T1RQ, respectively. The standard acquisition duration is the X-axis coordinate of point R to T1RQ minus the X-axis coordinate of point L to T1RQ. The line segment obtained by connecting points L to T1RQ and R to T1RQ is denoted as the intermittent input line segment T1RD. The Y-axis coordinate corresponding to the midpoint of the intermittent input line segment T1RD is denoted as the intermittent input velocity T1RS.

[0082] For specific implementation details, please refer to [link / reference]. Figure 3 As shown, TT1 is the intermittent input curve, the time occupied by the horizontal axis length corresponding to TT2 is the standard acquisition duration, TT3 is the intermittent input line segment, TT4 is the midpoint of the intermittent input line segment, and TT5 is the intermittent input flow rate.

[0083] Obtain the intermittent input velocity corresponding to the intermittent input data of all intermittent openings, and record them as intermittent input velocity 1RS to intermittent input velocity TRS respectively.

[0084] Rheological extraction methods also include:

[0085] For intermittent output data T1SC: Establish a Cartesian coordinate system, denoted as the output analysis coordinate system, where the unit of the X-axis of the output analysis coordinate system is time, and the unit of the Y-axis of the output analysis coordinate system is flow velocity; Based on the flow velocity of the circulating medium collected within the standard acquisition time in the intermittent output data T1SC, plot the corresponding curve in the output analysis coordinate system, denoted as the intermittent output curve T1CQ;

[0086] The leftmost and rightmost points of the intermittent output curve T1CQ are denoted as L~T1CQ and R~T1CQ, respectively. The standard acquisition duration is the X-axis coordinate of point R~T1CQ minus the X-axis coordinate of point L~T1CQ. The line segment obtained by connecting points L~T1CQ and R~T1CQ is denoted as the intermittent output line segment T1CD. The Y-axis coordinate corresponding to the midpoint of the intermittent output line segment T1CD is denoted as the intermittent output velocity T1CS.

[0087] Obtain the intermittent output velocity corresponding to the intermittent output data of all intermittent openings, and record them as intermittent output velocity 1CS to intermittent output velocity TCS respectively;

[0088] In summary, the method for obtaining the intermittent output flow rate is the same as the method for obtaining the intermittent input flow rate in the specific implementation process.

[0089] Rheological extraction methods also include:

[0090] For intermittent pressure data T1YL: the maximum pressure received by the pressure sensor in the intermittent pressure data T1YL is recorded as the intermittent pressure peak value T1FZ;

[0091] In the specific implementation process, by obtaining the intermittent pressure peak value, pressure-related data corresponding to each intermittent opening degree can be obtained, thus providing data support when correcting the flow characteristic curve in the future;

[0092] Obtain the peak intermittent pressure corresponding to all intermittent pressure data, and record them as intermittent pressure peak value 1FZ to intermittent pressure peak value TFZ respectively.

[0093] Rheological extraction methods also include:

[0094] Establish a Cartesian coordinate system, denoted as the flow analysis coordinate system, where the unit of the X-axis of the flow analysis coordinate system is the valve opening degree, and the unit of the Y-axis is the flow velocity;

[0095] Based on the intermittent input velocity corresponding to each intermittent opening, point markings are made in the flow analysis coordinate system, and the marked points are sequentially recorded as input point 1SRD to input point TSRD based on the abscissa from smallest to largest; based on the intermittent output velocity corresponding to each intermittent opening, point markings are made in the flow analysis coordinate system, and the marked points are sequentially recorded as output point 1SCD to output point TSCD based on the abscissa from smallest to largest.

[0096] For any input point T1SRD from input point 1SRD to input point TSRD, the average of the ordinate of input point T1SRD and the ordinate of output point T1SCD is denoted as input-output average T1, and the point corresponding to (intermittent opening T1, input-output average T1) is denoted as input-output average point T1JD, where intermittent opening T1 is the abscissa of input point T1SRD and output point T1SCD.

