A multi-objective optimization method for axial flow fans based on unsteady calculation

Abstract

The present invention provides a multi-objective optimization method for an axial flow fan based on unsteady calculation, which mainly includes the following steps: First, determine a function regarding efficiency; Second, determine a function regarding pressure pulsation; Third, determine a performance optimization function according to the first two functions; Fourth, use the particle swarm algorithm to optimize the final function, and use the Latin hypercube experimental design method to initialize the distribution of particles; Fifth, during the solution process, use a script program to call three-dimensional modeling, mesh generation, and numerical simulation programs; Sixth, when the optimization is completed, observe the position of the particles. If they are located on the boundary of the feasible region, return to step three to readjust the upper and lower limits of the feasible region and optimize again, so as to obtain the optimal solution. The present invention uses the particle swarm algorithm for automatic multi-objective optimization, realizes three-dimensional modeling, mesh generation, and unsteady numerical calculation of the axial flow fan by calling a script program, and can quickly realize multi-objective performance optimization of the axial flow fan within the feasible region.

Description

Technical Field

[0001] The present invention relates to an optimization method for axial flow fans based on unsteady calculations, and belongs to the technical field of fans Technical Background

[0002] An axial flow fan is an impeller machine that transports air with gas flowing along the axial direction of the fan. It is widely used in occasions where high flow rate and low pressure are required, such as exhaust ventilation and heat dissipation in metallurgy, chemical industry, mechanical equipment, and civil building scenarios. According to relevant data statistics, the actual operating efficiency of domestic fans is about 65% on average, and the power consumption of fans and water pumps accounts for 40%-50% of the total industrial electricity consumption in the country. Therefore, optimizing the fan to improve its performance is of great significance for energy conservation and emission reduction

[0003] The performance of an axial flow fan depends on the efficiency and the magnitude of pressure pulsation. In order to optimize the performance of the fan, unsteady numerical simulation calculations need to be carried out on the fan. To solve the problems of many target parameters of the axial flow fan and the high time consumption of unsteady numerical simulation, the present invention provides a multi-objective optimization method for axial flow fans based on unsteady calculations Summary of the Invention

[0004] The purpose of the present invention is to enable the axial flow fan to achieve rapid performance optimization

[0005] The present invention adopts the following technical solutions to achieve the above technical requirements: A multi-objective optimization method for axial flow fans based on unsteady calculations, including the following steps

[0006] Step 1: The design of the axial flow fan adopts the basic airfoil parametric design method. Select four variables, namely the blade inlet angle αF1, blade outlet angle βF1, blade thickness s, and blade chord length t, which have a greater impact on efficiency. Determine the range within which the design variables can vary. Use the above four parameters as function variables to determine a target function

[0007] f(αF1, βF1, s, t) = η

[0008] Step 2: Select four variables, namely the guide vane inlet angle αF2, guide vane outlet angle βF2, ratio of the number of blades to the number of guide vanes μ, and axial distance l between the blade and the guide vane, which have a greater impact on the pressure pulsation value of the axial flow fan. Determine the range within which the design variables can vary. Use the above four parameters as function variables to determine another target function

[0009] f(αF2, βF2, μ, l) = P

[0010] Step 3: Divide the above two target functions to form a performance optimization function

[0011]

[0012] Taking this function value as the optimization objective and the total pressure air volume as the constraint condition; selecting the blade inlet angle αF1, blade outlet angle βF1, blade thickness s, blade chord length t, guide vane inlet angle αF2, guide vane outlet angle βF2, ratio μ of the number of blades to the number of guide vanes, and axial distance l between the blade and the guide vane as the optimization input values, and determining the upper and lower limits of each variable to form a feasible region.

[0013] Step 4: Optimize the above performance optimization function using the particle swarm algorithm, determine the population size of the particle swarm algorithm, and use the Latin hypercube experimental design method to initialize the distribution of particles in the feasible region. Each particle is a design scheme for different airfoil guide vanes.

[0014] Step 5: Set the number of iteration steps of the particle swarm algorithm. During the solution process of the particle swarm algorithm, the particle swarm algorithm automatically calls the three-dimensional modeling, mesh generation, and numerical simulation programs using a script program and calculates the fitness (performance optimization function value) of each particle.

[0015] Step 6: When the number of iteration steps reaches the set value, the optimization ends. Observe the position of the particle in the population. If the particle position is on the boundary of the feasible region, return to Step 3 to readjust the upper and lower limits of the feasible region and optimize again to finally obtain the optimal airfoil guide vane scheme.

