Optimization design method, system and device for axial flow pump circumferential groove casing treatment and storage medium

Abstract

The invention belongs to the technical field of optimization design of axial flow pumps, and particularly provides an optimization design method, system and device for circumferential groove casing treatment of an axial flow pump and a storage medium, and the method comprises the steps of automatic CFD numerical simulation calculation of the circumferential groove casing treatment of the axial flow pump; quantitative calculation of blade top blockage of the axial flow pump; constructing an efficient optimization strategy and an optimization design problem of the operation stability of the axial flow pump; and a multi-dimensional nonlinear optimization design problem based on the multi-target EGO algorithm is efficiently solved. According to the method, the optimal design efficiency and design effect of circumferential groove casing treatment of the axial flow pump can be improved, and the influence of circumferential groove casing treatment on the operation efficiency of the axial flow pump is reduced while the stable operation working condition range of the axial flow pump is expanded to the maximum extent.

Description

Technical Field

[0001] This invention belongs to the field of axial flow pump optimization design technology, specifically, it relates to an optimization design method, system, device and storage medium for axial flow pump circumferential groove casing treatment. Background Technology

[0002] Axial flow pumps are a type of impeller pump characterized by high flow rate and low head. They are widely used in various fields such as military, industrial, and agricultural irrigation and drainage. High operating efficiency and a wide range of stable operating conditions are important design indicators for axial flow pumps. Conventionally designed axial flow pumps can only maintain stable operation near their design conditions. However, in actual operation, axial flow pumps often inevitably operate under non-design conditions. For example, the start-up and shutdown transition processes of axial flow pumps in pumped storage power stations are typical scenarios. When an axial flow pump operates at a low flow rate, complex flow phenomena such as backflow and impeller tip leakage will occur in the pump's flow channel. This will not only significantly deteriorate the internal flow stability of the pump but also cause significant noise and vibration, ultimately threatening the safe and stable operation of the axial flow pump.

[0003] To broaden the efficient and stable operating range of axial flow pumps, various active and passive flow control technologies have become a research hotspot in this field in recent years. Among them, circumferential groove casing treatment, as a simple and reliable passive stabilization enhancement technology, is widely used in the field of axial flow machinery. However, while improving the operational stability of axial flow machinery under low flow conditions, the design of circumferential groove casing treatment usually leads to a decrease in operating efficiency, i.e., there is a contradictory problem of "enhancing stability but reducing efficiency". In addition, if the design parameters of the circumferential groove casing treatment are unreasonable, it may even fail to improve operational stability.

[0004] Currently, the selection of optimal parameters for various casing treatment designs of axial flow pumps still largely relies on parametric research methods. This involves systematically changing the structural design parameters of the casing treatment, then analyzing their impact on various performance aspects of the axial flow pump (stable operating range, operating efficiency, etc.), and finally comparing and selecting a set of "optimal" structural design parameter combinations. Furthermore, calculating the stable operating range of the axial flow pump depends on time-consuming simulations of complete performance curves, further increasing the difficulty and computational workload of related optimization designs. Parametric research methods and existing methods for calculating stable operating ranges result in casing treatment optimization designs that are not only time-consuming and labor-intensive, but also only yield approximate "optimal" structural designs, rather than truly optimal ones. Moreover, parametric research methods struggle to consider the nonlinear coupling effects between design parameters and optimization objectives, especially when there are many design parameters and optimization objectives. This leads to a significant deviation between the obtained "optimal" design and the actual optimal design, failing to fully realize the true application effects of the casing treatment.

[0005] Therefore, there is an urgent need to develop efficient and feasible optimization design methods for circumferential groove casing treatment, so as to maximize the improvement of operational stability while reducing its impact on operational efficiency, especially for the circumferential groove casing treatment design of axial flow pumps, for which there is still relatively little research. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide an optimized design method, system, device and storage medium for the circumferential groove casing treatment of axial flow pumps, so as to improve the optimization design efficiency and design effect of the circumferential groove casing treatment of axial flow pumps, maximize the range of stable operating conditions of axial flow pumps, and reduce the impact of circumferential groove casing treatment on the operating efficiency of axial flow pumps.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: an optimized design method for the treatment of the circumferential groove casing of an axial flow pump, comprising the following steps: Step 1: Automatic CFD numerical simulation calculation of the circumferential groove casing treatment of the axial flow pump; Step 2: Quantitative calculation of axial flow pump impeller tip blockage; Step 3: Constructing efficient optimization strategies and optimization design problems for the operational stability of axial flow pumps; Step 4: Efficiently solve multidimensional nonlinear optimization design problems based on the multi-objective EGO algorithm.

