Design-optimization-manufacturing integration-based water pump management and control system and method

Abstract

The invention provides a water pump management and control system and method based on design-optimization-manufacturing integration, and the system comprises a data collection module which is used for collecting and preprocessing water pump design related parameters and actual production parameters; the three-dimensional model construction module is used for constructing a water pump impeller three-dimensional model by adopting a machine learning and optimization algorithm; the flow field reconstruction module is used for constructing a water pump impeller flow field graph by adopting an SST turbulence model; the digital twinborn model construction module is used for constructing and optimizing a water pump impeller process flow digital twinborn model to obtain optimal process parameters of the water pump impeller; and the management and control module is used for obtaining a water pump impeller structure meeting preset requirements based on the water pump impeller three-dimensional model, the water pump impeller flow field graph and the optimal process parameters of the water pump impeller, and completing management and control of the water pump based on design-optimization-manufacturing integration. According to the technical scheme, the matching degree of the designed impeller and the final impeller structure is high, and the precision and efficiency of the finally machined and manufactured impeller are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent optimization of water pumps, and particularly relates to a water pump control system and method based on design-optimization-manufacturing integration. Background Art

[0002] Water pumps are general-purpose machines with a large quantity, wide application, and high energy consumption in the national economic fields such as aerospace, industry, and medical treatment, and are used for fluid pressurized transportation. According to statistics, the power consumption of water pumps accounts for about 17% of the total power consumption in the country. Improving the pump efficiency is an important way to achieve energy conservation and emission reduction. The theory and technology of intelligent manufacturing of water pumps are beneficial to improving the operation efficiency and reliability of water pumps. In the prior art, there are many problems in the whole process of "design-optimization-manufacturing" of water pumps, such as high calculation cost, long optimization cycle, and imperfect manufacturing process. Summary of the Invention

[0003] In order to solve the problems existing in the prior art, the present invention provides a water pump control system and method based on design-optimization-manufacturing integration to improve the operation efficiency and reliability of water pumps.

[0004] To achieve the above object, the present invention provides the following solutions:

[0005] A water pump control system based on design-optimization-manufacturing integration, the system includes:

[0006] A data acquisition module, configured to collect and preprocess water pump design-related parameters and actual production parameters;

[0007] A three-dimensional model construction module, configured to construct a three-dimensional model of a water pump impeller based on the preprocessed water pump design-related parameters by using machine learning and optimization algorithms;

[0008] A flow field reconstruction module, configured to construct a flow field diagram of a water pump impeller by using the SST turbulence model based on the preprocessed water pump design-related parameters;

[0009] A digital twin model construction module, configured to construct and optimize a digital twin model of a water pump impeller process flow based on the actual production data to obtain the optimal process parameters of the water pump impeller;

[0010] A control module, configured to obtain a water pump impeller structure meeting preset requirements based on the three-dimensional model of the water pump impeller, the flow field diagram of the water pump impeller, and the optimal process parameters of the water pump impeller, and complete the control of the water pump based on design-optimization-manufacturing integration.

[0011] Preferably, in the data acquisition module, the water pump design-related parameters include: geometric parameters of water pump flow-through components, multi-condition efficiency of the water pump, pressure pulsation, and structural stress.

[0012] Preferably, the three-dimensional model building module includes:

[0013] A mathematical model building unit, used to establish a nonlinear mathematical relationship between the geometric parameters of the flow-through components of the water pump and the multi-operating efficiency of the water pump, the pressure pulsation, and the structural stress by using the Navier-Stokes equation and the flow continuity equation, and to build a nonlinear mathematical model based on boundary conditions through the nonlinear mathematical relationship;

[0014] A neural network model building unit, used to build an artificial neural network model of the hydraulic and structural performance of the water pump based on the nonlinear mathematical model;

[0015] The model optimization unit is used to optimize the artificial neural network model based on a multi-objective particle swarm algorithm to obtain the three-dimensional model of the water pump impeller.

