Optimal design method for cavitation erosion resistance of inducer based on optimal meta-model

Abstract

The invention discloses an inducer cavitation erosion resistance optimization design method based on an optimal meta model, which is applied to the technical field of pump inducer optimization design. Determining an inducer hydraulic performance simulation model by adjusting the coefficient of the cavitation model, and determining a cavitation calculation model by comparing the cavitation prediction error of the existing cavitation calculation model based on the unsteady numerical simulation result of the hydrofoil; on the basis of a small sample test design method, geometric parameters with the influence weight on the cavitation erosion area and the pump performance larger than a preset influence weight threshold value are determined, a hydraulic performance simulation model and a cavitation erosion calculation model are used for prediction, and a sample database is established; and training different meta-models based on the sample database, determining the meta-model with the best prediction precision, and solving the meta-model by using a global optimization algorithm. According to the induction wheel, the cavitation erosion phenomenon of the induction wheel can be effectively avoided, and the cavitation erosion resistance of the induction wheel is improved.

Description

Technical Field

[0001] The invention relates to the technical field of pump inducer optimization design, and in particular to an inducer cavitation resistance optimization design method based on an optimal meta-model. Background Art

[0002] Cavitation is an inevitable hydrodynamic phenomenon in the operation of pump machinery. As cavitation develops, the bubbles in the pump continue to collapse, which in turn causes erosion and damage to the pump components. In order to effectively alleviate and suppress the adverse effects of cavitation, adding an inducer in front of the pump inlet is widely used as an effective way to improve the cavitation performance of the pump. The working principle of the inducer is to accelerate and pressurize the fluid in advance at the front section of the pump inlet, thereby increasing the fluid pressure at the pump inlet and reducing the necessary cavitation margin of the pump.

[0003] Although the inducer can effectively improve the cavitation phenomenon in the pump, as a cavitation compensation component, its internal cavitation and cavitation phenomenon are inevitable, and the service life of the inducer is directly affected by the degree of cavitation. However, the current optimization of the inducer is mainly aimed at cavitation, and the optimization design for cavitation is rare. Therefore, it is of great significance to carry out the optimization design of the inducer's anti-cavitation performance for the long-term stable operation of the pump.

[0004] Therefore, how to provide an inducer anti-cavitation performance optimization design method based on an optimal meta-model that can effectively target the cavitation phenomenon of the inducer and improve the cavitation resistance of the inducer is an issue that technicians in this field urgently need to solve. Summary of the invention

[0005] In view of this, the present invention proposes an optimal design method for the anti-cavitation performance of an inducer based on an optimal meta-model.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] An optimal design method for the cavitation resistance of an inducer based on an optimal meta-model includes:

[0008] Step 1: Based on the turbulence model and cavitation model, the hydraulic performance simulation of the inducer and the unsteady numerical simulation of the hydrofoil are respectively performed, and based on the hydraulic performance simulation results of the inducer, the coefficients of the cavitation model are adjusted to determine the hydraulic performance simulation model of the inducer and the unsteady numerical simulation results of the hydrofoil, and the cavitation calculation model is determined by comparing the cavitation prediction errors of the existing cavitation calculation models;

[0009] Step 2: Based on the small sample test design method, determine the geometric parameters whose influence weights on the cavitation area and pump performance are greater than the preset influence weight threshold, and calculate the inducer hydraulic performance and cavitation area under the geometric parameters through the inducer hydraulic performance simulation model and the cavitation calculation model, and establish a sample database of geometric parameters, inducer hydraulic performance and cavitation area;

[0010] Step 3: Based on the sample database, different meta-models are trained, and the best meta-model is determined according to the prediction accuracy of the meta-model. The geometric parameters are used as optimization variables, and the cavitation area is minimized and the inducer hydraulic performance is maximized as the optimization target. Based on the preset value range of the geometric parameters and the optimization constraints, the best meta-model is solved using the global optimization algorithm;

[0011] Step 4: Use the inducer hydraulic performance simulation model and cavitation calculation model to calculate the inducer hydraulic performance and cavitation area under the solution results, and compare them with the predicted data of the solution results under the optimal meta-model. If the error is greater than the preset prediction error threshold, the predicted data of the optimal meta-model is added to the sample database, and the optimal meta-model is re-determined. If the error is less than the preset prediction error threshold, the optimization ends.

