Plane gate valve opening and closing capacity prediction method based on parameterization calculation

Abstract

The invention discloses a plane gate valve opening and closing capacity prediction method based on parameterization calculation. The method comprises the steps that a geometric parameter set and a hydraulic state parameter set of a to-be-analyzed plane gate valve are obtained; calculating the total gravity and the top vertical water pressure of the gate valve system; in a preset operation opening degree interval, extreme value solving is conducted on a pre-configured parameterized mapping model representing the corresponding relation between the relative opening degree of the plane gate valve and the bottom flowing water acting force, so that the most unfavorable extreme value of the bottom flowing water acting force is determined; based on the total gravity of the gate valve system, the top vertical water pressure, the most unfavorable extreme value and the total friction force determined by the hydraulic state parameter set, the maximum gate opening force is calculated according to the stress balance principle, and the predicted capacity of the hoist is determined. The most unfavorable working condition can be accurately recognized, and the accuracy and reliability of opening and closing capacity prediction are improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of water conservancy and hydropower engineering, and particularly relates to a plane gate valve opening and closing capacity prediction method based on parameterized calculation. BACKGROUND

[0002] The plane gate valve is a key structure for controlling water flow in water conservancy and hydropower engineering, and accurate prediction of the opening and closing capacity thereof is a cornerstone for ensuring safe and reliable operation of the engineering. The accuracy of the opening and closing capacity prediction is directly related to the rationality of the selection of the opening and closing machine equipment, and not only affects the engineering investment, but more importantly, if the opening force under the most unfavorable condition is underestimated, the gate may not be able to be opened at a critical moment. Therefore, developing a high-precision opening and closing capacity prediction method has important technical significance for improving the reliability and safety of water conservancy engineering design.

[0003] Currently, the prediction method of the opening and closing capacity of the plane gate valve mainly relies on physical model test or computational fluid dynamics (CFD) simulation. In engineering design, a table lookup method based on hydrodynamic force coefficient table is usually used. The method obtains the hydrodynamic force coefficients by pre-testing or simulating a plurality of standardized bottom edge geometries and specific opening points (for example, relative opening degree of 0.05, 0.1, 0.2, etc.), and arranges the obtained hydrodynamic force coefficients into a table or a database. The designer looks up the most similar working condition in the existing table according to the geometric parameters of the gate to be analyzed, obtains the corresponding hydrodynamic force coefficient, and performs subsequent mechanical balance calculation accordingly.

[0004] However, the above existing technology still has some technical problems in prediction accuracy, model generalization ability and decision reliability. Specifically, since the hydrodynamic force coefficient table is obtained based on discrete opening points, the discretization process may miss the real mechanical extreme point between two measurement points, and there is a technical risk of extreme value omission. In addition, the existing coefficient table is usually for standardized bottom edge configurations, and when the bottom edge of the actual gate has unique chamfer, round corner and other detailed structural features, the model is difficult to accurately reflect the influence of these subtle changes on the water power, resulting in insufficient model generalization ability. Moreover, the traditional prediction process is a deterministic calculation, and the output result does not contain uncertainty information, which cannot quantify the confidence of the prediction value, making it difficult for the designer to make risk-based robust decisions. SUMMARY

[0005] The application aims to provide a plane gate valve opening and closing capacity prediction method based on parameterized calculation to solve the above problems in the prior art.

[0006] The technical scheme is a plane gate valve opening and closing capacity prediction method based on parameterized calculation, which comprises:

[0007] obtaining a geometric parameter set and a hydraulic state parameter set of a plane gate valve to be analyzed;

[0008] calculate the total gravity of the gate valve system based on the set of geometric parameters, and calculate the top vertical water pressure based on the set of geometric parameters and the set of hydraulic state parameters;

[0009] In a preset operating opening interval, an extreme value of the bottom dynamic water force is determined by performing extreme value solving on a preconfigured parameterized mapping model representing the corresponding relationship between the relative opening of the flat gate valve and the bottom dynamic water force.

[0010] Based on the total gravity of the gate valve system, the top vertical water pressure, the most unfavorable extreme value, and the total friction force determined by the set of hydraulic state parameters, the maximum opening force is calculated according to the force balance principle.

[0011] The predicted capacity of the gate hoist is determined based on the maximum opening force.

[0012] According to an aspect of the present application, the total gravity of the gate valve system includes the self-weight of the flat gate valve, the counterweight, and the weight of the gate valve suspender system.

[0013] The total gravity of the gate valve system is calculated based on the set of geometric parameters, specifically by using the following formula:

[0014] G Z =G1+G2+G3;

[0015] Where G Z is the total gravity of the gate valve system, G1 is the self-weight of the flat gate valve, G2 is the counterweight, and G3 is the weight of the gate valve suspender system.

[0016] According to an aspect of the present application, the top vertical water pressure is calculated by the following formula:

[0017] F u =ρ×g×(Z j0 -Z d0 )×B×d;

[0018] Where F u is the top vertical water pressure, ρ is the density of water, g is the acceleration of gravity, Z j0 is the initial water level before operation of the valve well, Z d0 is the top elevation in the fully closed state of the valve, B is the width of the flat gate valve, and d is the thickness of the flat gate valve.

[0019] According to an aspect of the present application, the parameterized mapping model defines that the bottom dynamic water force is positive upward and negative downward, and selects the calculation logic based on the type of bottom edge inclination angle defined in the set of geometric parameters:

[0020] When the bottom edge is an upward inclined bottom edge: F d (n)=ρ×g×h×B×d×k up (n, α);

[0021] When the bottom edge is a downward inclined angle bottom edge: F d (n)=p x g x h x B x d x k down (n, a);

[0022] Where F d (n) is the bottom hydrodynamic force under the relative opening n, p is the density of water, g is the acceleration of gravity, B is the width of the flat gate valve, d is the thickness of the flat gate valve, h is the action water head, a is the bottom edge inclination angle, k up (n, a) and k down (n, a) are the hydrodynamic force coefficients corresponding to the opening and the inclination angle respectively.

[0023] According to one aspect of the present application, the total friction force includes the water stop friction resistance and the roller friction resistance;

[0024] The maximum opening force is calculated according to the force balance principle, specifically by using the following formula:

[0025] T=G Z +F u -F dmin +f z ;

[0026] Where T is the maximum opening force, G Z is the total weight of the gate valve system, F u is the top vertical water pressure, F dmin is the most unfavorable extreme value, f z is the total friction force.

[0027] According to one aspect of the present application, the parametric mapping model is constructed using a continuous function model containing fractional rational terms to represent the hydraulic surge characteristics at small opening, and the continuous function model is expressed as:

[0028] F d (n)=C hyd x[(a0+a1 x n) / (n+b)+a2 x n];

[0029] Where n is the relative opening, F d (n) is the bottom hydrodynamic force, C hyd is a coefficient term related to the water head, a0, a1, a2 and b are undetermined model parameters obtained by identification, wherein the term (a0+a1 x n) / (n+b) is used to represent the nonlinear surge behavior in the small opening interval, and the term a2 x n is used to represent the linear change trend in the large opening interval.

[0030] According to one aspect of the present application, the undetermined model parameters are associated with the bottom edge detail structure features defined in the geometric parameter set;

[0031] Determination of the to-be-determined model parameters includes:

[0032] Constructing a bottom edge detail feature vector X g , the bottom edge detail feature vector X g At least including the bottom edge inclination angle, the bottom edge chamfer length, the bottom edge fillet radius, the bottom edge gap and the water stop position category;

[0033] Constructing a hydraulic state vector X h , the hydraulic state vector X h At least including the acting water head and the flow velocity under the gate;

[0034] Establishing a regression mapping relationship with the bottom edge detail feature vector X g and the hydraulic state vector X h as inputs and the to-be-determined model parameters as outputs, to determine the specific values of a0, a1, a2 and b.

[0035] According to an aspect of the present application, the method further includes an online calibration step, specifically:

[0036] During the engineering commissioning or debugging stage of the flat gate valve, measured opening and closing force data under a predetermined group of measured opening degrees are obtained;

[0037] Based on the measured opening and closing force data and the force balance principle, the measured dynamic water acting force coefficient corresponding to the opening degree is inversely obtained;

[0038] The deviation between the measured dynamic water acting force coefficient and the parameterized mapping model output value is calculated to construct a deviation calibration term;

[0039] The parameterized mapping model is corrected online by using the deviation calibration term to obtain a calibrated bottom dynamic water acting force prediction value.

[0040] According to an aspect of the present application, the parameterized mapping model is a continuous function with respect to the relative opening degree;

[0041] Determining the most unfavorable extreme value of the bottom dynamic water acting force includes:

[0042] Taking the bottom dynamic water acting force as a continuous objective function of the relative opening degree;

[0043] Within the operating opening degree interval, a global optimization operation is performed on the continuous objective function to identify the most unfavorable opening degree that makes the continuous objective function reach an extreme value;

[0044] The most unfavorable opening degree is substituted into the continuous objective function to calculate the most unfavorable extreme value.

[0045] According to an aspect of the present application, identifying the most unfavorable opening degree that makes the continuous objective function reach an extreme value includes:

[0046] Deriving a derivative function by taking a first-order derivative of the continuous objective function with respect to the relative opening;

[0047] Solving a stationary point of the derivative function being equal to zero and checking whether the stationary point is located in the operating opening interval;

[0048] Comparing the function values at the stationary point and the boundary points of the operating opening interval to determine the most unfavorable opening.