[0016] Preferably, the flow field reconstruction module includes:

[0017] Numerical simulation unit, used to simulate the flow field of the pump impeller under adverse pressure gradient based on the two-equation SST turbulence model;

[0018] A pressure distribution acquisition unit, used to extract the characteristics of the water pump impeller flow field based on a machine learning algorithm, and to establish a mapping relationship between boundary conditions and numerical simulation data to train a proxy model, so as to predict the pressure distribution inside the water pump;

[0019] A flow field map acquisition unit is used to obtain the water pump impeller flow field map based on the internal pressure distribution of the water pump.

[0020] Preferably, the two-equation SST turbulence model includes two models: a low Reynolds number model k-ω and a high Reynolds number model k-ε; wherein the low Reynolds number model k-ω is used for numerical solution in the boundary layer region, and the high Reynolds number model k-ε is used for numerical solution in the free shear layer.

[0021] Preferably, the digital twin model building module includes:

[0022] A twin model building unit is used to build a digital twin model of the water pump impeller process based on the water pump impeller process design documents, equipment drawings and actual production data;

[0023] Optimization space acquisition unit, used to mine the potential relationship and optimization space between process parameters based on the historical data of the water pump impeller production process;

[0024] The industrial parameter acquisition unit is used to perform virtual experiments and optimization simulations in the digital twin model of the water pump impeller process flow based on the potential relationship and optimization space between the process parameters to obtain the optimal process parameters of the water pump impeller.

[0025] The present invention also provides a water pump control method based on integrated design-optimization-manufacturing. Applying the system, the method includes:

[0026] Collect and preprocess the parameters related to water pump design and the actual production parameters;

[0027] Based on the preprocessed parameters related to water pump design, construct a three-dimensional model of the water pump impeller using machine learning and optimization algorithms;

[0028] Based on the preprocessed parameters related to water pump design, construct a flow field diagram of the water pump impeller using the SST turbulence model;

[0029] Based on the actual production data, construct and optimize a digital twin model of the water pump impeller process flow to obtain the optimal process parameters of the water pump impeller;

[0030] Based on the three-dimensional model of the water pump impeller, the flow field diagram of the water pump impeller, and the optimal process parameters of the water pump impeller, obtain a water pump impeller structure that meets the preset requirements, and complete the control of the water pump based on integrated design-optimization-manufacturing.

[0031] Preferably, the parameters related to water pump design include: geometric parameters of the water pump flow-through components, multi-condition efficiency of the water pump, pressure pulsation, and structural stress.

[0032] Preferably, the method for obtaining the three-dimensional model of the water pump impeller includes:

[0033] Adopt the Navier-Stokes equation and the flow continuity equation to establish a non-linear mathematical relationship between the geometric parameters of the water pump flow-through components and the multi-condition efficiency of the water pump, the pressure pulsation, and the structural stress, and construct a non-linear mathematical model based on boundary conditions through the non-linear mathematical relationship;

[0034] Based on the non-linear mathematical model, construct an artificial neural network model for the hydraulic and structural performance of the water pump;

[0035] Optimize the artificial neural network model based on the multi-objective particle swarm algorithm to obtain the three-dimensional model of the water pump impeller.

[0036] Preferably, the method for obtaining the flow field diagram of the water pump impeller includes:

[0037] Based on the two-equation SST turbulence model, simulate the flow field of the water pump impeller under an adverse pressure gradient;

[0038] Based on the machine learning algorithm, extract the characteristics of the flow field of the water pump impeller, establish a mapping relationship between the boundary conditions and the numerical simulation data to train a surrogate model, and predict the internal pressure distribution of the water pump;

[0039] Based on the internal pressure distribution of the water pump, obtain the flow field diagram of the water pump impeller.