[0012] Optionally, in step 1, the inducer hydraulic performance simulation is performed based on the turbulence model and the cavitation model, and based on the inducer hydraulic performance simulation result, the inducer hydraulic performance simulation model is determined by adjusting the coefficient of the cavitation model, specifically:

[0013] The hydraulic performance of the inducer is simulated based on the RNG turbulence model and the ZGB cavitation model, and the simulation results of the inducer head and efficiency are compared with the test data. If the errors are less than the preset hydraulic performance simulation error threshold, the current model is determined as the inducer hydraulic performance simulation model; if the errors are greater than the preset hydraulic performance simulation error threshold, the evaporation and condensation coefficients of the ZGB cavitation model are adjusted until the errors are less than the preset hydraulic performance simulation error threshold, and the inducer hydraulic performance simulation model is obtained.

[0014] Optionally, in step 1, an unsteady numerical simulation of the hydrofoil is performed based on the turbulence model and the cavitation model, and based on the unsteady numerical simulation results of the hydrofoil, the cavitation calculation model is determined by comparing the cavitation prediction error of the existing cavitation calculation model, specifically:

[0015] Unsteady numerical simulation of the hydrofoil is carried out based on the RNG turbulence model and the ZGB cavitation model, and the cavitation area of ​​the hydrofoil is predicted by the existing cavitation calculation model. The errors between the cavitation area prediction results of different existing cavitation calculation models and the experimental data are compared, and the existing cavitation calculation model with the smallest error is determined as the cavitation calculation model; among them, the existing cavitation calculation models include: intensity function method, time-averaged cavitation erosion estimation method, grayscale method, and energy method.

[0016] Optionally, in step 2, the small sample experiment design method includes: orthogonal experiment method and Plackett-Burman experiment method.

[0017] Optionally, in step 2, the geometric parameters whose influence weights on the cavitation area and pump performance are greater than a preset influence weight threshold include: blade inlet placement angle, blade wrap angle, pitch, and number of blades.

[0018] Optionally, in step 3, different meta-models include: back propagation neural network, response surface method, naive Bayes method, and decision tree method.

[0019] Optionally, in step 3, the best meta-model is determined according to the prediction accuracy of the meta-model, specifically: the best meta-model is determined by comparing the mean square errors of different meta-models.

[0020] Optionally, in step 3, the preset value range of the geometric parameter is specifically:

[0021] The blade inlet placement angle is 5 to 15°;

[0022] The blade wrap angle is 300-380°;

[0023] The pitch is 0.75 to 1.5 times the original value;

[0024] The number of leaves is 2 to 4.

[0025] Optionally, in step 3, the optimization constraint is: the inducer head is within ±1%.

[0026] Optionally, in step 3, the global optimization algorithm includes: genetic algorithm, particle swarm algorithm, fish school algorithm, and annealing algorithm.

[0027] It can be seen from the above technical scheme that compared with the prior art, the present invention proposes an optimization design method for the cavitation resistance performance of the inducer based on the optimal meta-model, which can effectively target the cavitation phenomenon of the inducer and improve the cavitation resistance performance of the inducer. It also takes into account the hydraulic performance on the basis of optimizing the cavitation resistance performance of the inducer, which is of great significance to the long-term stable operation of the pump. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0029] Figure 1 It is a schematic diagram of the method flow of the present invention.

[0030] Figure 2 It is a schematic diagram for comparing the cavitation calculation results based on the grayscale method of the present invention with the experimental data.

[0031] Figure 3 It is a schematic diagram comparing the cavitation areas of the original model and the optimized model of the present invention. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] Embodiment 1:

[0034] Embodiment 1 of the present invention discloses a method for optimizing the anti-cavitation performance of an inducer based on an optimal meta-model, comprising:

[0035] Step 1: Based on the turbulence model and cavitation model, the hydraulic performance simulation of the inducer and the unsteady numerical simulation of the hydrofoil are performed respectively. Based on the simulation results of the hydraulic performance of the inducer, the hydraulic performance simulation model of the inducer and the unsteady numerical simulation results of the hydrofoil are determined by adjusting the coefficients of the cavitation model. The cavitation calculation model is determined by comparing the cavitation prediction errors of the existing cavitation calculation models.