[0008] In the preferred embodiment, the specific process of automatic CFD numerical simulation calculation for the circumferential groove casing treatment of the axial flow pump in step 1 is as follows: S11. Use 3D modeling software to perform geometric modeling of the fluid domain of the axial flow pump with circumferential groove casing. S12. Use a mesh generation tool to mesh the simulation computational fluid domain; S13. Based on the operating parameters of the optimized operating point and the flow characteristics of the axial flow pump, select a suitable physical model, set the boundary conditions for numerical simulation, and perform simulation calculations.

[0009] In a preferred embodiment, in step S11, the axial flow pump fluid domain includes rotor blade flow channels, stator blade flow channels, and circumferential groove casing processing flow channels.

[0010] In a preferred embodiment, during step S12, the rotation period of the circumferential groove computational domain is kept consistent with the rotation period of the rotor computational domain, and a complete mismatch technique is used to connect the circumferential groove mesh and the rotor flow channel computational domain mesh.

[0011] In the preferred embodiment, the specific process of quantifying the blockage at the tip of the axial flow pump in step 2 is as follows: S21. Selection of reference plane and velocity component for blockage quantification calculation: Select the z-axis, which is perpendicular to the rotation axis of the axial flow pump within the blade channel. The plane is used as the reference plane for blockage quantification calculation, and the axial velocity component is adopted. The end-wall blockage value is calculated, where, r Radial, i Tangential; S22. Flow field data export: Based on the CFD numerical simulation results, extract and export the product of density and axial velocity at different blade height positions. and its gradient The flow field data was obtained, and the data was preprocessed by deduplication and sorting to sort the data at different blade heights. i Arranged in ascending order; S23. Determination of the outer boundary of the end-wall blockage area: using... The gradient in the radial direction r and tangential i The L1 norm and threshold CV of the components are used to divide the blockage region and the mainstream region, and to determine the outer boundary of the end-wall blockage region. S24, The product of density and axial velocity at the outer boundary of the clogging area. Calculation: The average value within the quantized reference plane will be calculated. As The calculation formula is: ; in, m This indicates the mass flow rate of the fluid at the reference surface calculated using the selected blockage. A This represents the area of ​​the selected reference surface for blockage calculation; S25. Calculation of local blockage value: Combining the definition of boundary layer displacement thickness, using the preprocessed flow field data from step S22 and the values ​​calculated in step S24... Calculate the local blockage value at different blade height positions; S26. Calculation of tip blockage: Radial integration is performed on the local blockage at different blade height positions to obtain the tip blockage value at different axial positions within the blade channel.

[0012] In the preferred embodiment, the expression for determining the outer boundary of the end-wall blockage region in step S23 is: ; In the formula, CV represents the threshold, and the selection of the threshold CV is based on the reference plane. cloud map and The contour lines are used to determine the value.

[0013] In the preferred embodiment, in step S25, the local blockage value The calculation formula is: ; in,r Indicates different blade height positions in the radial direction. NB This indicates the number of blades in an axial flow pump.

[0014] In the preferred embodiment, step 3, specifically, involves constructing the optimization strategy as follows: using the reduction of the tip blockage value under near-stall conditions as the flow stability optimization strategy for the axial flow pump, and selecting a strategy that simultaneously reduces the maximum blockage B within the rotor flow channel. max Blockage near the trailing edge of the blade B end As the target for optimizing flow stability, and the operating efficiency under design conditions or With maximizing as another optimization objective, the optimization design problem for the circumferential groove casing of an axial flow pump is described as follows: ; In the formula, Indicates the first i Optimization design variables for circumferential slot casing processing and They represent the first i The lower and upper limits of the optimization design variables for the circumferential slot casing.

[0015] In the preferred embodiment, the optimized design variables DV This includes the width, height, tilt angle, and placement of the circumferential groove.

[0016] In the preferred embodiment, the specific process of efficiently solving the multidimensional nonlinear optimization design problem based on the multi-objective EGO algorithm in step 4 is as follows: S41. Use the optimal Latin hypercube experimental design method to uniformly distribute the initial sample point set in the design space; S42. Based on the values ​​of the design variables of the initial sample points, a three-dimensional geometric model of the circumferential groove casing is generated using three-dimensional modeling software. The automatic CFD numerical simulation calculation of the circumferential groove casing of the axial flow pump in step 1 is used for automatic mesh generation, model selection, simulation setting and numerical calculation. Combined with the quantification calculation method of the axial flow pump blade tip blockage in step 2, the objective function values ​​of each initial sample point are obtained, and the initial sample database is constructed. S43. Initiate the optimization iteration process: 1) The Pareto front solution set is calculated from the sample database using a non-dominated sorting algorithm; 2) Determine whether the optimization convergence condition is met. If not, use the sample database to build or update the Kriging surrogate model. 3) EIM based on the Kriging proxy model e The selection criteria for correction points are updated and combined with the differential evolution algorithm (DE) to search for the EIM. e The point with the largest value is used as the correction point; 4) The performance of the correction points is evaluated and the objective function value is calculated by automatic CFD numerical simulation and objective function calculation, respectively. The results are then added to the sample database to update the Pareto front solution set and the Kriging surrogate model. S44. Repeat S43 until the optimization convergence condition is met, and derive the final Pareto front solution set as the optimal solution set.