[0040] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention aims at the problems of high calculation cost, long optimization cycle and imperfect manufacturing process in the whole process of "design - optimization - manufacturing" of water pumps. It organically combines the theories, methods and technologies of multiple disciplines such as fluid mechanics and structural mechanics, proposes a multi - objective optimization design method for water pump performance based on artificial neural network and particle swarm algorithm, proposes a high - precision and fast solution method for water pump performance based on the modification of SST turbulence model, and establishes an accurate modeling method for the intelligent manufacturing process of water pump impellers based on digital twin. Through the system and method of the present invention, the designed impeller is highly matched with the final impeller structure after intelligent optimization, and the accuracy and efficiency of the finally machined impeller are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the present invention, the accompanying drawings required for use in the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0042] Figure 1 It is a schematic structural diagram of a water pump control system based on the integration of design - optimization - manufacturing according to an embodiment of the present invention;

[0043] Figure 2 It is a three - dimensional model diagram of an impeller according to an embodiment of the present invention;

[0044] Figure 3 It is a flow field diagram of an impeller according to an embodiment of the present invention;

[0045] Figure 4 It is a construction flow chart of a digital twin model of a process flow according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the present invention belong to the scope of protection of the present invention.

[0047] In order to make the above - mentioned objects, features and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0048] Embodiment 1

[0049] As Figure 1As shown in the figure, a water pump control system based on the integration of design-optimization-manufacturing, the system includes: a data acquisition module, a three-dimensional model construction module, a flow field reconstruction module, a digital twin model construction module, and a control module.

[0050] The data acquisition module is used to collect and preprocess the water pump design-related parameters and actual production parameters; in a further implementation manner, in the data acquisition module, the water pump design-related parameters include: the geometric parameters of the water pump's flow-through components, the multi-condition efficiency of the water pump, the pressure pulsation, and the structural stress.

[0051] The three-dimensional model construction module is used to construct a three-dimensional model of the water pump impeller based on the preprocessed water pump design-related parameters by using machine learning and optimization algorithms;

[0052] In a further implementation manner, the three-dimensional model construction module includes:

[0053] The mathematical model construction unit is used to establish a non-linear mathematical relationship between the geometric parameters of the water pump's flow-through components and the multi-condition efficiency, pressure pulsation, and structural stress of the water pump by using the Navier-Stokes equation and the flow continuity equation, and construct a non-linear mathematical model based on the boundary conditions through the non-linear mathematical relationship; in this embodiment, for an incompressible fluid, that is, the case where the density is a constant, the Navier-Stokes equation can be simplified as:

[0054]

[0055] where is the kinematic viscosity, p represents the pressure, and ρ represents the fluid density.

[0056] For an incompressible fluid, the continuity equation is simplified to the volume continuity equation, that is, the divergence of the velocity field is equal to zero, indicating that the local volume change rate is zero: ▽·u = 0, which means that the fluid will not be compressed or expanded at any point.

[0057] The neural network model construction unit is used to construct an artificial neural network model of the water pump's hydraulic and structural performance based on the non-linear mathematical model; in this embodiment, a genetic algorithm introducing an adaptive coding mechanism is used as the learning algorithm of the RBF network to train the artificial neural network model, further improving the network prediction accuracy. The specific steps are to divide the network structure optimization and parameter learning into two stages: training and evolution. First, randomly generate N individuals to form a population, and then learn the center c in the network corresponding to the number of hidden nodes chromosome in each individual i , the width parameter σ iAnd linear weights. When optimizing the number of hidden nodes using the genetic evolution algorithm, through the alternating execution of these two processes, an RBF network with basis functions that meet the minimum error requirement and have different width parameters is obtained. The genetic algorithm GA is used to solve the problem of optimizing the RBF network structure. The fitness function in GA is constructed by the following formula, that is, constructing an energy function so that its minimum value corresponds to the optimal solution of the problem:

[0058]

[0059] where m is the number of training samples; k is the number of neurons in the output layer; t mk is the expected output of the k-th neuron for the m-th sample; o mk is the RBF network output of the k-th neuron for the m-th sample. Finally, the obtained impeller design model (3D impeller model) is verified and optimized to determine the accuracy of prediction and the efficiency of design;

[0060] The model optimization unit is used to optimize the artificial neural network model based on the multi-objective particle swarm algorithm to obtain the 3D model of the water pump impeller. As Figure 2 shown, the 3D model diagram of the impeller designed through machine learning and intelligent optimization algorithms.