[0049] Advantages, the application can accurately identify the most unfavorable working condition, improve the accuracy and reliability of the opening and closing capacity prediction; eliminate the risk of extreme value missing judgment caused by interpolation or sparse data points, improve the accuracy of prediction; accurately reflect the influence of subtle structural changes on hydrodynamic force, improve the generalization ability and application range of the model. BRIEF DESCRIPTION OF DRAWINGS

[0050] Figure 1 A step flowchart of a plane gate valve opening and closing capacity prediction method based on parameterized calculation provided for the embodiments of the application.

[0051] Figure 2 A step flowchart of determining the to-be-determined model parameters provided for the embodiments of the application.

[0052] Figure 3 A step flowchart of online calibration provided for the embodiments of the application.

[0053] Figure 4 A step flowchart of determining the most unfavorable extreme value of the bottom hydrodynamic force provided for the embodiments of the application.

[0054] Figure 5 A gate chamber water level process line schematic diagram in the opening process of the drainage valve provided for the embodiments of the application.

[0055] Figure 6 A valve shaft water level process line schematic diagram in the opening process of the drainage valve provided for the embodiments of the application.

[0056] Figure 7 A measured valve opening and closing force process line schematic diagram in the opening process of the drainage valve provided for the embodiments of the application.

[0057] Figure 8 A plane valve bottom edge hydrodynamic force process line comparison and verification schematic diagram provided for the embodiments of the application. DETAILED DESCRIPTION

[0058] In the following, the technical solutions in the embodiments of the present application will be described clearly and completely with reference to the drawings in the embodiments of the present application, so that those skilled in the art can better understand the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work should fall within the protection scope of the present application.

[0059] It should be noted that the terms include and have as well as any variations thereof are intended to cover inclusive rather than exclusive inclusion, for example, a process, method, system, product, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements that are expressly identified as being included in the process, method, system, product, or apparatus and can include other steps or elements not expressly identified as being included in the process, method, system, product, or apparatus.

[0060] As shown in Figure 1 A parameterized calculation-based flat gate valve opening and closing capacity prediction method includes the following steps:

[0061] Obtain the geometric parameter set and the hydraulic state parameter set of the flat gate valve to be analyzed.

[0062] In the present embodiment, obtaining the geometric parameter set and the hydraulic state parameter set of the flat gate valve to be analyzed is the initialization stage of the entire prediction process. The geometric parameter set and the hydraulic state parameter set are the basis for all subsequent physical calculations and model construction. Specifically, the geometric parameter set can be organized as a structured data object, such as a JSON object or a dictionary structure in a programming language, to facilitate computer program calls. The geometric parameter set at least includes: body structure parameters such as gate width, thickness, height, and self-weight; bottom edge detail structure characteristics such as bottom edge type (e.g., upward angle or downward angle), bottom edge angle specific value, bottom edge chamfer length, bottom edge round corner radius, etc.; and accessory structure parameters such as seal device type, roller parameters, hanger system weight, etc. Correspondingly, the hydraulic state parameter set defines the external environmental conditions of the operation of the gate valve. The hydraulic state parameter set at least includes: upstream water level, downstream water level, initial flow velocity below the gate, water density, and water motion viscosity coefficient, etc. These parameters together determine the size of various water pressures and hydrodynamic forces acting on the gate.

[0063] Optionally, after obtaining the set of geometric parameters and the set of hydraulic state parameters, the system performs validity check on the input data. The check rules include: the geometric parameters such as gate width B, thickness d, height H should be positive real numbers and within a preset reasonable range, for example, B∈[1 meter, 6 meters], d∈[0.2 meters, 1.5 meters]; the acting water head h should be a non-negative real number and not exceed the maximum allowable value of the gate design water head; the bottom edge inclination angle a should be within the range of [-60 degrees, +60 degrees], where a positive value represents an upward inclination angle and a negative value represents a downward inclination angle; the fillet radius r and the chamfer length l should be non-negative real numbers. If the input parameters do not meet the check rules, the system will output an abnormal prompt information and terminate the subsequent calculation to avoid generating a prediction result without physical meaning.

[0064] The total gravity of the gate valve system is calculated based on the set of geometric parameters, and the top vertical water pressure is calculated based on the set of geometric parameters and the set of hydraulic state parameters.

[0065] Specifically, a basic statics analysis is performed on the gate to calculate several vertical forces that exist throughout the opening and closing process and are relatively stable. The total gravity of the gate valve system is the total force acting on the gate and its attached components in the gravitational field, and is a basic component of the gate opening force calculation. The top vertical water pressure is the static water pressure acting on the gate top sealing plate, and its size is directly related to the submergence depth of the gate top. It can be understood that the calculation of these forces usually follows the principles of standard fluid statics and solid mechanics.

[0066] Within a preset operating opening interval, the parameterized mapping model pre-configured to represent the corresponding relationship between the relative opening of the flat gate valve and the bottom dynamic water force is solved for extreme values to determine the most unfavorable extreme value of the bottom dynamic water force.

[0067] In this embodiment, the most unfavorable dynamic water force for the gate hoist is found within the entire gate opening process, i.e. the interval of the relative opening changing from 0 to 1. The most unfavorable extreme value usually refers to the maximum suction force (i.e. negative force), because both of these two cases will require the highest opening force. The method of extreme value solving is related to the specific form of the model. For discrete models, the process of extreme value solving is a traversal comparison; while for continuous function models, the process of extreme value solving is converted into a standard function optimization problem, which can be solved by analytical derivation or numerical optimization and other efficient methods.

[0068] Based on the total gravity of the gate valve system, the top vertical water pressure, the most unfavorable extreme value, and the total friction force determined by the set of hydraulic state parameters, the maximum gate opening force is calculated according to the force balance principle.

[0069] Specifically, the force analysis of the gate in the vertical direction is carried out, and the force system balance equation is established. The maximum gate opening force is the maximum pulling force required by the gate hoist to overcome all downward forces (including gravity, downward dynamic water force, friction force, etc.) and lift the gate upward under the most unfavorable working condition. The total friction force is another important resistance term, which usually includes the friction resistance of the water stop and the friction resistance of the roller, and its size is closely related to the water pressure and the structure of the gate. By algebraically summing all force components, the predicted value of the maximum gate opening force can be obtained according to the force balance principle.

[0070] The predicted capacity of the gate hoist is determined based on the maximum gate opening force.

[0071] In this embodiment, after calculating the maximum gate opening force, considering the safety redundancy in engineering design, a predetermined safety factor, for example, a value between 1.1 and 1.2, needs to be multiplied to obtain the design capacity or checking capacity of the gate hoist. The predicted capacity of the gate hoist can finally be used as a key design or checking basis to guide engineering practice. The output result can be a prediction report containing detailed calculation process and final capacity value.

[0072] Optionally, based on the bottom edge detail structure features defined in the geometric parameter set, a parameterized mapping model representing the corresponding relationship between the relative opening degree of the flat gate valve and the bottom dynamic water force can be constructed.

[0073] Specifically, by constructing a parameterized mathematical model, the continuous relationship between the bottom dynamic water force of the gate and the relative opening degree is accurately described. The construction of the parameterized mapping model can convert discrete and empirical data points into continuous functions or high-precision mappers that can reflect the internal physical laws and can be mathematically analyzed. The input of the parameterized mapping model includes not only the relative opening degree as the independent variable, but more importantly, the internal parameters or structure of the model itself are directly determined by the bottom edge detail structure features (such as inclination angle, chamfer, etc.). The model has good generalization ability and can predict different gate designs with different bottom edge designs. The parameterized mapping model can be specifically implemented as a discrete model or preferably a continuous function model. It solves the problem of low precision and inability to reflect the influence of bottom edge details in traditional lookup table method.

[0074] It should be noted that the parameterization calculation-based flat gate valve opening and closing capacity prediction method provided in the present application can run on a general computing device. As an example, the computing device can be a server or a personal computer configured with a central processing unit (CPU), a memory (RAM) with sufficient capacity, and a hard disk storage. At the software level, the parameterization calculation-based flat gate valve opening and closing capacity prediction method can be implemented on a mainstream operating system, such as a Windows Server or a Linux system, and deployed through a high-level programming language, such as Python or MATLAB, to facilitate complex numerical calculations and model building.

[0075] In one possible implementation, the total gravity of the gate valve system includes the self-weight of the flat gate valve, the counterweight, and the weight of the gate valve hanger system.

[0076] In the present embodiment, the total gravity of the gate valve system is a constant force acting vertically downward, and its value is not affected by the changes in the gate opening or the water flow state. The total gravity of the gate valve system is composed of three parts. As an example, the total gravity of the gate valve system is calculated based on the set of geometric parameters, specifically calculated by using the following formula:

[0077] G Z =G1+G2+G3;

[0078] where G Z is the total gravity of the gate valve system; G1 is the self-weight of the flat gate valve, i.e., the weight of the flat gate valve leaf itself; G2 is the counterweight, i.e., the weight of the counterweight block provided to balance the weight of the gate; and G3 is the weight of the gate valve hanger system, i.e., the weight of the hanger, lug, and other components connecting the gate valve to the gate leaf. These parameters are usually provided by the design drawings or the manufacturer of the gate valve and are known geometric parameters.

[0079] According to one aspect of the present application, the top vertical water pressure is obtained by calculation through the following formula:

[0080] F u =ρ×g×(Z j0 -Z d0 )×B×d;

[0081] where F u is the top vertical water pressure; ρ is the density of water, which is usually 1000 kg / m3; g is the acceleration of gravity, which is usually 9.8 m / s2; Z j0 is the initial water level elevation in the valve shaft before the gate is opened; Z d0 is the elevation of the top of the valve in the fully closed state, so (Z j0 -Z d0) represents the effective submergence depth of the gate top; B is the width of the flat gate, and d is the thickness of the flat gate. The vertical water pressure on the top is the total vertical upward static water pressure on the gate top due to its submergence in static water, and the calculation of the vertical water pressure on the top follows the basic principles of fluid statics.