[0097] In the specific implementation process, for example, in a data processing process, the intermittent opening degree 5 is 5%, the ordinate of the input point 5SRD is 3m / s, the ordinate of the output point 5SCD is 2m / s, then the average input and output value 5 is 2.5m / s. The point corresponding to (5, 2.5) in the flow analysis coordinate system is recorded as the average input and output point 5JD.

[0098] Obtain the average inflow and outflow points corresponding to all input points, and record the curve formed by all the average inflow and outflow points as the inflow and outflow flow curve;

[0099] For details, please refer to Figure 4 As shown, FF1 to FF8 represent intermittent opening degrees 1 to 8, △ represents the input point, □ represents the output point, and the curve corresponding to FF9 represents the inflow and outflow curves.

[0100] After obtaining the inlet and outlet flow curves, the valve opening is readjusted to 0. Based on the method of obtaining the inlet and outlet flow curves, the inlet and outlet flow curves are obtained multiple times while keeping the flow medium unchanged. All inlet and outlet flow curves are recorded as inlet and outlet flow curve 1 to inlet and outlet flow curve C respectively.

[0101] For any inflow / outflow curve C1 from inflow / outflow curve 2 to inflow / outflow curve C-1, and for any average inflow / outflow point α within inflow / outflow curve C1, the average inflow / outflow point α is processed using the mean judgment formula, and the ordinate of the average inflow / outflow point α is reassigned. The mean judgment formula is as follows: Where, α i Let α be the ordinate of the average point α. i-1Let α be the ordinate of the point whose x-coordinate is equal to the average inlet and outlet point α in the inlet and outlet flow curve C1-1. Based on the processing method of the average inlet and outlet point α, process all the average inlet and outlet points in the inlet and outlet flow curve C1. Then, based on the processing method of the inlet and outlet flow curve C1, process the inlet and outlet flow curve C1+1. Continue in this way until the inlet and outlet flow curve C is processed. The resulting inlet and outlet flow curve C is recorded as the flow characteristic curve.

[0102] In the specific implementation process, for example, in a data processing, the coordinates of an average point α in the inflow and outflow curve C1 are (60°, 2m / s), and the ordinate of the point with the horizontal coordinate of 60° in the inflow and outflow curve C1-1 is 2.1m / s. After processing by the mean value judgment formula, the ordinate of the average point α is reassigned to 2.05m / s.

[0103] The model training module is used to build a deep learning model and train the deep learning model based on the flow characteristic curve. The trained deep learning model is then used to simulate the flow characteristics of the control valve.

[0104] The model training module is configured with model building and optimization strategies, which include:

[0105] Build deep learning models;

[0106] Deep learning models are trained based on traffic characteristic curves;

[0107] The rheological extraction method and the flow characteristic curve are input into the deep learning model, and the flow characteristic curve is re-obtained based on the deep learning model using the rheological extraction method, which is denoted as the new simulated curve.

[0108] In the specific implementation process, whenever the deep learning model uses the rheological extraction method to re-acquire the flow characteristic curve, the valve opening of the regulating valve is readjusted to 0 and the flow medium is re-introduced. The flow characteristic curve is re-acquired based on the acquisition results of the pressure sensor and flow sensor during the process of increasing the valve opening.

[0109] The novel simulation curve and the flow characteristic curve are placed in the same flow analysis coordinate system. For any intermittent opening T1 from intermittent opening 1 to intermittent opening T in the flow analysis coordinate system, when the ordinate of the novel simulation curve with the intermittent opening T1 is not equal to the ordinate of the flow characteristic curve with the intermittent opening T1, the intermittent opening T1 is recorded as the calibration opening. Based on the calibration opening, the calibration method is used to adjust the ordinate of the flow characteristic curve with the calibrated opening.

[0110] The calibration method includes: recording the intermittent pressure peak value of the intermittent pressure data corresponding to the calibration opening in the rheological extraction method corresponding to the obtained flow characteristic curve as the standard calibration peak value;

[0111] The peak value of the intermittent pressure corresponding to the calibration opening in the rheological extraction method corresponding to the new simulation curve is recorded as the new simulation peak value.

[0112] When the standard calibration peak value equals the new simulated peak value, the ordinate of the flow characteristic curve with the horizontal axis corresponding to the standard opening degree is replaced with the ordinate of the new simulated curve with the horizontal axis corresponding to the standard opening degree.