[0061] The flow field reconstruction module is used to construct the flow field diagram of the water pump impeller using the SST turbulence model based on the preprocessed relevant parameters of the water pump design; A further implementation manner lies in that the flow field reconstruction module includes:

[0062] The numerical simulation unit is used to simulate the flow field of the water pump impeller under the adverse pressure gradient based on the two-equation SST turbulence model; Specifically, study the high-dimensional and non-linear mapping relationship between the input parameters of the SST turbulence model and the internal pressure of the water pump, and establish a high-precision SST turbulence surrogate model (two-equation SST turbulence model) of the water pump.

[0063] A further implementation manner lies in that the two-equation SST turbulence model includes two models: the low Reynolds number model k-ω and the high Reynolds number model k-ε; Among them, the low Reynolds number model k-ω is used for numerical solution in the boundary layer region, and the high Reynolds number model k-ε is used for numerical solution in the free shear layer.

[0064] In this embodiment, a data-driven high-precision and fast numerical calculation of the water pump is performed, using the two-equation SST turbulence model, and its mathematical expression is:

[0065]

[0066] where, P k is the generation term of the turbulent kinetic energy; μ and μ t are the viscous stress and the turbulent viscous stress respectively; σ ∈is an empirical coefficient, usually taking a value of 1.3; a i and a j represent spatial coordinates. ρ represents the fluid density, k represents the turbulent kinetic energy, and u i represents the velocity component, and C ∈ is an empirical coefficient, usually taking a value of 1.44, and ∈ is the dissipation rate.

[0067] The SST model combines the k-ω and k-ε models. These two models are numerically solved in different regions through a blending function. In the boundary layer region, the low Reynolds number model k-ω is used for calculation; while in the free shear layer, the high Reynolds number model k-ε with good adaptability is used for calculation. The SST model combines the advantages of the two models, enabling it to accurately simulate the flow separation under an adverse pressure gradient and having good performance for predicting complex flow fields.

[0068] The pressure distribution acquisition unit is used to extract the characteristics of the flow field of the water pump impeller based on a machine learning algorithm (neural network), and establish a mapping relationship between the boundary conditions and the numerical simulation data in a low-dimensional space to train a surrogate model to predict the internal pressure distribution of the water pump; in this embodiment, 1000 trainings are required, and the error metrics used include the normalized RMSE, normalized BIAS, and normalized STDE from the predictions of the ML surrogate model, and these data are based on the CFD simulation results of multiple time frames as references. The training set here consists of two-dimensional cylinders (N1 dataset) using the Mu model, and finally the best-performing parameters are selected among all models.

[0069] The training process of the neural network can be divided into 3 steps: forward propagation, loss function calculation, and backpropagation. At the output layer, the error between the predicted output value obtained by forward propagation and the actual value is used to measure the prediction effect. The error value depends on the loss function. Once the loss function is calculated, it will be backpropagated. In the current work, the most commonly used loss function is the root mean square error, which is defined as follows:

[0070]

[0071] where n k is the number of outputs, represents the predicted value of the i-th sample, represents the true value of the i-th sample, and δ i represents the difference between the predicted value and the true value of the i-th sample. During the backpropagation process, the gradient descent algorithm is used for parameter update.

[0072] Compare the performance of different feature extraction methods and different surrogate models, evaluate the accuracy and computational efficiency of the reconstructed flow field, and optimize the selection of feature extraction and surrogate models according to the evaluation results.

[0073] A flow field map acquisition unit for obtaining a flow field map of the pump impeller based on the internal pressure distribution of the pump. As Figure 3 shown, the flow field map of the impeller obtained through high-precision and fast numerical calculation of the pump driven by data.