[0017] In the preferred embodiment, in step S44, the criterion for determining convergence is: the Pareto front no longer increases the hypervolume relative to a given point.

[0018] In the preferred embodiment, the EIM e The selection criterion for the correction points is based on the improved expectation matrix method using Eulerian distance. The EIM function is quantized with reference to the definition of the improved EI function using Eulerian distance, resulting in the improved expectation matrix method EIM based on Eulerian distance. e The EIM calculation formula is as follows: ; In the formula, superscript k Indicates the number of optimal solutions, subscript m The dimension representing the optimal solution is shown below. Each element of the matrix is ​​an EI function, calculated as follows: ; In the formula, Indicates the first j The first point i The current optimal solution for each objective function value; and These represent the unknown points in the design space of the Kriging surrogate model. x The first i The predicted value and root mean square error of the objective function; and Let represent the cumulative function and probability density function of the standard normal distribution, respectively; Represents a point in the design space x Compared to the current Pareto front solution set, the th j The first point i The expected improvement in the value of the objective function.

[0019] In a preferred embodiment, the EIM includes a point within the design space. Compared to the improvement expectation of all points in the current Pareto front solution set on all objective functions, the EIM function is quantized by referring to the definition of the improved EI function based on Eulerian distance, resulting in the EIM method based on the improved expectation matrix of Eulerian distance. e ,as follows: .

[0020] This invention also provides an optimization design system for the circumferential groove casing treatment of an axial flow pump, used to execute the above-described optimization design method for the circumferential groove casing treatment of an axial flow pump, comprising: The automatic CFD numerical simulation module is used to perform geometric modeling, mesh generation, boundary condition and physical model setting, and simulation calculation of the fluid domain of an axial flow pump with a circumferential groove casing. The entire process is automatically implemented and called through macro files and batch files. The blade tip blockage quantification calculation module, based on classical boundary layer theory, selects the blockage quantification calculation reference surface and... Calculate the endwall clogging value, extract and preprocess the flow field data, divide the clogging region into the main flow region, and calculate the value at the outer boundary of the clogging region. Localized blockage values ​​and blade tip region blockage values; The optimization strategy and design problem building module aims to reduce the tip blockage value under near-stall conditions as a flow stability optimization strategy, setting the maximum blockage B in the rotor flow channel as the target. max Blockage near the trailing edge of the blade B end Maximize operating efficiency under design conditions or To optimize the target, we construct an optimization design problem for the circumferential groove casing of an axial flow pump. Based on the multi-objective EGO algorithm solution module, an optimal Latin hypercube experimental design is used to arrange the initial sample point set and construct an initial sample database. The Pareto front solution set is calculated using a non-dominated sorting algorithm, and the solution is based on the Kriging surrogate model and EIM. e The selection criteria for calibration points are combined with the differential evolution algorithm to search for calibration points. The sample database and model are iteratively updated until convergence, and the optimal solution set is output.

[0021] The present invention also provides an optimization design device for the treatment of the circumferential groove casing of an axial flow pump, comprising a processor and a memory, wherein the memory stores a computer program, and when the processor executes the computer program, it implements the above-described optimization design method for the treatment of the circumferential groove casing of an axial flow pump.

[0022] The present invention also provides an optimized design storage medium for the treatment of the circumferential groove casing of an axial flow pump. The storage medium is a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements all the steps of the optimized design method for the treatment of the circumferential groove casing of an axial flow pump described above.

[0023] The present invention provides an optimized design method, system, device, and storage medium for the treatment of the circumferential groove casing of an axial flow pump, which has the following beneficial effects: 1. It can significantly improve the optimization design efficiency and effect of the circumferential groove casing treatment of axial flow pumps, and at the same time, it can maximize the range of stable operating conditions of axial flow pumps while reducing the impact of the circumferential groove casing treatment on the operating efficiency of axial flow pumps, thus better expanding the application scope of axial flow pumps in military, industrial and agricultural irrigation and drainage fields.

[0024] 2. To address the shortcomings of existing parametric research methods that rely on systematically changing design parameters and analyzing performance effects one by one, this invention automates the entire process of CFD numerical simulation calculation, enabling automatic implementation and invocation of geometric modeling, mesh generation, simulation settings and calculations. This significantly reduces manual intervention, avoids repetitive and tedious manual operations, and shortens the design cycle.