[0082] In an exemplary embodiment, the parameterized mapping model defines the bottom hydrodynamic force as positive upward and negative downward, and selects the calculation logic based on the type of bottom edge inclination defined in the geometric parameter set:

[0083] In this embodiment, the parameterized mapping model is specifically implemented as a calculation method based on a discrete coefficient table. Specifically, the hydrodynamic force coefficients under different working conditions are obtained in advance through physical model tests or a series of computational fluid dynamics (CFD) simulations, and are sorted into a data table. During calculation, the corresponding calculation branch is selected according to the geometric configuration of the gate bottom edge.

[0084] When the bottom edge is an upward inclined bottom edge, its hydrodynamic force usually exhibits an upward lifting force, and the calculation formula is:

[0085] F d (n)=ρ×g×h×B×d×k up (n, a);

[0086] When the bottom edge is a downward inclined bottom edge, its hydrodynamic force usually exhibits a downward suction force at small opening, and the calculation formula is:

[0087] F d (n)=ρ×g×h×B×d×k down (n, a);

[0088] where F d (n) is the bottom hydrodynamic force under the relative opening n; h is the action water head, usually referring to the water level difference before and after the gate; a is the bottom edge inclination; k up (n, a) and k down (n, a) are the hydrodynamic force coefficients corresponding to the relative opening n and the bottom edge inclination a, respectively, which can be obtained by querying the pre-prepared coefficient table. It can be understood that the extreme value solving process is to find the maximum or minimum coefficient value in the coefficient table. As an optional implementation, the hydrodynamic force coefficient can also be approximately obtained by a semi-empirical formula based on physical principles, for example, for the hydrodynamic force coefficient of a downward inclined bottom edge, it can be approximately expressed in the form related to the sine function of the bottom edge inclination.

[0089] In one possible embodiment, the total friction force includes a water-stopping friction resistance and a roller friction resistance; based on the total gravity of the gate valve system, the top vertical water pressure, the most unfavorable extreme value, and the total friction force determined by the set of hydraulic state parameters, the maximum opening force is calculated according to the force balance principle, specifically by using the following formula:

[0090] T = G Z + F u - F dmin + f z ;

[0091] wherein T is the maximum opening force, G Z is the total gravity of the gate valve system, F u is the top vertical water pressure, F dmin is the most unfavorable extreme value, and f z is the total friction force.

[0092] Specifically, the total friction force is one of the main resistances that hinder the upward movement of the gate. Exemplarily, the total friction force f z is composed of two parts:

[0093] f z = f seal + f roller ;

[0094] wherein f seal is the water-stopping friction resistance, and f roller is the roller friction resistance.

[0095] Further, the water-stopping friction resistance f seal is the friction force generated due to the water pressure pressing the rubber water-stopping against the gate groove embedded part. Its size can be estimated by the following formula: f seal = μ seal × P seal × L seal ; wherein μ seal is the water-stopping friction coefficient, which is related to the material and the state of the contact surface; P seal is the average pressure on the water-stopping line, which is related to the acting water head h; and L seal is the total length of the water-stopping. Correspondingly, the roller friction resistance f roller is the resistance generated when the gate roller rolls on the track. Its size can be estimated by the following formula: f roller = μ roller × P h ; wherein μ roller is the comprehensive friction coefficient of the roller, and P h is the total horizontal static water pressure acting on the entire gate leaf. After all the force components are calculated, the maximum opening force T can be calculated according to the force balance principle in the vertical direction, wherein F dminFor the most unfavorable extreme value, under the calculation logic of the embodiment, it refers to the maximum hydrodynamic suction in the entire opening interval, which is calculated by looking up the coefficient table or empirical formula. Since it is specified to be negative downward, F dmin is negative.

[0096] As an example, the complete calculation process of the embodiment is described below through a specific numerical calculation case.

[0097] Suppose the geometric parameter set and the hydraulic state parameter set of a certain flat gate valve are as follows: the gate width B is 5 meters, the thickness d is 0.8 meters; the total gravity G Z of the gate valve system is 1500 kilonewtons; the top elevation Z d0 is 100 meters, the gate well water level Z j0 before opening is 102 meters; the acting water head h is 30 meters; the total friction f z is 200 kilonewtons. The gate is a downward inclined angle bottom edge. By looking up the corresponding hydrodynamic force coefficient table, it is found that when the relative opening n is 0.4-0.5, the most unfavorable hydrodynamic force coefficient k down is negative 0.1. Calculate the top vertical water pressure F u : F u = 1000 × 9.8 × (102-100) × 5 × 0.8 = 78400 newtons = 78.4 kilonewtons; calculate the most unfavorable bottom hydrodynamic force F dmin : F dmin = 1000 × 9.8 × 30 × 5 × 0.8 × (-0.1) = -117600 newtons = -117.6 kilonewtons; calculate the maximum opening force T according to the force balance formula: T = 1500-78.4-(-117.6)+200 = 1500-78.4+117.6+200 = 1739.2 kilonewtons. It is calculated that the required maximum opening force under the specific working condition is 1739.2 kilonewtons.

[0098] In another embodiment of the application, the parameterized mapping model is constructed using a continuous function model containing fractional rational terms to represent the hydraulic surge characteristics at small openings, which is expressed as:

[0099] F d (n) = C hyd × [(a0+a1×n) / (n+b)+a2×n];

[0100] where n is the relative opening, whose value range is between 0 and 1; F d (n) is the bottom hydrodynamic force that changes with the relative opening n; C hyd is a coefficient term related to the hydraulic condition; a0, a1, a2, and b are model parameters to be determined, which together determine the specific form of the hydrodynamic force curve.

[0101] In this embodiment, in order to accurately capture the nonlinear behavior of the water action force with the opening degree, especially the sharp change in the small opening degree area, a continuous function model with physical meaning is proposed. Specifically, the design of the continuous function structure incorporates physical insight. The fractional rational term (a0+a1×n) / (n+b) is used to specifically represent the nonlinear surge or sudden drop behavior caused by the sharp change of the gap between the gate bottom due to the local flow state disorder when the relative opening degree n approaches zero. The linear term a2×n is used to represent the trend that the water action force changes roughly linearly with the opening degree in the large opening degree interval. By superimposing these two parts, the model can take into account the nonlinear characteristics of the small opening degree area and the linear trend of the large opening degree area, and achieve high-precision fitting in the entire opening degree interval. Exemplarily, the coefficient term C hyd =ρ×g×h×B×d; where ρ is the density of water, g is the acceleration of gravity, h is the action water head, B is the gate width, and d is the gate thickness.

[0102] As Figure 2 shown, in further embodiments, the to-be-determined model parameters are associated with the bottom edge detail structure features defined in the geometric parameter set;

[0103] A parameterized mapping model representing the corresponding relationship between the relative opening degree of the flat gate valve and the bottom water action force is constructed, including determining the to-be-determined model parameters, specifically:

[0104] A bottom edge detail feature vector X g is constructed, which at least includes the bottom edge inclination angle, the bottom edge chamfer length, the bottom edge fillet radius, the bottom edge gap, and the water stop position category; g

[0105] A hydraulic state vector X h is constructed, which at least includes the action water head and the flow rate under the gate; h

[0106] A regression mapping relationship is established with the bottom edge detail feature vector X g and the hydraulic state vector X h as inputs and the to-be-determined model parameters as outputs, to determine the specific values of a0, a1, a2, and b.

[0107] In this embodiment, a meta-model is constructed, i.e., a regression model is established that can automatically predict the best parameters (a0, a1, a2, b) of the function model according to the specific geometry and hydraulic working conditions of the gate. Specifically, the physical factors affecting the water action force need to be numerized, i.e., the feature vector is constructed. The bottom edge detail feature vector X g ​​The geometry of the gate bottom is quantified. For example, the bottom edge inclination angle can be directly used as its angle value, and the bottom edge chamfer length and the fillet radius are used as their length values. For the water stop position category, a one-hot encoding can be used for processing, for example, using the vector [1, 0] to represent the bottom-mounted water stop, and [0, 1] to represent the side-mounted water stop. The hydraulic state vector X h is used to quantify the flow environment of the gate operation. The hydraulic state vector contains values such as the acting water head and the flow velocity below the gate. A regression mapping relationship is established from the input feature vector [X g , X h ] to the output parameters [a0, a1, a2, b].

[0108] In a preferred implementation, a regression mapping relationship is established with the bottom edge detail feature vector X g and the hydraulic state vector X h as inputs, and the to-be-determined model parameters as outputs, which is implemented by any of the following methods:

[0109] A multi-layer perception neural network is constructed with X g and X h as the input layer and a0, a1, a2, and b as the output layer for training; or a Gaussian process regression model is constructed to learn the nonlinear mapping law in the pre-stored sample set, and output the parameter prediction mean and prediction variance.

[0110] Specifically, two preferred schemes for constructing a regression mapping relationship are provided. The first scheme is to construct a multi-layer perception (MLP) neural network. The number of nodes in the input layer of the network is equal to the sum of the dimensions of the feature vectors X g and X h , and the output layer contains 4 nodes corresponding to the four model parameters a0, a1, a2, and b. Between the input layer and the output layer, one or more hidden layers can be set, for example, two hidden layers can be set, each containing 16 neurons, and using a rectified linear unit (ReLU) as the activation function. By training on a large number of working condition-parameter sample pairs, the neural network can learn the complex nonlinear mapping law.

[0111] The second scheme is to construct a Gaussian process regression (GPR) model. Unlike the neural network output deterministic prediction value, the advantage of the GPR model is that it can not only provide the prediction mean of the model parameters, but also give the confidence of the prediction result, i.e. the prediction variance. The prediction variance quantifies the uncertainty of the model under a certain input working condition, which is the key basis for subsequent risk assessment and robustness optimization.