[0074] A digital twin model construction module for constructing and optimizing a digital twin model of the technological process of the pump impeller based on actual production data, and obtaining the optimal process parameters of the pump impeller; as Figure 4 shown, it is a step diagram of constructing a digital twin model of the technological process based on process design documents, equipment drawings and actual production data. First is the design input stage, where the user inputs their requirements; then a series of product configuration designs and current flow component designs are carried out; among them, the product configuration design stage includes converting user requirements, selecting products according to the converted user requirements, making product configurations according to the product selection results and obtaining product configuration requirements; the product current flow component design stage includes impeller design and volute design, and the impeller design includes global parameter setting, impeller parameter setting, axis diagram design, blade design, streamline design, blade encryption and inlet chamfering. The volute design includes section design and diffuser section design. Finally, the design output is carried out. Figure 4 The flow chart on the right details the design processes of the impeller and the volute.

[0075] A further implementation manner is that the digital twin model construction module includes:

[0076] A twin model construction unit for constructing a digital twin model of the technological process of the pump impeller based on the process design document, equipment drawing and actual production data of the pump impeller;

[0077] An optimization space acquisition unit for mining the potential relationships and optimization space between process parameters based on the historical data of the pump impeller production process;

[0078] The industrial parameter acquisition unit is used to conduct virtual experiments and optimization simulations in the digital twin model of the water pump impeller process flow based on the potential relationships and optimization space among process parameters, and obtain the optimal process parameters of the water pump impeller. In this embodiment, the digital twin model and the existing production management system of the enterprise are used to analyze the historical data in the production process by using machine learning algorithms, excavate the potential relationships and optimization space among process parameters, conduct virtual experiments and optimization simulations in the digital twin model, and find the optimal process parameter settings; specifically, the PSO is used for dynamic parameter identification of the EHA system. EHA is the abbreviation of "Electro-Hydraulic Actuator", which is an important mechatronic system. It drives a hydraulic pump through an electric motor and uses hydraulic oil to transmit power, thereby achieving precise control of the load.

[0079] The steps are as follows:

[0080] a. Import data: Store the average impeller profile error and peak profile error obtained from the actual operation of the physical EHA at a certain moment into matrices respectively;

[0081] b. Objective function: A behavior model established with flow rate, head, specific speed, etc. as inputs and impeller machining error as output, where the behavior model includes a geometric model and a physical model. Therefore, given the initial data, running the behavior model in Simulink can obtain the impeller machining error. Based on the above analysis, take the error between the output value of the identification model and the actual impeller at the same moment as the objective function f:

[0082] f = |X i -T i |,

[0083] In the formula: X i - The output value of the rule model at the i-th moment; T i - The actual value of the impeller machining error at the i-th moment;

[0084] Finally, verify the feasibility of the processing technology and the optimization plan, and conduct a quality inspection model:

[0085] I(x, y) = I0(x, y) - I1(x, y),

[0086] where I(x, y) is the detected image, I0(x, y) is the standard image, I1(x, y) is the actual image, and adjust the process parameters to achieve the best optimization effect.

[0087] In this embodiment, the research focuses on the parameter and state identification technology of a high-precision machining system for water pump impellers based on digital twins. The state information and machining information of each intelligent manufacturing equipment are interacted and integrated to establish a collaborative control model for multiple intelligent manufacturing equipment groups and a method for joint debugging and control of intelligent manufacturing equipment. The management and control module is used to obtain the water pump impeller structure that meets the preset requirements based on the three-dimensional model of the water pump impeller, the flow field diagram of the water pump impeller, and the optimal process parameters of the water pump impeller, and complete the management and control of the water pump based on the integration of design-optimization-manufacturing.

[0088] The present invention addresses the problems of high computational cost, long optimization cycle, and imperfect manufacturing process in the whole process of "design-optimization-manufacturing" of water pumps. It organically combines the theories, methods, and technologies of multiple disciplines such as fluid mechanics and structural mechanics, and proposes a multi-objective optimization design method for water pump performance based on artificial neural networks and particle swarm algorithms, a high-precision rapid solution method for water pump performance based on the correction of the SST turbulence model, and a precise modeling method for the intelligent manufacturing process of water pump impellers based on digital twins. Finally, an impeller structure that fully meets the design parameters is obtained, forming an integrated design method for intelligent "design-optimization-manufacturing" of water pumps. Through this design method, the impeller designed by intelligent optimization has a high matching degree with the final impeller structure, and the accuracy and efficiency of the finally machined and manufactured impeller are improved.