[0113] In the specific implementation process, when the standard calibration peak value is equal to the new simulated peak value, it means that under the same intermittent pressure peak value, the flow velocity represented by the vertical axis corresponding to the standard opening on the horizontal axis of the flow characteristic curve may differ from the latest obtained flow velocity due to changes in the external environment and other reasons. Therefore, the latest obtained data should be used as the benchmark, that is, replaced by the vertical axis corresponding to the standard opening on the horizontal axis of the new simulated curve.

[0114] When the standard calibration peak value is not equal to the new simulated peak value, the ordinate corresponding to the standard opening degree in the flow characteristic curve is not replaced;

[0115] In specific implementation, for example, in a data processing, the calibration opening is set to an intermittent opening of 5. When using the rheological extraction method to obtain the flow characteristic curve, the intermittent pressure peak value of the intermittent pressure data corresponding to the intermittent opening of 5 is 10N. When using the rheological extraction method to obtain the new simulation curve, the intermittent pressure peak value of the intermittent pressure data corresponding to the intermittent opening of 5 is 8N. Therefore, the horizontal axis of the flow characteristic curve is not replaced with the vertical axis corresponding to the standard opening.

[0116] Model building and optimization strategies also include:

[0117] All calibrated openings from intermittent opening 1 to intermittent opening T in the flow analysis coordinate system are obtained, and the calibration method is used to process all calibrated openings. Based on the flow characteristic curve obtained after processing and the new simulation curve obtained by rheological extraction method using the deep learning model, the calibration method is used to adjust the vertical axis in the flow characteristic curve.

[0118] When the flow characteristic curve processed by the calibration method and the new simulated curve re-obtained by the deep learning model using the rheological extraction method are placed into the same flow analysis coordinate system, and all intermittent openings from intermittent opening 1 to intermittent opening T in the flow analysis coordinate system are not calibrated openings, the deep learning model is recorded as a deep learning model that has been trained.

[0119] The flow of the medium in the control valve is simulated using a trained deep learning model, and the flow characteristic curve is obtained using a rheological extraction method. The obtained flow characteristic curve is recorded as the flow characteristic of the control valve.

[0120] In practical implementation, since the flow characteristic of a control valve is defined as the relationship between the relative flow rate of the medium flowing through the control valve and its opening degree, and the unit of the Y-axis of the coordinate system where the flow characteristic curve is located is flow velocity, in actual processing, when the flow characteristic curve is recorded as a flow characteristic, the ordinate of the flow characteristic curve can be converted from flow velocity to relative flow rate before recording the flow characteristic curve as a flow characteristic. For example, the ordinate of a point A on the flow characteristic curve is 1 m / s and the slope is 2, while the relative flow rate corresponding to the medium flowing through the control valve with a velocity of 1 m / s and an acceleration of 2 m / s² is 1 m. 3 Then the ordinate of point A will be changed to 1m. 3 Based on the processing method for point A, all points in the flow characteristic curve are processed accordingly, and the processed flow characteristic curve is recorded as the flow characteristic. At this time, the vertical axis unit of the coordinate system where the flow characteristic is located becomes relative flow.

[0121] Example 2, Second Aspect, Please refer to Figure 2 As shown, the present invention also provides a numerical simulation method for the internal flow characteristics of a control valve, comprising:

[0122] Step S1: Install multiple flow sensors and pressure sensors in the regulating valve.

[0123] Step S1 includes: Step S101, the liquid flowing in the regulating valve is denoted as the flowing medium; the port in the valve body where the flowing medium enters is denoted as the input port; and the port in the valve body where the flowing medium exits is denoted as the output port.

[0124] Step S102: Place flow sensors in the pipes corresponding to the inlet and outlet, respectively denoted as input sensor and output sensor. The flow sensors are used to collect the velocity of the flowing medium.

[0125] Step S103: Place a pressure sensor in the valve core of the valve body. When the valve body contains multiple valve cores, place the pressure sensor in the valve core closest to the input sensor.