[0112] In a detailed numerical example, assume that the bottom edge detail feature vector X gFor example, the bottom edge inclination angle a = 30 degrees, the bottom edge chamfer length l = 0 mm, the bottom edge fillet radius r = 20 mm, the bottom edge gap g = 5 mm, and the sealing position category is bottom-pasting type (coded as [1, 0]). The hydraulic state vector X h For example, the action water head h = 30 m, and the downstream flow velocity v = 2 m / s. The above feature vector is input into the pre-trained multi-layer perceptron neural network or Gaussian process regression model, and the output undetermined model parameters are: a0 = -0.08, a1 = 0.15, a2 = 0.02, and b = 0.03. The undetermined model parameters are substituted into the continuous function model, and the hydrodynamic force-opening relationship curve of the gate is obtained. Further, if the Gaussian process regression model is used, in addition to the output parameter prediction mean, the parameter prediction variance can also be output, for example, s a0 = 0.01, s a1 = 0.02, s a2 = 0.005, and s b = 0.008. The variance information can be used to construct the conservative upper bound objective function.

[0113] In some optional embodiments, other machine learning models capable of realizing high-dimensional nonlinear mapping can also be used to construct the meta-model. For example, support vector regression (SVR) or gradient boosting decision tree (GBDT) algorithms can be used to establish the regression mapping relationship from the feature vector to the model parameters.

[0114] In an optional embodiment, the undetermined model parameters can be obtained by pre-identification through the following steps:

[0115] An offline sample data set containing the corresponding relationship between the predetermined group of relative opening and hydrodynamic force is obtained. A weighted least squares objective function containing physical constraints is constructed, and the physical constraints at least include the sign constraint of the parameters and the boundedness constraint of the function value in the definition domain. Different weight coefficients are assigned to the sample data in different relative opening intervals, and the weight coefficient set value in the small opening interval is higher than that in the large opening interval. The minimum value of the weighted least squares objective function is solved by a numerical optimization algorithm to determine the optimal parameter combination of a0, a1, a2, and b.

[0116] In this embodiment, a single specific working condition (i.e., a set of fixed X g and X h) Determine its corresponding optimal model parameters (a0, a1, a2, b), which provides a training sample for the meta-model. Specifically, the obtained offline sample dataset can be derived from a series of high-precision computational fluid dynamics (CFD) simulations or accurate physical hydraulic model tests. A weighted least squares (WLS) objective function is constructed for parameter identification. The most concerned in engineering is the extreme suction force that may occur in the small opening area, so the model needs to have the highest fitting accuracy in this area, and a weighting mechanism is preferably used. Exemplarily, the weight coefficient w can be determined by an exponential decay function with respect to the relative opening n, for example, w(n) = exp(-γ*n) + w0; where γ is a decay coefficient greater than 0, and w0 is the basic weight. The exponential decay function makes the sample points with n approaching to 0 obtain much larger weight than the sample points with n approaching to 1.

[0117] Further, introducing physical constraints into the objective function can prevent the model from having solutions that violate physical laws. For example, parameter b must be positive to avoid singularities of the function within the physically defined domain; for another example, the signs of parameters a0, a1, a2 can be constrained according to prior knowledge, or the function value of the model at a specific opening can be constrained to be bounded. Through a numerical optimization algorithm, such as the Levenberg-Marquardt (LM) algorithm, the weighted and constrained nonlinear least squares problem is solved to obtain the optimal parameter combination.

[0118] As shown in FIG. 1, Figure 3 In yet further embodiments, the method further comprises an online calibration step, specifically:

[0119] During the engineering commissioning or debugging phase of the flat gate valve, measured opening and closing force data at a predetermined set of openings are obtained;

[0120] Based on the measured opening and closing force data and the force balance principle, the measured hydrodynamic force coefficient at the corresponding opening is obtained by inversion;

[0121] The deviation between the measured hydrodynamic force coefficient and the output value of the parameterized mapping model is calculated, and a deviation calibration term is constructed;

[0122] The parameterized mapping model is corrected online using the deviation calibration term to obtain the calibrated bottom hydrodynamic force prediction value.

[0123] In this embodiment, a digital twin application is provided, which can eliminate the deviation between the pre-trained model and the actual physical entity. Due to factors such as machining errors, installation deviations or long-term running wear, there may be differences between the pre-trained theoretical model and the actual situation on site. The calibration process uses a small amount of high-precision measured opening and closing force data on site to obtain the dynamic water force under the real working condition through inverse calculation of the force balance equation. Compare the real value with the model prediction value to get a series of deviation data points. On this basis, a deviation calibration function δ(n) can be constructed, for example, by polynomial fitting or spline interpolation of the deviation data points. The model output F d_calibrated (n) after online calibration is:

[0124] F d_calibrated (n) = F d_model (n) + δ(n).

[0125] Where F d_model (n) is the model prediction value. This makes the model's prediction results highly consistent with the actual situation on site, achieving personalized and accurate correction of the model.

[0126] This embodiment solves the problems of limited accuracy, inability to interpolate working conditions not included in the table, and weak generalization ability of the method based on discrete coefficient table by constructing a continuous, physically law-constrained parameterized mapping model, achieving higher accuracy and higher generalization ability in predicting dynamic water force.

[0127] As shown in Figure 4 , in another embodiment of the present application, the parameterized mapping model is a continuous function with respect to the relative opening degree; within a predetermined operating opening degree interval, the parameterized mapping model is solved for extreme values to determine the most unfavorable extreme value of the bottom dynamic water force, including:

[0128] Taking the bottom dynamic water force as a continuous objective function of the relative opening degree;

[0129] Within the operating opening degree interval, a global optimization operation is performed on the continuous objective function to identify the most unfavorable opening degree that causes the continuous objective function to reach an extreme value;

[0130] Substitute the most unfavorable opening degree into the continuous objective function to calculate the most unfavorable extreme value.

[0131] In this embodiment, since the continuous function expression of the bottom dynamic water force F d (n) with respect to the relative opening degree n is obtained, the problem of finding the most unfavorable extreme value is transformed from a comparison problem of discrete data points to a global optimization problem of a standard continuous function on a given closed interval, i.e. [0, 1]. The mathematical problem is to mathematize the physical problem, i.e. to obtain the continuous function expression of the bottom dynamic water force F d(n) is defined as an objective function for optimization, whose argument is the relative opening n. The goal of the global optimization operation is to find one or more opening positions such that the objective function F d (n) reaches a global maximum or minimum value within the interval. Once the most unfavorable opening n* is determined, the corresponding most unfavorable extreme value, i.e. F d (n*) can be obtained by simply evaluating the function.

[0132] In a preferred embodiment, the global optimization operation is performed on the continuous objective function to identify the most unfavorable opening that makes the continuous objective function reach an extreme value, which includes the analytical derivation method, specifically:

[0133] First-order derivative of the continuous objective function with respect to the relative opening is obtained to obtain a derivative function;

[0134] Solve the stationary point of the derivative function equal to zero and check whether the stationary point is located within the operating opening interval;

[0135] Compare the function values at the stationary point and the boundary points of the operating opening interval to determine the most unfavorable opening.

[0136] Specifically, the first-order derivative of the function F d (n) with respect to the relative opening n is obtained to obtain a derivative function dF d / dn. Let the derivative function equal to zero, i.e. dF d / dn = 0, and solve the equation to obtain all stationary points. Filter out all stationary points located within the operating opening interval [0, 1]. Substitute the effective stationary points and the two boundary points of the interval, i.e. n = 0 and n = 1, into the original function F d (n) to calculate the function values. Compare the function values of all candidate points, and the maximum and minimum values are the global extreme values of the function within the operating opening interval. This embodiment can accurately find the extreme point, but the premise is that the model has no uncertainty, and it is suitable for the continuous objective function, i.e. F d (n) function is relatively simple and easy to derive.

[0137] In an alternative embodiment, the global optimization operation is performed on the continuous objective function to identify the most unfavorable opening that makes the continuous objective function reach an extreme value, which specifically includes the adaptive search method, which is used to solve the conservative upper bound extreme value of the gate force:

[0138] A conservative upper bound objective function T ub (n) of the gate force is constructed:

[0139] T ub (n) = T(n) + λ × σ T (n);

[0140] where T(n) is the predicted opening force varying with the relative opening degree n based on the parametric mapping model; λ is a preset confidence coefficient; σ T (n) is the prediction standard deviation of the opening force T(n), which is determined by the prediction variance of the parametric mapping model.

[0141] In the present embodiment, considering that the constructed parametric mapping model itself may have prediction errors, in order to obtain a more safe and conservative decision result in engineering, a self-adaptive search method considering model uncertainty is preferably adopted. Instead of directly optimizing the predicted opening force, a conservative upper bound objective function of the opening force is constructed and optimized, wherein the calculation method of the opening force T(n) is consistent with the force balance equation. The construction of the conservative upper bound objective function embodies the risk-based decision-making idea, which not only considers the predicted mean value T(n) of the opening force, but also adds a safety margin λ×σ T (n) through the second term. The safety margin is proportional to the uncertainty σ T (n) of the model at this point. Therefore, the optimization target is no longer to find the extreme value of the predicted value, but to find the extreme value of the combination of the predicted value + safety margin, so that the optimization result can fully consider the possible prediction deviation of the model, and the result is more safe and reliable.