[0016] In the above scheme, the airfoil parameter design method is used to determine an efficiency-related function with the four values of the blade inlet angle αF1, blade outlet angle βF1, blade thickness s, and blade chord length t as variables.

[0017] In the above scheme, four values of the guide vane inlet angle αF2, guide vane outlet angle βF2, ratio μ of the number of blades to the number of guide vanes, and axial distance l between the blade and the guide vane, which have a greater impact on pressure pulsation, are selected as variables to determine a function related to pressure pulsation.

[0018] In the above scheme, the pressure pulsation objective function and the efficiency objective function are divided to obtain a performance optimization function. The variables of this function are the eight values of the blade inlet angle αF1, blade outlet angle βF1, blade thickness s, blade chord length t, guide vane inlet angle αF2, guide vane outlet angle βF2, ratio μ of the number of blades to the number of guide vanes, and axial distance l between the blade and the rear guide vane.

[0019] In the above scheme, the script program called by the particle swarm algorithm can be written in languages such as MATLAB, LabVIEW, Python, etc.

[0020] In the above scheme, the three-dimensional modeling can be completed using software such as CFturbo, NX UG, Creo, BladeGen, etc., and the mesh generation can be realized using software such as ICEM, TurboGrid, STAR-ccm+, mesh, etc.

[0021] In the above solution, unsteady numerical calculations can be implemented using software such as CFX, Fluent, STAR-ccm+, Numeca, etc.

[0022] In the above solution, during the iteration process, the performance optimization function values of each particle in each step of the solution are the individual optimal values of each particle. The optimal solution of the population can be obtained in this step. At this time, the velocity and position of the particles are updated and the iteration continues. If the optimal solution appears at the boundary of the particle feasible domain, the feasible domain needs to be readjusted and the iteration starts until the optimal solution in the feasible domain, that is, the optimal solution of the airfoil guide vane design, is obtained.

[0023] The beneficial effects of the present invention are as follows: A multi-objective optimization method for axial flow fans based on unsteady calculations is provided. The present invention uses the particle swarm algorithm for automatic multi-objective optimization, and realizes the three-dimensional modeling, mesh generation, and unsteady numerical calculation of axial flow fans by calling script programs. It can quickly achieve the multi-objective performance optimization of axial flow fans within the range of the design variable feasible domain, with high calculation accuracy and good optimization efficiency. Description of the Drawings

[0024] Figure 1 It is a flowchart of a multi-objective optimization method for axial flow fans based on unsteady calculations.

[0025] Figure 2 It is a three-dimensional view of the impeller of an axial flow fan, where the number of blades is 7 and the number of guide vanes is 9. Detailed Embodiments

[0026] The present invention will be further described below in conjunction with the drawings and specific embodiments, but the protection scope of the present invention is not limited thereto.

[0027] The optimization method adopted by the present invention is as Figure 1 shown. Taking the ratio of the pressure pulsation to the efficiency of the axial flow fan as the optimization objective, a fast optimization method for the unsteady performance of an axial flow fan includes the following steps:

[0028] Embodiment: Initial data of the axial flow fan: (1) Airfoil part: Blade inlet angle 90°, blade outlet angle 20.3°, blade thickness 4 mm, blade chord length 1.2 m; (2) Guide vane part: Guide vane inlet angle 33.7°, guide vane outlet angle 90°, blade thickness 2.8 mm, blade chord length 0.42 m, ratio of the number of blades to the number of guide vanes 0.78, axial distance between the blade and the rear guide vane 35 cm.

[0029] (1) Select the parameters that have a greater impact on the airfoil efficiency as variables according to the airfoil parametric design experience and determine an objective function regarding efficiency;

[0030] (2) Select four values that have a greater impact on the pressure pulsation as variables to determine a function related to the pressure pulsation;

[0031] (3) Divide the pressure pulsation objective function by the efficiency objective function to obtain the performance optimization function;

[0032] (4) Select eight values, namely the blade inlet angle αF1, the blade outlet angle βF1, the blade thickness s, the blade chord length t, the inlet guide vane angle αF2, the outlet guide vane angle βF2, the ratio μ of the number of blades to the number of guide vanes, and the axial distance l between the blade and the rear guide vane, as design variables according to the performance optimization function and the axial flow fan design experience. Select the performance optimization function value as the optimization objective, and determine the specific upper and lower limits Max and Min of the design variables as shown in the following table.

[0033]

[0034] (5) Determine the number of particle populations, use the Latin hypercube experimental design method to initialize the distribution of particles in the feasible domain, set the number of iteration steps to 500 steps, and use the particle swarm algorithm to start optimization in the feasible domain.