[0126] Step S2: Adjust the valve based on the positioner control actuator, and obtain the flow characteristic curve based on the valve changes and the acquisition results of the flow sensor and pressure sensor.

[0127] Step S2 includes: Step S201, using a positioner to control the actuator to adjust the valve opening to 0, and introducing the flow medium into the valve body's inlet;

[0128] Step S202: Gradually increase the valve opening and acquire the data from the input sensor, output sensor, and pressure sensor.

[0129] Step S203: Based on the acquisition results of the input sensor, output sensor and pressure sensor, the flow characteristic curve of the control valve is obtained using the rheological extraction method.

[0130] The rheological extraction method includes: step S2031, during the process of increasing the valve opening from 0 to 100%, the valve opening is increased by a unit percentage from 0, and the valve opening at this time is stopped and recorded as the intermittent opening. When the valve opening is in the intermittent opening and continues for the standard acquisition time, the valve opening is increased again until the valve opening is increased to 100%.

[0131] All intermittent openings obtained during the process of increasing the valve opening are recorded in ascending order as intermittent opening 1 to intermittent opening T;

[0132] Step S2032: For any intermittent opening T1 from intermittent opening 1 to intermittent opening T, the data collected by the input sensor within the standard acquisition time corresponding to the valve opening at intermittent opening T1 is recorded as intermittent input data T1SR, the data collected by the output sensor within the standard acquisition time corresponding to the valve opening at intermittent opening T1 is recorded as intermittent output data T1SC, and the data collected by the pressure sensor within the standard acquisition time corresponding to the valve opening at intermittent opening T1 is recorded as intermittent pressure data T1YL, where T1 is a positive integer greater than or equal to 1 and less than or equal to T.

[0133] Step S2033, for intermittent input data T1SR: Establish a Cartesian coordinate system, denoted as the input analysis coordinate system, where the unit of the X-axis of the input analysis coordinate system is time, and the unit of the Y-axis of the input analysis coordinate system is flow velocity; Based on the flow velocity of the circulating medium collected within the standard acquisition time in the intermittent input data T1SR, plot the corresponding curve in the input analysis coordinate system, denoted as the intermittent input curve T1RQ;

[0134] Step S2034: The leftmost and rightmost points of the intermittent input curve T1RQ are denoted as points L~T1RQ and R~T1RQ, respectively, and the X-axis coordinates corresponding to points R~T1RQ minus the X-axis coordinates corresponding to points L~T1RQ are the standard acquisition duration; the line segment obtained by connecting points L~T1RQ and R~T1RQ is denoted as the intermittent input line segment T1RD, and the Y-axis coordinates corresponding to the midpoint of the intermittent input line segment T1RD are denoted as the intermittent input velocity T1RS;

[0135] Obtain the intermittent input velocity corresponding to the intermittent input data of all intermittent openings, and record them as intermittent input velocity 1RS to intermittent input velocity TRS respectively.

[0136] Step S2035, for intermittent output data T1SC: establish a Cartesian coordinate system, denoted as the output analysis coordinate system, where the unit of the X-axis of the output analysis coordinate system is time, and the unit of the Y-axis of the output analysis coordinate system is flow velocity; based on the flow velocity of the circulating medium collected within the standard acquisition time in the intermittent output data T1SC, plot the corresponding curve in the output analysis coordinate system, denoted as the intermittent output curve T1CQ;

[0137] The leftmost and rightmost points of the intermittent output curve T1CQ are denoted as L~T1CQ and R~T1CQ, respectively. The standard acquisition duration is the X-axis coordinate of point R~T1CQ minus the X-axis coordinate of point L~T1CQ. The line segment obtained by connecting points L~T1CQ and R~T1CQ is denoted as the intermittent output line segment T1CD. The Y-axis coordinate corresponding to the midpoint of the intermittent output line segment T1CD is denoted as the intermittent output velocity T1CS.

[0138] Step S2036: Obtain the intermittent output velocity corresponding to the intermittent output data of all intermittent openings, and record them as intermittent output velocity 1CS to intermittent output velocity TCS respectively.