[0142] It should be noted that the confidence coefficient λ is a dimensionless parameter, and its value reflects the risk preference of the decision maker. A larger λ value, for example, λ equal to 3, corresponds to an upper bound of a confidence interval of about 99.7%, which means that a more conservative attitude is taken towards uncertainty, which will make the optimization result more biased towards the region with larger model variance, and obtain a higher safety margin, which is suitable for high-risk or critical gate valve engineering. A smaller λ value, for example, λ equal to 1, indicates that more reliance is placed on the model prediction mean value.

[0143] In a preferred implementation, the prediction variance of the parametric mapping model is obtained in the following way:

[0144] Statistically identify the model residual distribution in the offline identification process, and calculate the standard deviation of the residual as a fixed prediction variance;

[0145] Or, when a Gaussian process regression model is used to construct the parametric mapping relationship, the posterior variance of the Gaussian process regression output is directly obtained as the prediction variance.

[0146] Specifically, two specific paths are provided for the source of the uncertainty term σ T (n). The first way is to statistically identify the residual between the model prediction value and the true value of all sample points in the parameter offline identification process, and calculate the overall standard deviation of the residual. The square of the overall standard deviation can be taken as a fixed prediction variance, representing the average uncertainty level of the model.

[0147] In a preferred second mode, when a Gaussian Process Regression (GPR) model is used to construct the meta-model, the GPR model itself can output the corresponding posterior variance for each new input condition. The posterior variance varies with the input, and can reflect that the model has higher uncertainty in areas with sparse training data, and lower uncertainty in areas with dense data. Directly using the posterior variance as the model prediction variance can more finely quantify the uncertainty.

[0148] For example, assume that the parameterized mapping model obtained by the Gaussian Process Regression model has a predicted mean T(0.05)=1500 kN of the opening force at a relative opening degree n=0.05, and a predicted standard deviation σ T (0.05)=80 kN. When a confidence coefficient λ=1 of a lower degree of conservatism is used, the conservative upper limit target function value is: T ub (0.05)=1500+1×80=1580 kN. When a confidence coefficient λ=3 of a higher degree of conservatism is used, the conservative upper limit target function value is: T ub (0.05)=1500+3×80=1740 kN. By adaptively searching T ub (n) in the entire operating opening degree interval [0, 1], assume that when λ=1, the most unfavorable opening degree n=0.06, and the corresponding maximum predicted opening force value is 1620 kN; when λ=3, the most unfavorable opening degree n=0.04, and the corresponding maximum predicted opening force value (conservative upper limit) is 1850 kN. It can be seen that a larger confidence coefficient leads to more conservative prediction results, and is suitable for engineering scenarios with higher safety requirements. It should be noted that this example is only used to demonstrate the conservative upper limit of the Gaussian Process Regression and the most unfavorable load search method, and the most unfavorable opening degree appearing in the small opening degree interval is only an example effect; in actual flat valve engineering, the most unfavorable load usually appears in the relative opening degree interval 0.4-0.5.

[0149] and uses an adaptive grid search algorithm or interval splitting algorithm to search for the opening degree position that maximizes the conservative upper limit target function T ub (n) in the operating opening degree interval, and takes the opening degree position as the most unfavorable opening degree n*.

[0150] As a preferred implementation, the adaptive search process can be implemented by a Bayesian optimization algorithm. Bayesian optimization is to intelligently select the next evaluation point by maximizing the acquisition function, and the T ub (n) function in this embodiment is exactly the Upper Confidence Bound (UCB) acquisition function. The adaptive grid search algorithm can find the opening degree position n* that maximizes the conservative upper limit target function T ub (n) in the global range with very high efficiency and the least number of function evaluations.

[0151] The embodiment can more accurately and reliably locate the global most unfavorable point, effectively avoiding the problem of missed judgment caused by sparse data points.

[0152] According to another aspect of the present application, the construction of the continuous function model is realized based on the jet contraction theory, comprising:

[0153] A relationship model of the jet contraction coefficient and the relative opening degree is established, and the jet contraction coefficient is defined as the ratio of the actual jet cross-sectional area to the geometric opening area.

[0154] In the embodiment, when the flat gate valve is partially opened, the upstream water body is ejected from the gate bottom gap to form a contracting jet. The jet contraction coefficient is a core physical quantity representing the degree of jet contraction. The variation law of the jet contraction coefficient with the relative opening degree can be obtained from the asymptotic analysis of potential flow theory. At a very small opening degree, due to the sharp turning of the boundary, the jet contraction is most severe, and the contraction coefficient tends to a certain limit value; as the opening degree increases, the boundary turning effect weakens, the jet contraction tends to disappear, and the contraction coefficient asymptotically tends to 1. In a preferred implementation, a relationship model of the jet contraction coefficient and the relative opening degree is established, specifically:

[0155] C c (n)=1-(1-C c0 )×exp(-β×n / n ref );

[0156] Where C c (n) is the jet contraction coefficient at the relative opening degree n; C c0 is the zero opening limit contraction coefficient, which is usually in the range of 0.61 to 0.85, and the specific value is determined by the bottom edge geometry; β is the contraction characteristic index, representing the speed of contraction effect decay with opening degree, and its typical value range is 2 to 5; n ref is the reference opening degree, defined as the ratio of the bottom edge characteristic length to the net height of the gate hole, for dimensionless; exp is the natural exponential function.

[0157] The zero opening limit contraction coefficient in the relationship model is determined based on the pre-stored bottom edge geometric parameters, including the bottom edge inclination angle, the fillet radius, and the chamfer length.

[0158] In the embodiment, a relationship model of the zero opening limit contraction coefficient and the bottom edge geometric parameters is established. The zero opening limit contraction coefficient C c0 is determined by the detailed geometry of the bottom edge. Illustratively, the zero opening limit contraction coefficient in the relationship model is determined based on the bottom edge geometric parameters, specifically:

[0159] C c0 =C c_base ×[1+Δ α +Δr + Δ l ];

[0160] wherein C c0 is the zero opening limit contraction coefficient; C c_base is the reference contraction coefficient, which is π / (π+2) for a right-angle sharp-edged bottom lip according to the Borda contraction theory, approximately equal to 0.611; Δ α is the bottom lip inclination correction term, and a is the bottom lip inclination; Δ r is the bottom lip rounding correction term, and r is the bottom lip rounding radius; Δ l is the bottom lip chamfer correction term, and l is the bottom lip chamfer length;

[0161] The bottom lip inclination correction term is:

[0162] For an upwardly inclined bottom lip, i.e. the inclination is inclined in the upstream direction: Δ α = k α_up × sin(a) × (1 + cos(a));

[0163] For a downwardly inclined bottom lip, i.e. the inclination is inclined in the downstream direction: Δ α = k α_down × sin(a) × (1 - cos(a));

[0164] wherein a is the bottom lip inclination; k α_up is the upward inclination influence coefficient; k α_down is the downward inclination influence coefficient, and sin is the sine function and cos is the cosine function.

[0165] In the present embodiment, the specific form of the bottom lip inclination correction term is determined according to the inclination type. For an upwardly inclined bottom lip, the inclination correction term is positive, indicating that the upward inclination can slow down the jet contraction and increase the contraction coefficient; for a downwardly inclined bottom lip, the inclination correction term is negative, indicating that the downward inclination will intensify the jet contraction. The typical value of the upward inclination influence coefficient k α_up is 0.3, and the typical value of the downward inclination influence coefficient k α_down is 0.2.

[0166] The bottom lip rounding correction term Δ r is:

[0167] Δ r = k r × (r / d) 0.5 × [1 - exp(-(r / d) / 0.1)];

[0168] wherein r is the bottom lip rounding radius, d is the gate thickness, k r is the rounding influence coefficient, and the typical value is 0.4; the existence of the rounding can guide the smooth transition of the water flow and slow down the jet contraction.

[0169] Bottom chamfer correction term Δ l is:

[0170] Δ l = k l × (l / d) / (1+l / d);

[0171] where l is the length of the bottom chamfer, k l is the chamfer influence coefficient, typically 0.15. The effect of chamfer is similar to that of fillet, but relatively weak.

[0172] It should be noted that the typical values of the influence coefficients (k α_up = 0.3, k α_down = 0.2, k r = 0.4, k l = 0.15) are based on a large number of hydraulic model test data statistics. In practical application, these coefficients can be adjusted within a certain range: the value range of k α_up is usually 0.2 to 0.4, and the larger value is suitable for the case where the inclination angle is larger and the streamline bending effect is more significant; the value range of k α_down is usually 0.1 to 0.3, and the larger value is suitable for the case where the downstream flow pattern is more turbulent; the value range of k r is usually 0.3 to 0.5, and the larger the fillet radius, the more significant the influence of the fillet correction term; the value range of k l is usually 0.1 to 0.2, and the larger the chamfer length, the more significant the influence of the chamfer correction term. Sensitivity analysis of parameters shows that the fillet correction coefficient k r has the most significant influence on the zero opening limit contraction coefficient C c0 , and the accurate measurement of the bottom fillet radius should be prioritized in engineering design.

[0173] Based on the momentum theorem, the analytical relationship between the dimensionless hydrodynamic force coefficient and the relative opening is derived using the jet contraction coefficient, and a continuous function model is obtained.

[0174] In this embodiment, after obtaining the jet contraction coefficient, the hydrodynamic force acting on the gate bottom can be derived according to the momentum theorem of fluid mechanics. The gap area at the bottom of the gate is selected as the control body, the momentum flux and pressure distribution at the inlet and outlet of the control body are considered, and the momentum balance equation is established according to Newton's second law. After derivation, the bottom hydrodynamic force can be expressed as:

[0175] F d (n)=ρ×g×h×B×d×K d (n);

[0176] where F d(n) is the bottom hydrodynamic force at relative opening n, p is the density of water, g is the acceleration of gravity, h is the water head in front of the gate, B is the gate width, d is the gate thickness, K d (n) is the dimensionless hydrodynamic force coefficient.