[0036] The inducer hydraulic performance simulation is carried out based on the turbulence model and the cavitation model. Based on the inducer hydraulic performance simulation results, the inducer hydraulic performance simulation model is determined by adjusting the coefficient of the cavitation model. Specifically, it is as follows:

[0037] Based on the RNG turbulence model and the ZGB cavitation model, the hydraulic performance of the inducer is simulated under multiple working conditions, and the simulation results of the inducer head and efficiency are compared with the test data. If the errors are less than the preset hydraulic performance simulation error threshold (the preset hydraulic performance simulation error threshold in the present invention is 5%), the current model is determined as the inducer hydraulic performance simulation model. If the error is greater than the preset hydraulic performance simulation error threshold, the evaporation and condensation coefficients of the ZGB cavitation model are adjusted until the errors are less than the preset hydraulic performance simulation error threshold, and the inducer hydraulic performance simulation model is obtained.

[0038] Based on the turbulence model and cavitation model, the unsteady numerical simulation of the hydrofoil is carried out. Based on the unsteady numerical simulation results of the hydrofoil, the cavitation calculation model is determined by comparing the cavitation prediction error of the existing cavitation calculation model, which is specifically:

[0039] Unsteady numerical simulation of the hydrofoil is carried out based on the RNG turbulence model and the ZGB cavitation model, and the cavitation area of ​​the hydrofoil is predicted by the existing cavitation calculation model. The errors between the cavitation area prediction results of different existing cavitation calculation models and the experimental data are compared, and the existing cavitation calculation model with the smallest error is determined as the cavitation calculation model; among them, the existing cavitation calculation models include: intensity function method, time-averaged cavitation erosion estimation method, grayscale method, and energy method.

[0040] Step 2: Based on the small sample test design method, determine the geometric parameters whose influence weights on the cavitation area and pump performance are greater than the preset influence weight threshold, and calculate the inducer hydraulic performance and cavitation area under the geometric parameters through the inducer hydraulic performance simulation model and the cavitation calculation model, and establish a sample database of geometric parameters, inducer hydraulic performance and cavitation area.

[0041] Before calculating the inducer hydraulic performance and cavitation area under geometric parameters through the inducer hydraulic performance simulation model and the cavitation calculation model, and establishing a sample database of geometric parameters, inducer hydraulic performance and cavitation area, it also includes: performing Latin hypercube sampling on the geometric parameters.

[0042] Small sample experimental design methods include: orthogonal experimental method and Plackett-Burman experimental method.

[0043] The geometric parameters whose influence weights on cavitation area and pump performance are greater than the preset influence weight thresholds include: blade inlet placement angle, blade wrap angle, pitch, and number of blades.

[0044] Step 3: Based on the sample database, different meta-models are trained, and the best meta-model is determined according to the prediction accuracy of the meta-model. The geometric parameters are used as optimization variables, and the cavitation area is minimized and the inducer hydraulic performance is highest (highest efficiency) or only the cavitation area is minimized as the optimization goal. Based on the preset value range of the geometric parameters and the optimization constraints, the best meta-model is solved using the global optimization algorithm.

[0045] Different meta-models, including: back propagation neural network, response surface method, naive Bayes method, decision tree method.

[0046] The best meta-model is determined according to the prediction accuracy of the meta-model, specifically: the best meta-model is determined by comparing the mean square errors of different meta-models.

[0047] The preset value range of geometric parameters is as follows:

[0048] The blade inlet placement angle is 5 to 15°;

[0049] The blade wrap angle is 300-380°;

[0050] The pitch is 0.75 to 1.5 times the original value;

[0051] The number of leaves is 2 to 4.

[0052] The optimization constraint is: the inducer head is within ±1%.

[0053] Global optimization algorithms include: genetic algorithm, particle swarm algorithm, fish school algorithm, and annealing algorithm.

[0054] Step 4: Calculate the hydraulic performance of the inducer and the cavitation area of ​​the solution using the inducer hydraulic performance simulation model and the cavitation calculation model, and compare them with the predicted data of the solution under the optimal meta-model. If the error is greater than the preset prediction error threshold (the preset prediction error threshold in the present invention is 5%), add the predicted data of the optimal meta-model to the sample database, and redefine the optimal meta-model. If the error is less than the preset prediction error threshold, the optimization ends.

[0055] Embodiment 2:

[0056] Embodiment 2 of the present invention discloses a specific application of an optimal meta-model-based method for optimizing the design of inducer cavitation resistance, as follows:

[0057] Application: A three-blade equal pitch inducer with a design flow rate of 8m 3 / h, lift is 5m, speed is 2900r / min.