[0025] 3. Traditional calculations of stable operating conditions rely on time-consuming simulations of complete performance curves. This invention proposes a stability optimization strategy centered on reducing tip blockage in near-stall conditions, eliminating the need for complete performance curve simulations for every new design. Combined with the efficient solution capabilities of the multi-objective EGO algorithm, it achieves this through optimal Latin hypercube experimental design with uniform initial sample points, Kriging surrogate model construction and iterative updates, and EIM... e The search method that combines the selection criterion for correction points with the differential evolution algorithm can quickly converge to the optimal solution set, significantly reducing the workload and time cost of optimization calculations, and solving the problems of large computational load and low efficiency in multidimensional nonlinear optimization design.

[0026] 4. Existing parametric research methods are difficult to consider the nonlinear coupling effects between design parameters and optimization objectives, and can only obtain approximate optimal designs. However, this invention uses the multi-objective EGO algorithm to efficiently solve multidimensional nonlinear optimization problems. Combined with the accuracy of blade tip blockage quantification calculation, it fully captures the coupling relationship between design variables and optimization objectives. The final Pareto front solution set effectively avoids local optima and ensures that the circumferential groove casing treatment design can fully play its application role.

[0027] To reduce the maximum blockage of the rotor flow channel B max Blockage near the trailing edge of the blade B end Synchronous control of tip leakage flow and its trajectory ensures a wider range of stable operating conditions; while maintaining design operating efficiency. or With the goal of maximizing efficiency, the design achieves a two-way optimization of stability and efficiency, breaking the contradiction that the two cannot be achieved simultaneously in traditional design. This allows the axial flow pump to expand its stable operating range while minimizing efficiency loss.

[0028] 5. This invention is based on a blade tip blockage quantification calculation method developed from classical boundary layer theory, by selecting... r-θThe reference plane and axial velocity components accurately extract flow field data, divide clogging areas, and calculate local clogging values ​​and blade tip clogging values, providing precise quantitative basis for flow stability optimization. It can specifically suppress complex flow phenomena such as backflow and blade tip leakage under low flow conditions, fundamentally improve the flow stability inside the pump, reduce noise and vibration caused by flow instability, reduce the risk of equipment failure, and ensure the safe and stable operation of axial flow pumps. Attached Figure Description

[0029] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a flowchart of the technical solution of the present invention; Figure 2 This is a schematic diagram of the fluid domain of an axial flow pump with a circumferentially grooved casing. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention.

[0031] Example 1: An optimization design method for the circumferential groove casing treatment of an axial flow pump includes modules such as automatic CFD numerical simulation calculation of the circumferential groove casing treatment, quantitative calculation of axial flow pump impeller tip blockage, efficient optimization strategy for axial flow pump operating stability and construction of optimization design problem, and efficient solution of multidimensional nonlinear optimization design problem based on multi-objective EGO algorithm. Figure 1 As shown.

[0032] Specifically, it includes the following steps: Step 1: Automatic CFD numerical simulation calculation of the circumferential groove casing treatment of the axial flow pump. The specific process is as follows: S11. A 3D modeling software is used to geometrically model the fluid domain of the axial flow pump with a circumferentially grooved casing, including the rotor blade flow channel, stator blade flow channel, and the flow channel of the circumferentially grooved casing, etc. Figure 2 As shown.

[0033] S12. The computational fluid domain for simulation is meshed using a meshing tool. During meshing, the rotation period of the circumferential groove computational domain is kept consistent with that of the rotor computational domain, and a completely mismatched mesh technique is used to connect the circumferential groove mesh and the rotor flow channel computational domain mesh. The completely mismatched mesh technique is an advanced multi-block mesh connection method that allows fluid calculations to be performed on non-matched mesh interfaces without forced mesh alignment or interpolation. Compared with traditional matched meshing methods, the completely mismatched technique can significantly reduce the difficulty and time of mesh generation, and is especially suitable for parametric design or complex geometric deformation scenarios, such as the geometric optimization design of the current circumferential groove casing.

[0034] S13. Based on the operating parameters of the optimized operating point and the flow characteristics of the axial flow pump, select a suitable physical model, set the boundary conditions for numerical simulation, and perform simulation calculations.

[0035] In addition, to facilitate the implementation of automated optimization processes, macro files are recorded and batch files are written to enable the automated implementation of the entire CFD numerical simulation calculation process and its on-demand recall during optimization.

[0036] Step 2: Quantitative calculation of axial flow pump impeller tip blockage.

[0037] To accurately quantify tip flow blockage within the rotor blade passage of an axial flow pump, a quantitative calculation method for tip blockage within the impeller passage of an axial flow pump was developed based on classical boundary layer theory. The specific process is as follows: S21. Selection of reference plane and velocity component for blockage quantification calculation: Select the z-axis, which is perpendicular to the rotation axis of the axial flow pump within the blade channel. The plane is used as the reference plane for blockage quantification calculation, and the axial velocity component is adopted. The end-wall blockage value is calculated, where, r Radial, i Tangential.