[0177] In a preferred implementation, the analytical relationship between the dimensionless hydrodynamic force coefficient and the relative opening is derived based on the momentum theorem, specifically:

[0178] K d (n) = (1 - C c (n)) / C c (n) - [(1 - C c (n)) 2 / (2 x C c (n) 2 )] x [1 / (n + ε) 2 ] x Φ(n);

[0179] where K d (n) is the dimensionless hydrodynamic force coefficient at relative opening n, C c (n) is the jet contraction coefficient at relative opening n, n is the relative opening, and ε is a regularization parameter with a typical value of 0.01 to avoid numerical singularity at n equal to 0. The first term (1 - C c (n)) / C c (n) in the formula reflects the static pressure difference effect due to jet contraction. When the jet contracts, the average pressure at the jet cross-section is lower than the pressure at the gate bottom surface, resulting in an upward thrust or a reduction in the suction force. The second term in the formula reflects the dynamic pressure effect produced by jet acceleration. Water flow accelerates from the low-speed area in front of the gate to the high-speed jet area at the gate bottom, and according to Bernoulli's equation, the pressure decreases with the increase in velocity, and the dynamic pressure effect produces local negative pressure in the jet contraction area, which is the main source of suction force. Φ(n) is a boundary layer correction function used to correct the deviation caused by the limited thickness of the boundary layer, and its expression is:

[0180] Φ(n) = 1 - exp(-n / n δ );

[0181] where n δ is the characteristic opening of the boundary layer, typically 0.02, and exp is the natural exponential function.

[0182] In a specific numerical case, assume that the bottom edge of a flat gate valve is designed with a downward inclination angle of 30 degrees, a round corner radius of 20 mm, a chamfer length of 0, and a gate thickness of 800 mm. Calculate each correction term. The inclination correction term is: Δ α= 0.2 x sin(30°) x (1 - cos(30°)) = 0.2 x 0.5 x (1 - 0.866) = 0.0134; the corner correction term is: Δ r = 0.4 x (0.02 / 0.8) 0.5 x [1 - exp(-0.25)] = 0.4 x 0.158 x 0.221 = 0.0140; the chamfer correction term is: Δ l = 0; the zero opening limit contraction coefficient is: C c0 = 0.611 x (1 + 0.0134 + 0.0140 + 0) = 0.611 x 1.0274 = 0.628. Taking the contraction characteristic index β equal to 3, the reference opening n ref equal to 0.1, then at the relative opening n equal to 0.05, the jet contraction coefficient is: C c (0.05) = 1 - (1 - 0.628) x exp(-3 x 0.05 / 0.1) = 1 - 0.372 x exp(-1.5) = 1 - 0.372 x 0.223 = 0.917. Substituting into the dynamic force formula, taking ε equal to 0.01, n δ equal to 0.02, the boundary layer correction function is: Φ(0.05) = 1 - exp(-0.05 / 0.02) = 1 - exp(-2.5) = 1 - 0.082 = 0.918. The dimensionless dynamic water force coefficient is: K d (0.05) = (1 - 0.917) / 0.917 - [(1 - 0.917) 2 / (2 x 0.917 2 )] x [1 / (0.05 + 0.01) 2 ] x 0.918 = 0.0905 - 1.045 = -0.955. The negative value indicates that the dynamic water force direction is downward at this opening, i.e. the downward suction force.

[0183] In an alternative embodiment, the continuous function model is constructed in a piecewise coupled form, including:

[0184] The jet velocity is calculated based on the jet contraction coefficient, and the local Froude number is determined by the ratio of the jet velocity to the local water depth.

[0185] In this embodiment, the flow regime identification criterion is defined. The flow regime characteristics at different openings are essentially different, and it is difficult to accurately describe the dynamic water force characteristics in the full opening interval with a single model. The local Froude number can be introduced as a flow regime discrimination criterion:

[0186] Fr local (n) = V jet (n) / sqrt(g x h local (n));

[0187] where Frlocal (n) is the local Froude number at relative opening n, V jet (n) is the jet velocity, g is the gravitational acceleration, h local (n) is the local water depth in the jet zone, sqrt is the square root function. The jet velocity V jet (n) can be obtained from the continuity equation:

[0188] V jet (n) = Q / (B x e x n x C c (n)) = mu x sqrt (2 x g x h) / (n x C c (n));

[0189] where V jet (n) is the jet velocity, Q is the discharge, B is the gate width, e is the gate height, n is the relative opening, C c (n) is the jet contraction coefficient, mu is the gate discharge coefficient, g is the gravitational acceleration, h is the upstream water head, sqrt is the square root function. For the submerged outflow condition, the local water depth h local (n) can be approximated as the downstream water depth h d .

[0190] According to the comparison result of the local Froude number and the preset critical Froude number, the first critical opening and the second critical opening of the flow regime transition are determined.

[0191] Specifically, the critical opening of the flow regime transition is determined: two critical Froude numbers are defined to distinguish different flow regime intervals. The first critical Froude number Fr cr1 is used to distinguish the boundary between the jet-dominated zone and the transition zone, and the typical value is 2.5. The second critical Froude number Fr cr2 is used to distinguish the boundary between the transition zone and the pressure difference-dominated zone, and the typical value is 1.0. The first critical opening n1* of the flow regime transition is determined by solving the following equation:

[0192] Fr local (n1*) = Fr cr1 ;

[0193] where n1* is the critical opening of the jet-dominated zone and the transition zone, i.e. the first critical opening; Fr local (n1*) is the local Froude number at the first critical opening, Fr cr1 is the first critical Froude number.

[0194] The second critical opening n2* of the flow regime transition is determined by solving the following equation:

[0195] Fr local (n2*) = Fr cr2 ;

[0196] where n2* is the critical opening of the transition zone and the pressure-difference dominant zone, i.e., the second critical opening; Fr local (n2*) is the local Froude number at the second critical opening, Fr cr2 is the second critical Froude number. Since Fr local (n) is a monotonically decreasing function of n, i.e., the larger the opening, the lower the jet velocity and the smaller the Froude number, the above equation has a unique solution in physics. The bisection method or Newton iteration method can be used to solve the critical opening.

[0197] For the interval less than or equal to the first critical opening, a jet-dominant model based on the jet contraction coefficient is used; for the interval greater than or equal to the second critical opening, a quasi-static pressure-difference dominant model is used; for the transition interval between the two critical openings, a smooth transition function is used to weight the two models, obtaining a continuous and derivable segmented coupling model in the full opening interval.

[0198] In this embodiment, a segmented coupling model is established. According to the flow state recognition result, a targeted physical model is used for different intervals. For the jet-dominant zone, i.e., the interval where the relative opening n satisfies 0 < n ≤ n1*, a complete model based on the jet contraction theory is used:

[0199] F d_I (n) = p x g x h x B x d x K1 x n d (n);

[0200] where F d_I (n) is the bottom hydrodynamic force of the jet-dominant zone, p is the density of water, g is the acceleration of gravity, h is the water head in front of the gate, B is the gate width, d is the gate thickness, K1 is the jet-dominant zone coefficient, and n is the relative opening. d (n) is the dimensionless hydrodynamic force coefficient.

[0201] For the pressure-difference dominant zone, i.e., the interval where the relative opening n satisfies n ≥ n2*, when the opening is large, the jet contraction effect basically disappears, and the hydrodynamic force is mainly determined by the static pressure difference before and after the gate. At this time, a simplified quasi-static model can be used, and the expression of the pressure-difference dominant model is:

[0202] F d_III (n) = p x g x h x B x d x K3 x (1-n) m ;

[0203] where F d_III (n) is the bottom hydrodynamic force of the pressure-difference dominant zone, p is the density of water, g is the acceleration of gravity, h is the water head in front of the gate, B is the gate width, d is the gate thickness, K3 is the pressure-difference dominant zone coefficient, and n is the relative opening.

[0204] For the transition zone, i.e. the interval where the relative opening n satisfies n1*<n<n2*, a differentiable smooth transition function is used to realize the mixing of the two models:

[0205] F d_II (n) = σ(n)xF d_I (n) + [1-σ(n)]xF d_III (n);

[0206] where F d_II (n) is the bottom hydrodynamic force of the transition zone, σ(n) is the smooth transition function, F d_I (n) is the calculated value of the jet-dominant zone model, and F d_III (n) is the calculated value of the pressure-dominant zone model. The smooth transition function σ(n) adopts a hyperbolic tangent function form:

[0207] σ(n) = (1 / 2) × [1-tanh((n-n m ) / Δn)];

[0208] where σ(n) is the transition weight at the relative opening n; tanh is the hyperbolic tangent function, n is the relative opening, n m is the center opening of the transition zone, and Δn is the transition bandwidth parameter.

[0209] The center opening of the transition zone and the transition bandwidth parameter are determined according to the first critical opening and the second critical opening:

[0210] n m = (n1*+n2*) / 2;

[0211] where n m is the center opening of the transition zone, n1* is the first critical opening, and n2* is the second critical opening.

[0212] Δn = (n2*-n1*) / 4;

[0213] where Δn is the transition bandwidth parameter. The smooth function satisfies the following properties: when n tends to n1*, σ(n) tends to 1, and the model output tends to the jet model; when n tends to n2*, σ(n) tends to 0, and the model output tends to the pressure model; σ(n) is a continuous and differentiable function, ensuring the continuity and differentiability of the overall model.

[0214] Further, after using the piecewise coupling model, the extreme value solving process needs to be adjusted accordingly. The definition of the overall hydrodynamic force function F d (n) in each interval is as follows:

[0215] F d (n) = F d_I (n), when 0<n≤n1*;

[0216] F d (n)=F d_II (n), when n1* < n < n2*;

[0217] F d (n)=F d_III (n), when n ≥ n2*.