[0058] Step 1: Using the default parameters of the ZGB cavitation model, the calculated head is 5.3m and the efficiency is 61.2%. The test head is 5.2m and the efficiency is 60.3%. The calculated relative error of the head is 1.9%, and the relative error of the efficiency is 1.5%. The relative error is less than 5%, and the current model is determined as the inducer hydraulic performance simulation model.

[0059] The unsteady numerical simulation of Twist11 hydrofoil is carried out, the inlet velocity is 14m / s, the cavitation number is 1.2, and a total of 10 cycles are calculated. After the error comparison of the existing cavitation calculation model, the grayscale method with an area error of 8.6% is selected as the cavitation calculation model. The grayscale method is as follows:

[0060]

[0061] Where N is the number of moments exceeding the threshold; P d is environmental pressure; P V is the saturated vapor pressure; α V is the steam volume fraction; t is the time interval.

[0062] Comparison chart of cavitation calculation results based on grayscale method and experimental data, such as Figure 2 shown.

[0063] Step 2: The geometric parameters whose influence weights on the cavitation area and pump performance are greater than the preset influence weight threshold are shown in Table 1.

[0064] Table 1 Geometric parameters

[0065] Main geometric parameters Original value Value range 1 Blade inlet placement angle β 10° [5°,15°] 2 Blade wrap angle φ 320° [300°,380°] 3 Pitch d 75mm [56mm,150mm] 4 Number of blades Z 3 [2,4]

[0066] Step 3: After comparing the mean square errors of each meta-model, as shown in Table 2, the back propagation neural network with the lowest mean square error is selected as the best meta-model.

[0067] Table 2 Mean square error of each meta-model

[0068]

[0069]

[0070] The geometric parameters are used as optimization variables, the cavitation area is minimized and the inducer hydraulic performance is the highest (highest efficiency) or only the cavitation area is minimized as the optimization target, and the optimal meta-model is solved using a global optimization algorithm based on the preset value range of the geometric parameters and the optimization constraints. In Example 2 of the present invention, the cavitation area is minimized as the optimization target, and the head is limited to not less than 5m, and the global optimization algorithm selected is a particle swarm algorithm.

[0071] Step 4: Use the inducer hydraulic performance simulation model and cavitation calculation model to calculate the inducer hydraulic performance and cavitation area under the solution results, and compare them with the predicted data of the solution results under the optimal meta-model. The prediction errors of the head, efficiency, and cavitation area are all less than 5%, and the optimization is completed.

[0072] The parameters of the original model and the optimized model are shown in Table 3. The cavitation area of ​​the original model and the optimized model are compared. Figure 3 As shown, the cavitation area of ​​the optimized model is reduced by 17.4% compared with the original model, which proves the superiority of the present invention.

[0073] Table 3 Parameters of the original model and the optimized model

[0074]

[0075] The embodiment of the present invention discloses an optimal meta-model-based design method for optimizing the cavitation resistance of an inducer wheel, which can effectively target the cavitation phenomenon of the inducer wheel and improve the cavitation resistance of the inducer wheel. It also takes the hydraulic performance into consideration on the basis of optimizing the cavitation resistance of the inducer wheel, and is of great significance to the long-term stable operation of the pump.

[0076] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0077] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for optimizing the anti-cavitation performance of an inducer based on an optimal meta-model, characterized in that: include: Step 1: Based on the turbulence model and the cavitation model, the hydraulic performance simulation of the inducer and the unsteady numerical simulation of the hydrofoil are respectively performed, and based on the hydraulic performance simulation results of the inducer, the hydraulic performance simulation model of the inducer and the unsteady numerical simulation results of the hydrofoil are determined by adjusting the coefficients of the cavitation model, and the cavitation calculation model is determined by comparing the cavitation prediction error of the existing cavitation calculation model; Step 2: Based on the small sample test design method, determine the geometric parameters whose influence weights on the cavitation area and pump performance are greater than the preset influence weight threshold, and calculate the inducer hydraulic performance and cavitation area under the geometric parameters through the inducer hydraulic performance simulation model and the cavitation calculation model, and establish a sample database of the geometric parameters, the inducer hydraulic performance and the cavitation area; Step 3: Based on the sample database, different meta-models are trained, and the best meta-model is determined according to the prediction accuracy of the meta-model. The geometric parameters are used as optimization variables, and the cavitation area is minimized and the inducer hydraulic performance is maximized as the optimization target. Based on the preset value range of the geometric parameters and the optimization constraints, the best meta-model is solved using a global optimization algorithm; Step 4: Calculate the inducer hydraulic performance and cavitation area under the solution result by using the inducer hydraulic performance simulation model and the cavitation calculation model, and compare them with the predicted data of the solution result under the optimal meta-model. If the error is greater than the preset prediction error threshold, add the predicted data of the optimal meta-model to the sample database, and redefine the optimal meta-model. If the error is less than the preset prediction error threshold, the optimization ends.