[0139] Step S2037, for intermittent pressure data T1YL: record the maximum pressure received by the pressure sensor in the intermittent pressure data T1YL as the intermittent pressure peak value T1FZ;

[0140] Obtain the peak intermittent pressure corresponding to all intermittent pressure data, and record them as intermittent pressure peak value 1FZ to intermittent pressure peak value TFZ respectively.

[0141] Step S2038: Establish a plane rectangular coordinate system, denoted as the flow analysis coordinate system, where the unit of the X-axis of the flow analysis coordinate system is the valve opening degree, and the unit of the Y-axis is the flow velocity;

[0142] Based on the intermittent input velocity corresponding to each intermittent opening, point markings are made in the flow analysis coordinate system, and the marked points are sequentially recorded as input point 1SRD to input point TSRD based on the abscissa from smallest to largest; based on the intermittent output velocity corresponding to each intermittent opening, point markings are made in the flow analysis coordinate system, and the marked points are sequentially recorded as output point 1SCD to output point TSCD based on the abscissa from smallest to largest.

[0143] Step S2039: For any input point T1SRD from input point 1SRD to input point TSRD, the average of the ordinate of input point T1SRD and the ordinate of output point T1SCD is recorded as the input-output average value T1, and the point corresponding to (intermittent opening T1, input-output average value T1) is recorded as the input-output average point T1JD, where the intermittent opening T1 is the abscissa of input point T1SRD and output point T1SCD;

[0144] Obtain the average inflow and outflow points corresponding to all input points, and denote the curve formed by all the average inflow and outflow points as the inflow and outflow flow curve.

[0145] Step S20391: After obtaining the inlet and outlet flow curves, readjust the valve opening to 0, and obtain the inlet and outlet flow curves multiple times while keeping the flow medium unchanged, based on the method of obtaining the inlet and outlet flow curves, and record all the inlet and outlet flow curves as inlet and outlet flow curve 1 to inlet and outlet flow curve C respectively.

[0146] For any inflow / outflow curve C1 from inflow / outflow curve 2 to inflow / outflow curve C-1, and for any average inflow / outflow point α within inflow / outflow curve C1, the average inflow / outflow point α is processed using the mean judgment formula, and the ordinate of the average inflow / outflow point α is reassigned. The mean judgment formula is as follows: Where, α i Let α be the ordinate of the average point α. i-1 Let α be the ordinate of the point whose x-coordinate is equal to the average inlet and outlet point α in the inlet and outlet flow curve C1-1. Based on the processing method of the average inlet and outlet point α, process all the average inlet and outlet points in the inlet and outlet flow curve C1. Then, based on the processing method of the inlet and outlet flow curve C1, process the inlet and outlet flow curve C1+1. Continue in this way until the inlet and outlet flow curve C is processed. The resulting inlet and outlet flow curve C is recorded as the flow characteristic curve.

[0147] Step S3: Establish a deep learning model and train the deep learning model based on the flow characteristic curve. Use the trained deep learning model to simulate the flow characteristics of the control valve.

[0148] Step S3 includes: Step S301, establishing a deep learning model;

[0149] Step S302: Train the deep learning model based on the traffic characteristic curve;

[0150] Step S302 includes: Step S3021, inputting the rheological extraction method and the flow characteristic curve into the deep learning model, and using the rheological extraction method to re-obtain the flow characteristic curve based on the deep learning model, which is denoted as the new simulated curve;

[0151] Step S3022: Place the new simulation curve and the flow characteristic curve into the same flow analysis coordinate system. For any intermittent opening T1 from intermittent opening 1 to intermittent opening T in the flow analysis coordinate system, when the ordinate of the new simulation curve with the intermittent opening T1 is not equal to the ordinate of the flow characteristic curve with the intermittent opening T1, the intermittent opening T1 is recorded as the calibration opening. Based on the calibration opening, the calibration method is used to adjust the ordinate of the flow characteristic curve with the calibration opening as the ordinate.