[0218] Since the piecewise model is continuously differentiable within each interval and continuous at the interval boundaries, analytical differentiation or adaptive search can still be used to find the extrema. The candidate points for extrema include: the stationary point in the jet-dominated region where the derivative is zero, the stationary point in the transition region where the derivative is zero, the stationary point in the pressure-dominated region where the derivative is zero, the function value at the boundary point n equal to 0, the function value at the boundary point n equal to 1. It should be noted that the function behavior near the flow regime transition critical openings n1* and n2* should be checked, because a sharp change in the hydrodynamic characteristics can occur at the flow regime transition.

[0219] In a detailed numerical example, assume that the orifice discharge coefficient μ is equal to 0.62, the upstream water head h is equal to 10 meters, the downstream water depth h d is equal to 3 meters, and the gravitational acceleration g is equal to 9.81 meters per second squared. The jet velocity and the local Froude number are calculated for each opening. At a relative opening n equal to 0.05: V jet (0.05) = 0.62 x sqrt(2 x 9.81 x 10) / (0.05 x 0.917) = 0.62 x 14.01 / 0.0459 = 189.3 meters per second; Fr local (0.05) = 189.3 / sqrt(9.81 x 3) = 189.3 / 5.42 = 34.9; this Froude number is much larger than Fr cr1 (2.5) and belongs to the jet-dominated region. At a relative opening n equal to 0.3, assume that C c (0.3) is equal to 0.985: V jet (0.3) = 0.62 x 14.01 / (0.3 x 0.985) = 8.69 / 0.296 = 29.4 meters per second; Fr local (0.3) = 29.4 / 5.42 = 5.4; this Froude number is larger than Fr cr1 (2.5) and still belongs to the jet-dominated region. At a relative opening n equal to 0.6, assume that C c (0.6) is equal to 0.998: V jet (0.6) = 0.62 x 14.01 / (0.6 x 0.998) = 8.69 / 0.599 = 14.5 meters per second; Fr local (0.6) = 14.5 / 5.42 = 2.7; this Froude number is slightly larger than Fr cr1(2.5), located near the boundary between the jet-dominant region and the transition region. Interpolation yields the first critical aperture n1* to be approximately 0.62, and the second critical aperture n2* to be approximately 0.85. The center aperture n of the transition region... m The value is 0.735, and the transition band width parameter Δn is 0.0575.

[0220] According to another aspect of this application, the method for predicting the opening and closing capacity of a planar gate valve based on parametric calculation can also be:

[0221] S1: Obtain the total gravity G of the gate valve system based on the self-weight of the plane gate valve, its counterweight, and the self-weight of the lifting rod. Z .

[0222] Specifically, the total gravity G of the gate valve system Z The weight should include the weight of the gate valve itself (G1) and its counterweight (G2), as well as the weight of the gate valve's lifting rod system (G3). Z = G1 + G2 + G3.

[0223] S2: Obtain the pressure F at the top of the gate valve based on the water level in the gate well before the gate valve opens. u .

[0224] In this embodiment, the pressure F at the top of the gate valve u It can be obtained by the following formula: F u =ρg(Z j0 - Z d0 Bd; where ρ is the density of water, in kg / m³ 3 g is the acceleration due to gravity, with units of m / s². 2 Z j0 Z represents the initial water level of the valve well before operation, in meters (m). d0 B is the top elevation of the valve in the fully closed state, in meters; B is the width of the flat gate valve, in meters; d is the thickness of the flat gate valve, in meters.

[0225] S3: Based on the bottom flange form of the flat gate valve, select the corresponding dynamic water force prediction formula for different bottom flange shapes, and calculate the downward suction force or upward support force F at the bottom of the valve under each typical opening degree. d Extract the worst-case value F dmin .

[0226] Specifically, the force F of the bottom water flow in a gate valve under different opening conditions d The most unfavorable hydrodynamic force F is obtained from calculations based on hydraulic and geometric parameters. dmin Here, it is defined that the force of dynamic water is positive when it is upward and negative when it is downward;

[0227] Top slant bottom edge: F d =(3.91n2 -0.0004a 2 +0.0022na-4.57n+0.053a-0.05)pg h Bd+F f ;

[0228] Dip angle bottom edge: F d =(1.83n 2 -0.000011a 2 +0.000034na-2.06n+0.0009a+0.093)pg h Bd+F f ;

[0229] Wherein n is the gate opening of the flat gate valve, i.e. the ratio of the height of the flow area to the height of the water conveying corridor, n = 1 when the gate is fully open; a is the bottom edge inclination angle; h is the action water head, unit: m; F f is the buoyancy of the flat gate valve.

[0230] S4: Obtain the total friction force f z according to the relevant specification table calculation.

[0231] S5: Obtain the maximum opening force T of the flat gate valve according to the force balance principle.

[0232] In this embodiment, the maximum opening force T of the flat gate valve is calculated as follows: T = G Z +F u -F dmin +f z .

[0233] S6: According to the maximum opening force of the flat gate valve, round up to the standard series to select the predicted capacity of the hoist.

[0234] The prediction method of the embodiment takes into account the comprehensive influence of the key parameters such as the weight, hydraulic parameters, geometric parameters of the flat gate valve and its suspender system on the capacity of the flat gate valve hoist, wherein the hydraulic parameters include the action water head and the flow velocity below the gate, the geometric parameters include the basic dimensions such as the length, width and height of the flat gate valve, and also include the detailed structural dimensions of the bottom edge. The high-precision bottom edge hydrodynamic load is taken as the core algorithm and embedded into the technical system to accurately design the valve and select the capacity of the hoist.

[0235] According to another aspect of this application, a planar gate valve opening and closing capacity prediction system based on parametric calculation is also provided, comprising: a planar gate valve basic data acquisition unit, used to acquire the operating water level and elevation of the project, the self-weight and counterweight of the planar gate valve, structural dimensions, and design data including waterstop and roller materials; a planar gate valve and hanger system total weight calculation unit; a planar gate valve top water pressure calculation unit; a planar gate valve bottom dynamic water force calculation unit; a planar gate valve total friction force calculation unit; a planar gate valve maximum opening force calculation unit; and a hoist capacity selection prediction processing unit.

[0236] In a detailed embodiment, a lock project is designed with a maximum head of 17.8m. The water conveyance system uses planar gate valves with a sill elevation of -6.65m and gate leaf dimensions of 5.46m × 5.22m. The gate thickness is 0.92m, and the waterstop dimensions are 4.42m × 5.12m. The valve bottom edge is designed as a sharp-flange type, with the bottom edge inclined upstream at a 42° angle to the horizontal plane. The total gravity G of the gate valve system is obtained based on the self-weight and counterweight of the planar gate valve, and the weight of the lifting rod. Z G Z = G1 + G2 + G3 = 211.8 + 0 + 50 = 261.8 kN; where G1 is the self-weight of the flat gate valve, which is 211.8 kN; G2 is the counterweight, which is 0 kN; and G3 is the weight of the lifting rod, which is 50 kN. The pressure F at the top of the gate valve is obtained based on the water level in the gate well before the flat gate valve opens. u ;F u =ρg(Z j0 - Z d0 Bd = 839 kN; where ρ is the density of water, which is 1000 kg / m³. 3 g is the acceleration due to gravity, which is 9.81 m / s². 2 Z j0 The initial water level in the valve well before operation was 20.6m; Z d0 B represents the top elevation of the valve in its fully closed state, which is -0.52m; B is the width of the flat gate valve, which is 4.4m; and d is the thickness of the flat gate valve, which is 0.92m. The bottom edge of the flat gate valve slopes upstream. Using the hydrodynamic force prediction formula, the downward suction force or upward support force F acting on the bottom of the valve at various typical opening degrees is calculated. d Extract the worst-case value F dmin ;F d =(3.91n 2 -0.0004α 2 +0.0022 nα-4.57n+0.053α-0.05)ρghBd+F f ;F dmin= 274.20 kN. The calculated bottom edge hydrodynamic force is as follows: when the flat gate valve opening n is 0.10, the dimensionless correction coefficient k of the bottom edge hydrodynamic force is 1.03, and the bottom edge hydrodynamic load is 890.14 kN; when the flat gate valve opening n is 0.20, the dimensionless correction coefficient k of the bottom edge hydrodynamic force is 0.70, and the bottom edge hydrodynamic load is 656.46 kN; when the flat gate valve opening n is 0.30, the dimensionless correction coefficient k of the bottom edge hydrodynamic force is 0.45, and the bottom edge hydrodynamic load is 478.02 kN; when the flat gate valve opening n is 0.40, the dimensionless correction coefficient k of the bottom edge hydrodynamic force is 0.28, and the bottom edge hydrodynamic load is 354.84 kN; when the flat gate valve opening n is 0.50, the dimensionless correction coefficient k of the bottom edge hydrodynamic force is 0.18, and the bottom edge hydrodynamic load is 286.90 kN; when the flat gate valve opening n is 0.60, the dimensionless correction coefficient k of the bottom edge hydrodynamic force is 0.16, and the bottom edge hydrodynamic load is 274.20 kN; when the flat gate valve opening n is 0.70, the dimensionless correction coefficient k of the bottom edge hydrodynamic force is 0.22, and the bottom edge hydrodynamic load is 316.75 kN; when the flat gate valve opening n is 0.80, the dimensionless correction coefficient k of the bottom edge hydrodynamic force is 0.36, and the bottom edge hydrodynamic load is 414.55 kN. Wherein, α is the bottom edge inclination, which is 42°; h is the designed maximum water head, which is 17.8 m; F f is the flat gate valve buoyancy, which is 160 kN; the dimensionless correction coefficient k of the bottom edge hydrodynamic force is 3.91n 2 -0.0004α 2 +0.0022 nα-4.57n+0.053α-0.05. According to the relevant specification table, the sliding friction resistance and rolling friction resistance coefficients are obtained, and the total friction force f z is calculated. z ; f zs = μ1 K1 K2 ρgh (2H zs +B zs ) Δ + P / R (μ2r + l) = 26.3 + 92 = 118.3 kN; wherein, μ1 is the rubber to stainless steel water sealing friction coefficient, which is 0.2; K1 is the rubber water sealing friction force correction coefficient, which is 1.8; K2 is the rubber water sealing width correction coefficient, which is 0.6; H zs is the valve water sealing height, which is 3.95 m; B Z is the valve water sealing width, which is 3.74 m; Δ is the rubber water sealing water pressure action width, which is 0.6 m; P is the horizontal pressure, which is 3609.5 kN; R is the roller radius, which is 450 mm; μ2 is the bearing sliding friction coefficient, which is 0.14; r is the roller shaft radius, which is 75 mm; l is the rolling friction force arm, which is 1 mm. According to the force balance principle, the maximum gate opening force T of the flat gate valve is obtained; T = Gu -F dmin +f z =261.8 + 839 - 274.20 + 118.3 = 944.9 kN. Based on the maximum opening force of the plane gate valve, rounding up to the standard series gives the predicted capacity of the hoist. The hoist capacity for the water supply valves in this lock project can be selected as 1000 kN.