[0089] Embodiment 2

[0090] The present invention also provides a water pump management and control method based on the integration of design-optimization-manufacturing. The application system and method include:

[0091] Collect and preprocess the water pump design-related parameters and actual production parameters;

[0092] Based on the preprocessed water pump design-related parameters, use machine learning and optimization algorithms to construct a three-dimensional model of the water pump impeller;

[0093] Based on the preprocessed water pump design-related parameters, use the SST turbulence model to construct a flow field diagram of the water pump impeller;

[0094] Based on the actual production data, construct and optimize the digital twin model of the water pump impeller process flow to obtain the optimal process parameters of the water pump impeller;

[0095] Based on the three-dimensional model of the water pump impeller, the flow field diagram of the water pump impeller, and the optimal process parameters of the water pump impeller, obtain the water pump impeller structure that meets the preset requirements, and complete the management and control of the water pump based on the integration of design-optimization-manufacturing.

[0096] A further implementation method is that the water pump design-related parameters include: geometric parameters of the water pump flow-through components, multi-condition efficiency of the water pump, pressure pulsation, and structural stress.

[0097] A further embodiment lies in that the method for obtaining the three-dimensional model of the water pump impeller includes:

[0098] Using the Navier-Stokes equation and the flow continuity equation, establishing the non-linear mathematical relationship between the geometric parameters of the water pump flow-through components and the multi-condition efficiency, pressure pulsation, and structural stress of the water pump, and constructing a non-linear mathematical model based on the boundary conditions through the non-linear mathematical relationship;

[0099] Based on the non-linear mathematical model, constructing an artificial neural network model for the hydraulic and structural performance of the water pump;

[0100] Optimizing the artificial neural network model based on the multi-objective particle swarm algorithm to obtain the three-dimensional model of the water pump impeller.

[0101] A further embodiment lies in that the method for obtaining the flow field diagram of the water pump impeller includes:

[0102] Simulating the flow field of the water pump impeller under the inverse pressure gradient based on the two-equation SST turbulence model;

[0103] Extracting the characteristics of the flow field of the water pump impeller based on the machine learning algorithm, and establishing the mapping relationship between the boundary conditions and the numerical simulation data to train the surrogate model to predict the internal pressure distribution of the water pump;

[0104] Based on the internal pressure distribution of the water pump, obtaining the flow field diagram of the water pump impeller.

[0105] The above-described embodiments are only descriptions of the preferred embodiments of the present invention, and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A water pump control system based on design-optimization-manufacturing integration, characterized in that: The system comprises: Data acquisition module, used to collect and pre-process pump design-related parameters and actual production parameters; A 3D model building module is used to build a 3D model of the pump impeller based on the pre-processed pump design related parameters using machine learning and optimization algorithms; The flow field reconstruction module is used to construct the pump impeller flow field diagram based on the pre-processed pump design related parameters using the SST turbulence model; A digital twin model building module is used to build and optimize a digital twin model of a water pump impeller process flow based on the actual production data to obtain optimal process parameters of the water pump impeller; The control module is used to obtain a water pump impeller structure that meets preset requirements based on the three-dimensional model of the water pump impeller, the flow field diagram of the water pump impeller and the optimal process parameters of the water pump impeller, and complete the control of the water pump based on the integration of design, optimization and manufacturing.

2. The system according to claim 1, characterized in that In the data acquisition module, the water pump design related parameters include: geometric parameters of water pump flow components, water pump multi-operating efficiency, pressure pulsation and structural stress.