[0152] The calibration method includes: recording the intermittent pressure peak value of the intermittent pressure data corresponding to the calibration opening in the rheological extraction method corresponding to the obtained flow characteristic curve as the standard calibration peak value;

[0153] The peak value of the intermittent pressure corresponding to the calibration opening in the rheological extraction method corresponding to the new simulation curve is recorded as the new simulation peak value.

[0154] When the standard calibration peak value equals the new simulated peak value, the ordinate of the flow characteristic curve with the horizontal axis corresponding to the standard opening degree is replaced with the ordinate of the new simulated curve with the horizontal axis corresponding to the standard opening degree.

[0155] When the standard calibration peak value is not equal to the new simulated peak value, the horizontal axis of the flow characteristic curve is not replaced with the vertical axis corresponding to the standard opening degree.

[0156] Step S3023: Obtain all calibrated openings in the flow analysis coordinate system with the horizontal axis as intermittent opening 1 to intermittent opening T, and process them using the calibration method based on all calibrated openings. Based on the flow characteristic curve obtained after processing and the new simulation curve re-obtained by the rheological extraction method using the deep learning model, continue to use the calibration method to adjust the vertical axis in the flow characteristic curve.

[0157] Step S3024: When the flow characteristic curve processed by the calibration method and the new simulation curve re-obtained by the deep learning model using the rheological extraction method are put into the same flow analysis coordinate system, and all the intermittent openings from intermittent opening 1 to intermittent opening T in the flow analysis coordinate system are not calibration openings, the deep learning model is recorded as a deep learning model that has been trained.

[0158] Step S303: Use the trained deep learning model to simulate the flow of the medium in the control valve, and use the rheological extraction method to obtain the flow characteristic curve. The obtained flow characteristic curve is recorded as the flow characteristic of the control valve.

[0159] Working principle: First, multiple flow sensors and pressure sensors are set in the control valve; then, the valve is adjusted by the actuator controlled by the positioner, and the flow characteristic curve is obtained based on the valve changes and the acquisition results of the flow and pressure sensors; finally, a deep learning model is established, and the deep learning model is trained based on the flow characteristic curve, and the trained deep learning model is used to simulate the flow characteristics of the control valve.

[0160] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0161] The above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and are not intended to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the scope of the technology disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of protection of the claims.