[0038] S22. Flow field data export: Based on the CFD numerical simulation results, and combined with the selection of the reference plane and velocity components in step S21, the product of density and axial velocity at different blade height positions is extracted and exported. and its gradient The flow field data was obtained, and the data was preprocessed by deduplication and sorting to sort the data at different blade heights. i Arrange them in ascending order for easier subsequent calculations.

[0039] S23. Determination of the outer boundary of the end-wall blockage area: Based on... The gradient in the congested region is significantly larger than its gradient in the mainstream region. This can be utilized... The gradient in the radial direction r and tangential iThe L1 norm and threshold CV of the components are used to divide the blockage region into the mainstream region and determine the outer boundary of the end-wall blockage region.

[0040] The expression for determining the outer boundary of the end-wall blockage region is: ; In the formula, CV represents the threshold, and the selection of the threshold CV is based on the reference plane. cloud map and The contour lines are used to determine the location. The colors in the contour map effectively distinguish between blocked and mainstream areas, and this, combined with the characteristics of fluid flow, further helps to differentiate them. The value will be significantly larger in the congested region than in the mainstream region, therefore it will be located near the outer boundary of the congested region. The value corresponding to the contour line can be selected as the threshold.

[0041] S24, The product of density and axial velocity at the outer boundary of the clogging area. Calculation: The average value within the quantized reference plane will be calculated. As The calculation formula is: ; in, m This indicates the mass flow rate of the fluid at the reference surface calculated using the selected blockage. A This represents the area of ​​the selected reference surface for blockage calculation.

[0042] S25. Calculation of local blockage value: Combining the definition of boundary layer displacement thickness, using the preprocessed flow field data from step S22 and the values ​​calculated in step S24... Calculate the local blockage value at different leaf height positions.

[0043] Local blockage value The calculation formula is: ; in, r Indicates different blade height positions in the radial direction. NB This indicates the number of blades in an axial flow pump.

[0044] S26. Calculation of tip blockage: Radial integration is performed on the local blockage at different blade height positions to obtain the tip blockage value at different axial positions within the blade channel.

[0045] Step 3: Constructing efficient optimization strategies and optimization design problems for the operational stability of axial flow pumps.

[0046] Based on the analysis of the flow instability characteristics of axial flow pumps under low flow conditions, the flow instability of axial flow pumps is mainly related to the magnitude of flow blockage in the blade tip region caused by tip leakage. Therefore, a flow stability optimization strategy for axial flow pumps is proposed, which involves reducing the tip blockage value near stall conditions. This avoids the need for simulation calculations of the stable operating range for each new design during the optimization process, significantly reducing the computational load. Furthermore, considering the importance of the tip blockage location to the flow instability of axial flow pumps, the flow stability optimization objective is chosen to simultaneously reduce the maximum blockage within the rotor channel. B max and blockage near the trailing edge of the blade B end This enables synchronous control of tip leakage flow-related blockage and tip leakage flow trajectory. The maximum blockage and blockage near the blade trailing edge can be calculated by setting multiple tip blockage quantification reference surfaces in the rotor blade flow channel (for example, setting one reference surface every 5% of the tip chord length).

[0047] Meanwhile, in order to reduce the impact of the circumferential groove casing design on the operating efficiency of the axial flow pump, the operating efficiency under the design conditions will also be considered during the optimization process. or Maximizing this is another optimization objective; therefore, the complete optimization design problem for the circumferential groove casing of the axial flow pump can be formulated as follows: ; In the formula, Indicates the first i Optimization design variables for circumferential slot casing processing and They represent the first i The lower and upper limits of the optimization design variables for the circumferential slot casing.

[0048] The optimized design variables DV This includes the width, height, tilt angle, and placement of the circumferential groove, such as... Figure 2 As shown in the image.

[0049] Step 4: Efficiently solve multidimensional nonlinear optimization design problems based on the multi-objective EGO algorithm.

[0050] The specific process is as follows: S41. Optimization begins with DOE (Design of Experiments), employing the optimal Latin hypercube experimental design method to uniformly distribute the initial sample point set within the design space. S42. Based on the values ​​of the design variables of the initial sample points, a three-dimensional geometric model of the circumferential groove casing is generated using three-dimensional modeling software. The automatic CFD numerical simulation calculation of the circumferential groove casing of the axial flow pump in step 1 is used for automatic mesh generation, model selection, simulation setting and numerical calculation. Combined with the quantification calculation method of the axial flow pump blade tip blockage in step 2, the objective function values ​​of each initial sample point are obtained, and the initial sample database is constructed.

[0051] S43. Initiate the optimization iteration process: 1) First, the Pareto front solution set is calculated from the sample database using the non-dominated sorting algorithm.

[0052] 2) Next, determine whether the optimization convergence condition is met. If not, use the sample database to build or update the Kriging surrogate model. For designs that do not meet the optimization constraints, use the penalty function method to handle them.