[0237] like Figures 5 to 7 As shown, the pressure change curve during valve opening in engineering is illustrated, and the water level and pressure at each opening degree are extracted. The opening and closing forces at each opening degree are calculated using the method proposed in this embodiment, and the results match the calculated results of the measured opening and closing forces, demonstrating the reliability of the prediction method. According to... Figure 8 The measured maximum valve opening force is 850 kN, while the predicted maximum opening force in this embodiment is 944.9 kN. The hoist capacity selected by rounding upwards is 1000 kN. Considering the effective operating efficiency of the hoist, the hoist capacity selection is reasonable and has a certain safety margin. Simultaneously, the prediction of the hydrodynamic force at the valve's bottom edge in this embodiment is calculated using the measured water level process line during the opening of the lock project's spillway valve. This calculation is then compared with the empirical values ​​recommended in current lock valve design specifications. It can be seen that the hydrodynamic force at the bottom edge in this embodiment matches the trend of the measured process item, demonstrating high prediction accuracy.

[0238] This invention constructs a continuous function model containing rational fractional terms to accurately describe the complete and continuous curve of bottom hydrodynamic force as a function of gate opening. The problem of finding the most unfavorable operating condition is transformed from comparing discrete points to solving a global optimization problem of the continuous function. By employing mathematical optimization algorithms such as analytical differentiation or adaptive search, the true global most unfavorable point can be accurately located throughout the entire operating range, eliminating the risk of missed extreme values ​​due to interpolation or sparse data points, thus improving prediction accuracy. Detailed geometric features such as bottom edge inclination, chamfers, and fillets are vectorized, and a nonlinear mapping relationship is established between these geometric features and key parameters of the hydrodynamic model. This allows the model to customize the optimal hydrodynamic curve for any given gate geometry, accurately reflecting the impact of subtle structural changes on hydrodynamics, and improving the model's generalization ability and applicability. Uncertainty quantification is introduced; by employing techniques such as Gaussian process regression, the model can output not only the predicted mean but also the variance of the predicted values ​​simultaneously. Building upon this, an optimization is performed by constructing a conservative upper bound objective function that incorporates the prediction variance, ensuring that the final decision incorporates consideration of model uncertainty. This provides designers with a tool to quantify risk, enabling more robust and reliable gate capacity decisions based on the project's safety level.

[0239] The preferred embodiments of the present application are described in detail above, but the present application is not limited to the specific details of the above-described embodiments, and various equivalent transformations of the technical solutions of the present application can be made within the technical concept of the present application, and these equivalent transformations all belong to the protection scope of the present application.

Claims

1. A method for predicting opening and closing capacity of a flat gate valve based on parameterized calculation, characterized in that, The method comprises: acquiring a set of geometric parameters and a set of hydraulic state parameters of a flat gate valve to be analyzed; calculating total gravity of the gate valve system based on the set of geometric parameters, and calculating top vertical water pressure based on the set of geometric parameters and the set of hydraulic state parameters; in a preset operating opening interval, performing extreme value solving on a preconfigured parameterized mapping model representing the corresponding relationship between the relative opening of the flat gate valve and the bottom hydrodynamic force, to determine the most unfavorable extreme value of the bottom hydrodynamic force; calculating the maximum opening force according to the force balance principle based on the total gravity of the gate valve system, the top vertical water pressure, the most unfavorable extreme value, and the total friction force determined by the set of hydraulic state parameters; determining the predicted capacity of the gate hoist based on the maximum opening force.

2. The method of claim 1, wherein, The total gravity of the gate valve system comprises the self-weight of the flat gate valve, the counterweight, and the weight of the gate valve suspender system; The total gravity of the gate valve system is calculated based on the set of geometric parameters, specifically by using the following formula: G Z = G1+ G2+ G3; where G Z G1 is the weight of the flat gate valve, G2 is the counterweight, and G3 is the weight of the gate valve hoist system.

3. The method of claim 1, wherein, The top vertical water pressure is calculated by using the following formula: F u = p x g x (Z j0 -Z d0 ) x B x d; where F u is the top vertical water pressure, p is the density of water, g is the acceleration of gravity, Z j0 is the initial water level in the valve shaft before operation, Z d0 is the top elevation in the fully closed state of the valve, B is the width of the flat gate valve, and d is the thickness of the flat gate valve.

4. The method of claim 1, wherein, The parameterized mapping model defines that the bottom hydrodynamic force is positive upward and negative downward, and selects the calculation logic based on the bottom edge inclination type defined in the set of geometric parameters: When the bottom edge is an upwardly inclined angle bottom edge: F d (n) = p x g x h x B x d x k up (n, a); When the bottom edge is a downwardly inclined angle bottom edge: F d (n) = p x g x h x B x d x k down (n, a); where F d (n) is the bottom hydrodynamic force at the relative opening n, p is the density of water, g is the acceleration of gravity, B is the flat gate width, d is the flat gate thickness, h is the action water head, a is the bottom edge inclination, k up (n, a) and k down (n, a) are the hydrodynamic force coefficients at the corresponding opening and inclination, respectively.

5. The method of claim 1, wherein, The total friction force comprises the water stop friction resistance and the roller friction resistance; The maximum opening force is calculated according to the force balance principle, specifically by using the following formula: T = G Z + F u - F dmin + f z ; Where T is the maximum gate opening force, G Z is the total gate system weight force, F u is the top vertical water pressure force, F dmin is the most unfavorable extreme, f z is the total friction force.

6. The method of claim 1, wherein, The parameterized mapping model is constructed by using a continuous function model containing fractional rational terms to represent the hydraulic surge characteristics at small openings, and the continuous function model is expressed as: F d (n)=C hyd ×[(a0+a1×n) / (n+b)+a2×n] where n is the relative opening, F d (n) is the bottom hydrodynamic force, C hyd is the water head dependent coefficient term, a0, a1, a2 and b are the undetermined model parameters obtained by identification, where the term (a0+a1xn) / (n+b) is used to represent the non-linear surge behavior in the small opening interval, and the a2xn term is used to represent the linear variation trend in the large opening interval.

7. The method of claim 6, wherein, The to-be-determined model parameters are associated with the bottom edge detail structure features defined in the set of geometric parameters; The determination of the to-be-determined model parameters comprises: Constructing the bottom edge detail feature vector X g , the bottom edge detail feature vector X g At least including the bottom edge inclination angle, the bottom edge chamfer length, the bottom edge fillet radius, the bottom edge gap and the water stop position category; Constructing the hydraulic state vector X h , the hydraulic state vector X h at least includes the acting water head and the flow velocity downstream the lock. The bottom edge detail feature vector X g and the hydraulic state vector X h The regression mapping relationship with the undetermined model parameters as the output and the bottom edge detail feature vector X and the hydraulic state vector X as the input is determined to determine the specific values of a0, a1, a2, and b.

8. The method of claim 6, wherein, The method further comprises an online calibration step, specifically: During the engineering trial operation or debugging of the flat gate valve, measured opening and closing force data at a predetermined group of measured openings are acquired; Based on the measured opening and closing force data and the force balance principle, the measured hydrodynamic force coefficients at the corresponding openings are inversely obtained; The deviation between the measured hydrodynamic force coefficients and the output values of the parameterized mapping model is calculated to construct a deviation calibration term; The parameterized mapping model is modified online by using the deviation calibration term to obtain calibrated bottom hydrodynamic force prediction values.

9. The method of claim 1, wherein, The parameterized mapping model is a continuous function with respect to the relative opening; The most unfavorable extreme value of the bottom hydrodynamic force is determined by: regarding the bottom hydrodynamic force as a continuous objective function of the relative opening; performing global optimization operation on the continuous objective function within the operating opening interval to identify the most unfavorable opening that makes the continuous objective function reach an extreme value; substituting the most unfavorable opening into the continuous objective function to calculate the most unfavorable extreme value.

10. The method of claim 9, wherein, The most unfavorable opening that makes the continuous objective function reach an extreme value is identified by: taking the first derivative of the continuous objective function with respect to the relative opening to obtain a derivative function; solving the stationary point of the derivative function equal to zero and verifying whether the stationary point is located within the operating opening interval; comparing the function values at the stationary point and the boundary points of the operating opening interval to determine the most unfavorable opening.

Images