Claims

1. A numerical simulation method for the internal flow characteristics of a control valve, characterized in that, include: The flow characteristic data of the control valve is simulated using the rheological extraction method, and the flow characteristic curve is obtained by simulation based on the flow characteristic data; A deep learning model is established and trained based on the flow characteristic curve. The trained deep learning model is then used to simulate the flow characteristics of the control valve. Multiple flow sensors and pressure sensors are installed in the regulating valve; The liquid flowing through the control valve is denoted as the flowing medium; the port through which the flowing medium enters the valve body is denoted as the inlet, and the port through which the flowing medium exits the valve body is denoted as the outlet. Flow sensors are placed in the pipes corresponding to the inlet and outlet, and are referred to as the input sensor and the output sensor, respectively. The flow sensors are used to collect the velocity of the flowing medium. A pressure sensor is placed in the valve core of the valve body. When the valve body contains multiple valve cores, the pressure sensor is placed in the valve core closest to the input sensor. The flow characteristic data of the control valve are simulated using rheological extraction method, and the flow characteristic curve of the control valve is obtained, including: control valve based on positioner control actuator; The positioner controls the actuator to adjust the valve opening to 0 and allows the flowing medium to enter the valve body's inlet. The valve opening is gradually increased, and the data collected from the input sensor, output sensor, and pressure sensor are obtained. The flow characteristic curve of the control valve is obtained using the rheological extraction method based on the acquisition results of the input sensor, output sensor and pressure sensor; The rheological extraction method includes: During the process of increasing the valve opening from 0 to 100%, the valve opening is stopped for each unit percentage increase starting from 0, and the valve opening at this point is recorded as the intermittent opening. When the valve opening is in the intermittent opening and continues for the standard sampling time, the valve opening is increased again until the valve opening is increased to 100%. All intermittent openings obtained during the process of increasing the valve opening are recorded in ascending order as intermittent opening 1 to intermittent opening T; For any intermittent opening T1 from intermittent opening 1 to intermittent opening T, the data collected by the input sensor within the standard acquisition time corresponding to the valve opening at intermittent opening T1 is recorded as intermittent input data T1SR, the data collected by the output sensor within the standard acquisition time corresponding to the valve opening at intermittent opening T1 is recorded as intermittent output data T1SC, and the data collected by the pressure sensor within the standard acquisition time corresponding to the valve opening at intermittent opening T1 is recorded as intermittent pressure data T1YL, where T1 is a positive integer greater than or equal to 1 and less than or equal to T; For intermittent input data T1SR: Establish a Cartesian coordinate system, denoted as the input analysis coordinate system, where the X-axis of the input analysis coordinate system is in units of time, and the Y-axis is in units of flow velocity; Based on the flow velocity of the circulating medium collected within the standard acquisition time in the intermittent input data T1SR, plot the corresponding curve in the input analysis coordinate system, denoted as the intermittent input curve T1RQ; The leftmost and rightmost points of the intermittent input curve T1RQ are denoted as L~T1RQ and R~T1RQ, respectively, and the X-axis coordinate of point R~T1RQ minus the X-axis coordinate of point L~T1RQ is the standard acquisition duration; the line segment obtained by connecting points L~T1RQ and R~T1RQ is denoted as the intermittent input line segment T1RD, and the Y-axis coordinate corresponding to the midpoint of the intermittent input line segment T1RD is denoted as the intermittent input velocity T1RS; Obtain the intermittent input velocity corresponding to the intermittent input data of all intermittent openings, and record them as intermittent input velocity 1RS to intermittent input velocity TRS respectively; For intermittent output data T1SC: Establish a Cartesian coordinate system, denoted as the output analysis coordinate system, where the unit of the X-axis of the output analysis coordinate system is time, and the unit of the Y-axis of the output analysis coordinate system is flow velocity; Based on the flow velocity of the circulating medium collected within the standard acquisition time in the intermittent output data T1SC, plot the corresponding curve in the output analysis coordinate system, denoted as the intermittent output curve T1CQ; The leftmost and rightmost points of the intermittent output curve T1CQ are denoted as L~T1CQ and R~T1CQ, respectively, and the X-axis coordinate of point R~T1CQ minus the X-axis coordinate of point L~T1CQ is the standard acquisition duration; the line segment obtained by connecting points L~T1CQ and R~T1CQ is denoted as the intermittent output line segment T1CD, and the Y-axis coordinate corresponding to the midpoint of the intermittent output line segment T1CD is denoted as the intermittent output velocity T1CS; Obtain the intermittent output velocity corresponding to the intermittent output data of all intermittent openings, and record them as intermittent output velocity 1CS to intermittent output velocity TCS respectively; For intermittent pressure data T1YL: the maximum pressure received by the pressure sensor in the intermittent pressure data T1YL is recorded as the intermittent pressure peak value T1FZ; Obtain the peak intermittent pressure corresponding to all intermittent pressure data, and record them as intermittent pressure peak value 1FZ to intermittent pressure peak value TFZ respectively; Establish a Cartesian coordinate system, denoted as the flow analysis coordinate system, where the unit of the X-axis of the flow analysis coordinate system is the valve opening degree, and the unit of the Y-axis is the flow velocity; Based on the intermittent input velocity corresponding to each intermittent opening, point markings are made in the flow analysis coordinate system, and the marked points are sequentially recorded as input point 1SRD to input point TSRD based on the abscissa from smallest to largest; based on the intermittent output velocity corresponding to each intermittent opening, point markings are made in the flow analysis coordinate system, and the marked points are sequentially recorded as output point 1SCD to output point TSCD based on the abscissa from smallest to largest. For any input point T1SRD from input point 1SRD to input point TSRD, the average of the ordinate of input point T1SRD and the ordinate of output point T1SCD is denoted as input-output average value T1, and the point corresponding to (intermittent opening T1, input-output average value T1) is denoted as input-output average point T1JD, where intermittent opening T1 is the x-coordinate of input point T1SRD and output point T1SCD; Obtain the average inflow and outflow points corresponding to all input points, and record the curve formed by all the average inflow and outflow points as the inflow and outflow flow curve; After obtaining the inlet and outlet flow curves, the valve opening is readjusted to 0, and the inlet and outlet flow curves are obtained multiple times while keeping the flow medium unchanged, based on the method of obtaining the inlet and outlet flow curves. All inlet and outlet flow curves are recorded as inlet and outlet flow curve 1 to inlet and outlet flow curve C respectively. For any inflow / outflow curve C1 from inflow / outflow curve 2 to inflow / outflow curve C-1, and for any average inflow / outflow point α within inflow / outflow curve C1, the average inflow / outflow point α is processed using the mean judgment formula, and the ordinate of the average inflow / outflow point α is reassigned. The mean judgment formula is as follows: , where α i Let α be the ordinate of the average point α. i-1 Let α be the ordinate of the point whose x-coordinate is equal to the average inlet and outlet point α in the inlet and outlet flow curve C1-1. Based on the processing method of the average inlet and outlet point α, process all the average inlet and outlet points in the inlet and outlet flow curve C1. Then, based on the processing method of the inlet and outlet flow curve C1, process the inlet and outlet flow curve C1+1. Continue in this way until the inlet and outlet flow curve C is processed. The resulting inlet and outlet flow curve C is recorded as the flow characteristic curve.