[0053] 3) EIM based on the Kriging proxy model e (An improved expectation matrix method based on Eulerian distance) The correction point selection criterion is updated and combined with the Differential Evolution (DE) algorithm to search for the EIM. e The point with the largest value is used as the correction point.

[0054] EIM e The selection criterion for the correction points is based on the improved expectation matrix method using Eulerian distance. The EIM function is quantized with reference to the definition of the improved EI function using Eulerian distance, resulting in the improved expectation matrix method EIM based on Eulerian distance. e EIM is a multi-objective correction point selection criterion derived from the single-objective EI correction point selection criterion. The calculation method is as follows: ; In the formula, superscript k Indicates the number of optimal solutions, subscript m The dimension representing the optimal solution is shown below. Each element of the matrix is ​​an EI function, calculated as follows: ; In the formula, Indicates the first j The first point i The current optimal solution for each objective function value; and These represent the unknown points in the design space of the Kriging surrogate model. x The first i The predicted value and root mean square error of the objective function; and Let represent the cumulative function and probability density function of the standard normal distribution, respectively; Represents a point in the design space x Compared to the current Pareto front solution set, the th j The first point i The expected improvement in the value of the objective function.

[0055] Therefore, EIM includes a point within the design space. The expected improvement of all points in the current Pareto front solution set across all objective functions is compared with the expected improvement of the EI function. Then, the EIM function is quantified by referring to the definition of the improved EI function based on the Eulerian distance, in order to measure the improvement of a point in the design space. Compared to the overall improvement of the current Pareto front solution set, we obtain the improved expectation matrix method EIM based on Eulerian distance. e ,as follows: .

[0056] 4) The performance of the correction points is evaluated and the objective function value is calculated by automatic CFD numerical simulation and objective function calculation, respectively. The results are then added to the sample database to update the Pareto front solution set and the Kriging surrogate model. S44. Repeat S43 until the optimization convergence condition is met. The convergence of the optimization iteration is judged by the trend of the change of the hypervolume of the Pareto front relative to the given point. When the hypervolume of the Pareto front no longer increases, the optimization is considered to have converged, and the final Pareto front solution set is derived as the optimal solution set.

[0057] Example 2: This embodiment provides an optimized design system for the treatment of the circumferential groove casing of an axial flow pump, characterized in that it includes a method for performing the optimized design of the circumferential groove casing treatment of an axial flow pump as described in Embodiment 1, comprising: The automatic CFD numerical simulation module is used to perform geometric modeling, mesh generation, boundary condition and physical model setting, and simulation calculation of the fluid domain of an axial flow pump with a circumferential groove casing. The entire process is automatically implemented and called through macro files and batch files. The blade tip blockage quantization calculation module, based on classical boundary layer theory, selects a blockage quantization reference plane perpendicular to the rotation axis and w z Calculate the endwall clogging value, extract and preprocess the flow field data, divide the clogging region into the main flow region, and calculate the value at the outer boundary of the clogging region. Localized blockage values ​​and blade tip region blockage values; The optimization strategy and design problem building module aims to reduce the tip blockage value under near-stall conditions as a flow stability optimization strategy, setting the maximum blockage B in the rotor flow channel as the target.max Blockage near the trailing edge of the blade B end Maximize operating efficiency under design conditions or To optimize the target, we construct an optimization design problem for the circumferential groove casing of an axial flow pump. Based on the multi-objective EGO algorithm solution module, an optimal Latin hypercube experimental design is used to arrange the initial sample point set and construct an initial sample database. The Pareto front solution set is calculated using a non-dominated sorting algorithm, and the solution is based on the Kriging surrogate model and EIM. e The selection criteria for calibration points are combined with the differential evolution algorithm to search for calibration points. The sample database and model are iteratively updated until convergence, and the optimal solution set is output.

[0058] Example 3: This embodiment provides an optimization design device for the treatment of the circumferential groove casing of an axial flow pump, including a processor and a memory. The memory stores a computer program, and when the processor executes the computer program, it implements the optimization design method for the treatment of the circumferential groove casing of an axial flow pump as described in Embodiment 1.

[0059] Example 4: This embodiment provides an optimized design storage medium for the treatment of the circumferential groove casing of an axial flow pump. The storage medium is a computer-readable storage medium that stores a computer program. When the computer program is executed by a processor, it implements all the steps of the optimized design method for the treatment of the circumferential groove casing of an axial flow pump described in Embodiment 1.

[0060] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An optimized design method for the circumferential groove casing treatment of an axial flow pump, characterized in that, Includes the following steps: Step 1: Automatic CFD numerical simulation calculation of the circumferential groove casing treatment of the axial flow pump; Step 2: Quantitative calculation of axial flow pump impeller tip blockage; Step 3: Constructing efficient optimization strategies and optimization design problems for the operational stability of axial flow pumps; Step 4: Efficiently solve multidimensional nonlinear optimization design problems based on the multi-objective EGO algorithm.