2. The method for optimizing the anti-cavitation performance of an inducer based on an optimal meta-model according to claim 1, characterized in that: In step 1, the hydraulic performance simulation of the inducer is performed based on the turbulence model and the cavitation model, and based on the simulation result of the hydraulic performance of the inducer, the coefficient of the cavitation model is adjusted to determine the hydraulic performance simulation model of the inducer, specifically: The hydraulic performance of the inducer is simulated based on the RNG turbulence model and the ZGB cavitation model, and the simulation results of the inducer head and efficiency are compared with the test data. If the errors are less than the preset hydraulic performance simulation error threshold, the current model is determined as the inducer hydraulic performance simulation model. If the error is greater than the preset hydraulic performance simulation error threshold, the evaporation and condensation coefficients of the ZGB cavitation model are adjusted until the errors are less than the preset hydraulic performance simulation error threshold, and the inducer hydraulic performance simulation model is obtained.

3. The method for optimizing the anti-cavitation performance of an inducer based on an optimal meta-model according to claim 1, characterized in that: In step 1, an unsteady numerical simulation of the hydrofoil is performed based on the turbulence model and the cavitation model. Based on the unsteady numerical simulation results of the hydrofoil, the cavitation calculation model is determined by comparing the cavitation prediction error of the existing cavitation calculation model, which is specifically: Unsteady numerical simulation of the hydrofoil is carried out based on the RNG turbulence model and the ZGB cavitation model, and the cavitation area of ​​the hydrofoil is predicted by using the existing cavitation calculation model. The errors between the cavitation area prediction results of different existing cavitation calculation models and the experimental data are compared, and the existing cavitation calculation model with the smallest error is determined as the cavitation calculation model; wherein, the existing cavitation calculation models include: intensity function method, time-averaged cavitation erosion estimation method, grayscale method, and energy method.

4. The method for optimizing the anti-cavitation performance of an inducer based on an optimal meta-model according to claim 1, characterized in that: In step 2, the small sample test design method includes: orthogonal test method and Plackett-Burman test method.

5. The method for optimizing the anti-cavitation performance of an inducer based on an optimal meta-model according to claim 1, characterized in that: In step 2, the geometric parameters whose influence weights on the cavitation area and pump performance are greater than the preset influence weight thresholds include: blade inlet placement angle, blade wrap angle, pitch, and number of blades.

6. The method for optimizing the design of the inducer cavitation resistance based on the optimal meta-model according to claim 1, characterized in that: In step 3, the different meta-models include: back propagation neural network, response surface method, naive Bayes method, and decision tree method.

7. The method for optimizing the design of the inducer cavitation resistance based on the best meta-model according to claim 1, characterized in that: In step 3, the best meta-model is determined according to the prediction accuracy of the meta-model, specifically: the best meta-model is determined by comparing the mean square errors of different meta-models.

8. The method for optimizing the design of the inducer cavitation resistance based on the best meta-model according to claim 5, characterized in that: In step 3, the preset value range of the geometric parameter is specifically: The blade inlet placement angle is 5 to 15°; The blade wrap angle is 300-380°; The pitch is 0.75 to 1.5 times the original value; The number of the blades is 2 to 4.

9. The method for optimizing the design of the inducer cavitation resistance based on the best meta-model according to claim 8, characterized in that: In step 3, the optimization constraint condition is: the inducer head is within ±1%.

10. The method for optimizing the design of the inducer cavitation resistance based on the best meta-model according to claim 1, characterized in that: In step 3, the global optimization algorithm includes: genetic algorithm, particle swarm algorithm, fish swarm algorithm, and annealing algorithm.

Images