2. The numerical simulation method for the internal flow characteristics of a control valve according to claim 1, characterized in that, A deep learning model is established and trained based on the flow characteristic curve. The trained deep learning model is then used to simulate the flow characteristics of the control valve, including: Build deep learning models; Deep learning models are trained based on traffic characteristic curves; The flow of the medium in the control valve is simulated using a trained deep learning model, and the flow characteristic curve is obtained using a rheological extraction method. The obtained flow characteristic curve is recorded as the flow characteristic of the control valve.

3. The numerical simulation method for the internal flow characteristics of a regulating valve according to claim 2, characterized in that, Training deep learning models based on traffic characteristic curves includes: The rheological extraction method and the flow characteristic curve are input into the deep learning model, and the flow characteristic curve is re-obtained based on the deep learning model using the rheological extraction method, which is denoted as the new simulated curve. The novel simulation curve and the flow characteristic curve are placed in the same flow analysis coordinate system. For any intermittent opening T1 from intermittent opening 1 to intermittent opening T in the flow analysis coordinate system, when the ordinate of the novel simulation curve with the intermittent opening T1 is not equal to the ordinate of the flow characteristic curve with the intermittent opening T1, the intermittent opening T1 is recorded as the calibration opening. Based on the calibration opening, the calibration method is used to adjust the ordinate of the flow characteristic curve with the calibrated opening. The calibration method includes: recording the intermittent pressure peak value of the intermittent pressure data corresponding to the calibration opening in the rheological extraction method corresponding to the obtained flow characteristic curve as the standard calibration peak value; The peak value of the intermittent pressure corresponding to the calibration opening in the rheological extraction method corresponding to the new simulation curve is recorded as the new simulation peak value. When the standard calibration peak value equals the new simulated peak value, the ordinate of the flow characteristic curve with the horizontal axis corresponding to the standard opening degree is replaced with the ordinate of the new simulated curve with the horizontal axis corresponding to the standard opening degree. When the standard calibration peak value is not equal to the new simulated peak value, the horizontal axis of the flow characteristic curve is not replaced with the vertical axis corresponding to the standard opening degree; All calibrated openings from intermittent opening 1 to intermittent opening T in the flow analysis coordinate system are obtained, and the calibration method is used to process all calibrated openings. Based on the flow characteristic curve obtained after processing and the new simulation curve obtained by rheological extraction method using the deep learning model, the calibration method is used to adjust the vertical axis in the flow characteristic curve. When the flow characteristic curve processed by the calibration method and the new simulated curve re-obtained by the deep learning model using the rheological extraction method are placed into the same flow analysis coordinate system, and all intermittent openings from intermittent opening 1 to intermittent opening T in the flow analysis coordinate system are not calibrated openings, the deep learning model is recorded as a deep learning model that has been trained.

Images