3. The system according to claim 2, characterized in that The three-dimensional model building module includes: A mathematical model building unit, used to establish a nonlinear mathematical relationship between the geometric parameters of the flow-through components of the water pump and the multi-operating efficiency of the water pump, the pressure pulsation, and the structural stress by using the Navier-Stokes equation and the flow continuity equation, and to build a nonlinear mathematical model based on boundary conditions through the nonlinear mathematical relationship; A neural network model building unit, used to build an artificial neural network model of the hydraulic and structural performance of the water pump based on the nonlinear mathematical model; The model optimization unit is used to optimize the artificial neural network model based on a multi-objective particle swarm algorithm to obtain the three-dimensional model of the water pump impeller.

4. The system according to claim 1, characterized in that The flow field reconstruction module comprises: Numerical simulation unit, used to simulate the flow field of the pump impeller under adverse pressure gradient based on the two-equation SST turbulence model; A pressure distribution acquisition unit, used to extract the characteristics of the water pump impeller flow field based on a machine learning algorithm, and to establish a mapping relationship between boundary conditions and numerical simulation data to train a proxy model, so as to predict the pressure distribution inside the water pump; A flow field map acquisition unit is used to obtain the water pump impeller flow field map based on the internal pressure distribution of the water pump.

5. The system according to claim 4, characterized in that The two-equation SST turbulence model includes two models: a low Reynolds number model k-ω and a high Reynolds number model k-ε; wherein the low Reynolds number model k-ω is used for numerical solution in the boundary layer region, and the high Reynolds number model k-ε is used for numerical solution in the free shear layer.

6. The system according to claim 1, characterized in that The digital twin model building module includes: A twin model building unit is used to build a digital twin model of the water pump impeller process based on the water pump impeller process design documents, equipment drawings and actual production data; Optimization space acquisition unit, used to mine the potential relationship and optimization space between process parameters based on the historical data of the water pump impeller production process; The industrial parameter acquisition unit is used to perform virtual experiments and optimization simulations in the digital twin model of the water pump impeller process flow based on the potential relationship and optimization space between the process parameters to obtain the optimal process parameters of the water pump impeller.

7. A water pump control method based on design-optimization-manufacturing integration, using the system described in any one of claims 1 to 6, characterized in that: The method comprises: Collect and pre-process pump design-related parameters and actual production parameters; Based on the preprocessed pump design parameters, a three-dimensional model of the pump impeller is constructed using machine learning and optimization algorithms; Based on the pre-processed pump design parameters, the SST turbulence model is used to construct the pump impeller flow field diagram; Based on the actual production data, a digital twin model of the water pump impeller process flow is constructed and optimized to obtain the optimal process parameters of the water pump impeller; Based on the water pump impeller three-dimensional model, the water pump impeller flow field diagram and the water pump impeller optimal process parameters, a water pump impeller structure that meets the preset requirements is obtained, and the water pump is managed and controlled based on the integration of design, optimization and manufacturing.

8. The method according to claim 7, characterized in that The water pump design related parameters include: geometric parameters of water pump flow components, water pump multi-operating efficiency, pressure pulsation and structural stress.

9. The method according to claim 8, characterized in that The method for obtaining the three-dimensional model of the water pump impeller includes: Using the Navier-Stokes equation and the flow continuity equation, a nonlinear mathematical relationship between the geometric parameters of the flow-through components of the water pump and the multi-operating efficiency of the water pump, the pressure pulsation, and the structural stress is established, and a nonlinear mathematical model based on boundary conditions is constructed through the nonlinear mathematical relationship; Based on the nonlinear mathematical model, an artificial neural network model of the hydraulic and structural performance of the water pump is constructed; The artificial neural network model is optimized based on a multi-objective particle swarm algorithm to obtain a three-dimensional model of the water pump impeller.

10. The method according to claim 9, characterized in that The method for obtaining the water pump impeller flow field diagram comprises: The flow field of the pump impeller under adverse pressure gradient is simulated based on the two-equation SST turbulence model; Extracting the characteristics of the water pump impeller flow field based on a machine learning algorithm, and establishing a mapping relationship between boundary conditions and numerical simulation data to train a proxy model to predict the internal pressure distribution of the water pump; Based on the internal pressure distribution of the water pump, the water pump impeller flow field diagram is obtained.

Images