2. The optimized design method for the circumferential groove casing treatment of an axial flow pump according to claim 1, characterized in that, In step 1, the specific process of automatic CFD numerical simulation calculation for the circumferential groove casing treatment of the axial flow pump is as follows: S11. Use 3D modeling software to perform geometric modeling of the fluid domain of the axial flow pump with circumferential groove casing. S12. Use a mesh generation tool to mesh the simulation computational fluid domain; S13. Based on the operating parameters of the optimized operating point and the flow characteristics of the axial flow pump, select a suitable physical model, set the boundary conditions for numerical simulation, and perform simulation calculations.

3. The optimized design method for the circumferential groove casing treatment of an axial flow pump according to claim 2, characterized in that, In step S11, the axial flow pump fluid domain includes rotor blade flow channel, stator blade flow channel and circumferential groove casing processing flow channel.

4. The optimized design method for the circumferential groove casing treatment of an axial flow pump according to claim 2, characterized in that, In step S12, during mesh generation, the rotation period of the circumferential groove computational domain is kept consistent with the rotation period of the rotor computational domain, and a complete mismatch technique is used to connect the circumferential groove mesh and the rotor flow channel computational domain mesh.

5. The optimized design method for the circumferential groove casing treatment of an axial flow pump according to claim 1, characterized in that, In step 2, the specific process of quantifying the blockage at the tip of the axial flow pump is as follows: S21. Selection of reference plane and velocity component for blockage quantification calculation: Select the z-axis, which is perpendicular to the rotation axis of the axial flow pump within the blade channel. The plane is used as the reference plane for blockage quantification calculation, and the axial velocity component is adopted. The end-wall blockage value is calculated, where, r Radial, θ Tangential; S22. Flow field data export: Based on the CFD numerical simulation results, extract and export the product of density and axial velocity at different blade height positions. and its gradient The flow field data was obtained, and the data was preprocessed by deduplication and sorting to sort the data at different blade heights. θ Arranged in ascending order; S23. Determination of the outer boundary of the end-wall blockage area: using... The gradient in the radial direction r and tangential θ The L1 norm and threshold CV of the components are used to divide the blockage region and the mainstream region, and to determine the outer boundary of the end-wall blockage region. S24, the product of density and axial velocity at the outer boundary of the clogging area. Calculation: The average value within the quantized reference plane will be calculated. As The calculation formula is: ； in, m This indicates the mass flow rate of the fluid at the reference surface calculated using the selected blockage. A This represents the area of ​​the selected reference surface for blockage calculation; S25. Calculation of local blockage value: Combining the definition of boundary layer displacement thickness, using the preprocessed flow field data from step S22 and the values ​​calculated in step S24... Calculate the local blockage value at different blade height positions; S26. Calculation of tip blockage: Radial integration is performed on the local blockage at different blade height positions to obtain the tip blockage value at different axial positions within the blade channel.

6. The optimized design method for the circumferential groove casing treatment of an axial flow pump according to claim 5, characterized in that, In step S23, the expression for determining the outer boundary of the end-wall blockage region is: ； In the formula, CV represents the threshold, and the selection of the threshold CV is based on the reference plane. cloud map and The contour lines are used to determine the value.

7. The optimized design method for the circumferential groove casing treatment of an axial flow pump according to claim 5, characterized in that, In step S25, the local blockage value The calculation formula is: ； in, r Indicates different blade height positions in the radial direction. NB This indicates the number of blades in an axial flow pump.

8. The optimized design method for the circumferential groove casing treatment of an axial flow pump according to claim 1, characterized in that, In step 3, the optimization strategy is specifically constructed as follows: The strategy for optimizing the flow stability of the axial flow pump is to reduce the tip blockage value under near-stall conditions, and to simultaneously reduce the maximum blockage B within the rotor flow channel. max Blockage near the trailing edge of the blade B end As the target for optimizing flow stability, and the operating efficiency under design conditions η With maximizing as another optimization objective, the optimization design problem for the circumferential groove casing of an axial flow pump is described as follows: ； In the formula, Indicates the first i Optimization design variables for circumferential slot casing processing and They represent the first i The lower and upper limits of the optimization design variables for the circumferential slot casing.

9. The optimized design method for the circumferential groove casing treatment of an axial flow pump according to claim 8, characterized in that, The optimized design variables DV This includes the width, height, tilt angle, and placement of the circumferential groove.