[0035] (6) Write a script program using LabVIEW during the algorithm optimization process, automatically call the CFturbo software to perform 3D modeling of the airfoil guide vane, use the ICEM software for meshing, and use the STAR-ccm+ software to perform unsteady numerical calculations for the design conditions using the harmonic balance method.

[0036] (7) When the number of iteration steps reaches and the optimization is completed, output the optimized model solution: the blade inlet angle αF1 = 90°, the blade outlet angle βF1 = 34.5°, the blade thickness s = 3.8 mm, the blade chord length t = 1.25 m, the inlet guide vane angle αF2 = 34.5°, the outlet guide vane angle βF2 = 90°, the ratio μ of the number of blades to the number of guide vanes = 0.78, and the axial distance l between the blade and the rear guide vane = 37 cm. Compare the multi-objective optimized model with the original model. Finally, the total pressure efficiency of the axial flow fan is increased by 5.7 percentage points compared with that before optimization, and the inlet noise value of the fan is reduced by 5.8 dB.

Claims

1. A multi-objective optimization method for axial flow fans based on unsteady calculations, comprising the following steps: Step 1: In the design of the axial flow fan, the basic airfoil parametric design method is adopted. Four variables, namely the blade inlet angle αF1, blade outlet angle βF1, blade thickness s, and blade chord length t, which have a greater impact on efficiency, are selected. The range within which the design variables can vary is determined, and the above four parameters are used as function variables to determine an objective function: f(αF1, βF1, s, t) = η Step 2: Four variables, namely the guide vane inlet angle αF2, guide vane outlet angle βF2, ratio μ of the number of blades to the number of guide vanes, and axial distance l between the blade and the guide vane, which have a greater impact on the pressure pulsation value of the axial flow fan, are selected. The range within which the design variables can vary is determined, and the above four parameters are used as function variables to determine another objective function: f(αF2, βF2, μ, l) = P Step 3: Divide the two objective functions in Step 1 and Step 2 to form a performance optimization function Take this function value as the optimization objective, with the full-pressure air volume as the constraint condition; The blade inlet angle αF1, blade outlet angle βF1, blade thickness s, blade chord length t, guide vane inlet angle αF2, guide vane outlet angle βF2, ratio μ of the number of blades to the number of guide vanes, and axial distance l between the blade and the guide vane are selected as the optimization input values, and the upper and lower limits of each variable are determined to form a feasible region; Step 4: Use the particle swarm algorithm to optimize the performance optimization function in Step 3, determine the population size of the particle swarm algorithm, and use the Latin hypercube experimental design method to initialize the distribution of particles in the feasible region. Each particle is a design scheme for different airfoil guide vanes; Step 5: Set the number of iteration steps of the particle swarm algorithm. During the solution process of the particle swarm algorithm, the particle swarm algorithm automatically calls the three-dimensional modeling, mesh generation, and numerical simulation programs using a script program and calculates the performance optimization function value of each particle; Step 6: When the number of iteration steps reaches the set value, the optimization ends. Observe the position of the particle in the population. If the particle position is on the boundary of the feasible region, return to Step 3 to readjust the upper and lower limits of the feasible region and perform optimization again to finally obtain the optimal airfoil guide vane scheme.

2. The multi-objective optimization method for an axial flow fan based on unsteady calculation according to claim 1, wherein: In Step 4, the particle swarm algorithm is replaced by a hybrid gravitational search algorithm or a bat algorithm.

3. The multi-objective optimization method of an axial flow fan based on unsteady calculation according to claim 1, characterized in that: The script program in Step 5 is written in MATLAB, LabVIEW, or Python language.

4. The multi-objective optimization method of an axial flow fan based on unsteady calculation according to claim 1, characterized in that: The three-dimensional modeling in Step 5 is completed using software such as CFturbo, NX UG, Creo, or BladeGen.

5. The multi-objective optimization method of an axial flow fan based on unsteady calculation according to claim 1, wherein: The mesh generation in Step 5 is implemented using software such as ICEM, TurboGrid, STAR-ccm+, or mesh.

6. The multi-objective optimization method of an axial flow fan based on unsteady calculation according to claim 1, characterized in that: The numerical calculation method is completed using numerical calculations based on time steps or frequency-domain unsteady numerical calculations using the harmonic balance method.

7. The multi-objective optimization method for axial flow fans based on unsteady calculation according to claim 6, characterized in that: The numerical calculation method based on time steps uses software such as CFX or fluent, and the frequency-domain unsteady numerical calculation using the harmonic balance method uses software such as STAR-ccm+ or Numeca.

Images