10. The optimized design method for the circumferential groove casing treatment of an axial flow pump according to claim 1, characterized in that, In step 4, the specific process for efficiently solving the multidimensional nonlinear optimization design problem based on the multi-objective EGO algorithm is as follows: S41. Use the optimal Latin hypercube experimental design method to uniformly distribute the initial sample point set in the design space; S42. Based on the values ​​of the design variables of the initial sample points, a three-dimensional geometric model of the circumferential groove casing is generated using three-dimensional modeling software. The automatic CFD numerical simulation calculation of the circumferential groove casing of the axial flow pump in step 1 is used for automatic mesh generation, model selection, simulation setting and numerical calculation. Combined with the quantification calculation method of the axial flow pump blade tip blockage in step 2, the objective function values ​​of each initial sample point are obtained, and the initial sample database is constructed. S43. Initiate the optimization iteration process: 1) The Pareto front solution set is calculated from the sample database using a non-dominated sorting algorithm; 2) Determine whether the optimization convergence condition is met. If not, use the sample database to build or update the Kriging surrogate model. 3) EIM based on the Kriging proxy model e The selection criteria for correction points are updated and combined with the differential evolution algorithm (DE) to search for the EIM. e The point with the largest value is used as the correction point; 4) The performance of the correction points is evaluated and the objective function value is calculated by automatic CFD numerical simulation and objective function calculation, respectively. The results are then added to the sample database to update the Pareto front solution set and the Kriging surrogate model. S44. Repeat S43 until the optimization convergence condition is met, and derive the final Pareto front solution set as the optimal solution set.

11. The optimized design method for the circumferential groove casing treatment of an axial flow pump according to claim 10, characterized in that, In step S44, the criterion for determining convergence is that the hypervolume of the Pareto front no longer increases relative to a given point.

12. The optimized design method for the circumferential groove casing treatment of an axial flow pump according to claim 10, characterized in that, The EIM e The selection criterion for the correction points is based on the improved expectation matrix method using Eulerian distance. The EIM function is quantized with reference to the definition of the improved EI function using Eulerian distance, resulting in the improved expectation matrix method EIM based on Eulerian distance. e The EIM calculation formula is as follows: ； In the formula, superscript k Indicates the number of optimal solutions, subscript m The dimension representing the optimal solution is shown below. Each element of the matrix is ​​an EI function, calculated as follows: ； In the formula, Indicates the first j The first point i The current optimal solution for each objective function value; and These represent the unknown points in the design space of the Kriging surrogate model. x The first i The predicted value and root mean square error of the objective function; and Let represent the cumulative function and probability density function of the standard normal distribution, respectively; Represents a point in the design space x Compared to the current Pareto front solution set, the th j The first point i The expected improvement in the value of the objective function.

13. The optimized design method for the circumferential groove casing treatment of an axial flow pump according to claim 12, characterized in that, The EIM includes a point within the design space. Compared to the improvement expectation of all points in the current Pareto front solution set on all objective functions, the EIM function is quantized by referring to the definition of the improved EI function based on Eulerian distance, resulting in the EIM method based on the improved expectation matrix of Eulerian distance. e ,as follows: 。 14. An optimized design system for the circumferential groove casing treatment of an axial flow pump, characterized in that, An optimized design method for processing the circumferential groove casing of an axial flow pump according to any one of claims 1 to 13 includes: The automatic CFD numerical simulation module is used to perform geometric modeling, mesh generation, boundary condition and physical model setting, and simulation calculation of the fluid domain of an axial flow pump with a circumferential groove casing. The entire process is automatically implemented and called through macro files and batch files. The blade tip blockage quantization calculation module, based on classical boundary layer theory, selects a blockage quantization reference plane perpendicular to the rotation axis and w z Calculate the endwall clogging value, extract and preprocess the flow field data, divide the clogging region into the main flow region, and calculate the value at the outer boundary of the clogging region. Localized blockage values ​​and blade tip region blockage values; The optimization strategy and design problem building module aims to reduce the overall tip blockage value under near-stall conditions as a flow stability optimization strategy, setting the maximum blockage B in the rotor flow channel as the target. max Blockage near the trailing edge of the blade B end Maximize operating efficiency under design conditions η To optimize the target, we construct an optimization design problem for the circumferential groove casing of an axial flow pump. Based on the multi-objective EGO algorithm solution module, an optimal Latin hypercube experimental design is used to arrange the initial sample point set and construct an initial sample database. The Pareto front solution set is calculated using a non-dominated sorting algorithm, and the solution is based on the Kriging surrogate model and EIM. e The selection criteria for calibration points are combined with the differential evolution algorithm to search for calibration points. The sample database and model are iteratively updated until convergence, and the optimal solution set is output.

15. An optimized design device for the treatment of the circumferential groove casing of an axial flow pump, characterized in that, It includes a processor and a memory, the memory storing a computer program, and when the processor executes the computer program, it implements the optimized design method for the circumferential groove casing treatment of an axial flow pump as described in any one of claims 1 to 13.

16. An optimized design storage medium for the circumferential groove casing treatment of an axial flow pump, characterized in that, The storage medium is a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements all the steps of the optimized design method for the circumferential groove casing treatment of the axial flow pump as described in any one of claims 1 to 13.

Images