Multi-target and multi-component integrated optimization method and equipment for counter-rotating axial flow fan

Abstract

The invention discloses a multi-target and multi-component integrated optimization method and equipment for a counter-rotating axial flow fan, and belongs to the technical field of turbomachinery optimization design. According to the method, parametric modeling is carried out on multiple components such as two rows of blades, an inlet current collector and an outlet diffuser of the contra-rotating axial flow fan, geometric parameter combinations of all the components are selected as multi-dimensional decision variables, and a multi-dimensional decision space is determined; obtaining decision sample sets according to the multi-dimensional decision space, solving a plurality of aerodynamic performances of the counter-rotating fan corresponding to different decision sample sets through numerical simulation, and forming a multi-target aerodynamic sample set; combining a genetic algorithm to optimize a weight and a threshold to train a neural network agent model; performing global search and nonlinear optimization through a genetic algorithm based on a neural network agent model by taking the optimal multi-target aerodynamic performance as an optimization target, and outputting a Pareto optimal solution set; and determining a global optimal solution and carrying out numerical simulation verification to obtain an optimized geometric model of the counter-rotating fan. The pneumatic performance of the contra-rotating axial flow fan can be effectively improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of impeller machinery optimization design, and in particular relates to a multi-objective and multi-component integrated optimization method and equipment for a counter-rotating axial flow fan. Background Art

[0002] The counter-rotating axial flow fan is a special axial flow fan composed of two-stage impellers rotating in opposite directions with the same design pressure, hub diameter, impeller outer diameter and impeller speed. The counter-rotating axial flow fan can effectively ventilate and collect and discharge dust to maintain a good environment; at the same time, the counter-rotating axial flow fan has a simple structure and no guide vanes, which makes its overall structure compact and easy to install and move.

[0003] Due to the advantages of high efficiency, high single-stage pressure ratio, flexible use and good anti-wind performance, counter-rotating axial flow fans are mainly used in mines, industrial sectors that transport flammable gases and coal dust, and are also used for ventilation and air exchange in tunnels and ships, as well as in cooling towers, wind tunnels and boilers.

[0004] Although the superiority of counter-rotating axial flow fans has led to an increasingly wide range of applications, the counter-rotating axial flow fans currently used in various industries generally have low efficiency and a narrow range of stable operating conditions. Most of the existing counter-rotating axial flow fan design optimization methods are summarized by combining manual experience, experiments and simulations to study the impact of certain parameters on performance. They are often limited by the setting of initial parameters and the test range, making it difficult to ensure that the global optimal solution is found. In addition, the optimization results may be affected by assumptions and accumulated errors, making it difficult to improve the comprehensive performance of counter-rotating axial flow fans through existing methods. Therefore, it is very difficult for existing methods to handle multi-parameter optimization problems, and there is an urgent need to establish an intelligent optimization method system for counter-rotating axial flow fans. Summary of the invention

[0005] The object of the present invention is to provide a multi-objective and multi-component integrated optimization method and device for a counter-rotating axial flow fan, so as to solve the technical problem that existing methods have difficulty in dealing with multi-parameter optimization problems.

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

[0007] The present invention discloses a multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan, comprising the following steps:

[0008] Obtain the optimization target of the counter-rotating axial flow fan;

[0009] Parametric modeling is performed on multiple components of a counter-rotating axial flow fan, geometric parameter combinations of the multiple components are selected through parametric modeling, and a multidimensional decision space is determined based on the geometric parameter combinations of the multiple components as multidimensional decision variables;

[0010] An integrated sampling is performed in the multi-dimensional decision space to obtain a decision sample set; then the aerodynamic performance of the counter-rotating axial flow fan corresponding to the decision sample set is solved to form a multi-objective aerodynamic sample set;

[0011] With the decision sample set as the input layer and the aerodynamic sample set corresponding to the decision sample set as the output layer, the genetic algorithm is used to optimize the weights and thresholds of the neural network, determine the hidden layer structure, and obtain the optimized neural network proxy model;

[0012] According to the determined optimization goal, the genetic algorithm is used to perform global search and nonlinear optimization in the multidimensional decision space based on the optimization neural network agent model, and the Pareto optimal solution set is output;

[0013] The global optimal solution is determined from the Pareto optimal solution set and verified by numerical simulation, thereby obtaining the geometric model of the optimized counter-rotating axial flow fan.

[0014] Furthermore, the components of the counter-rotating axial flow fan are parametrically modeled, and the components of the counter-rotating axial flow fan include two rows of blades, an inlet collector and an outlet diffuser.

[0015] Furthermore, parameter modeling of two rows of blades of the counter-rotating axial flow fan includes the following steps:

[0016] The end wall is defined by fitting the wheel disc and hub profiles of the two rows of blades with Bezier curves;

[0017] Define the cross section from the root to the tip of the two rows of blades to define the flow surface;

[0018] The stacking and bending rules of the two rows of blades are defined, and then the blade shape is defined to complete the parametric modeling of the two rows of blades of the counter-rotating axial flow fan.

[0019] Furthermore, parameterized modeling of the inlet collector of the counter-rotating axial flow fan is performed, including the following steps:

[0020] In the meridian flow surface, a straight line, a Bezier curve or a B-spline curve is selected to fit the inlet collector flow channel shape; or a composite line consisting of a straight line, a Bezier curve or a B-spline curve is used to perform segmented fitting on the complex inlet collector flow channel shape.

[0021] Furthermore, parameterized modeling of the outlet diffuser of the counter-rotating axial flow fan is performed, including the following steps:

[0022] In the meridian flow surface, a straight line, Bezier curve or B-spline curve is selected to fit the outlet diffuser flow path shape; or a composite line composed of a straight line, Bezier curve or B-spline curve is used to perform segmented fitting on the complex outlet diffuser flow path line shape.

[0023] Furthermore, obtaining the decision sample set according to the multidimensional decision space is to obtain the decision sample set by performing integrated Latin hypercube sampling in the multidimensional decision space.

[0024] Furthermore, the number of the decision sample sets is not less than 20 times the number of decision variables.

[0025] Furthermore, the specific steps of performing integrated Latin hypercube sampling in the multidimensional decision space to obtain a decision sample set are:

[0026] Latin hypercube sampling is performed in the multidimensional decision space and the sampling number is determined. Latin hypercube sampling divides each decision space into several intervals with equal probability based on uniform distribution. A point is randomly selected from several intervals corresponding to each decision space as the initial sampling point to reflect the characteristics of the multidimensional parameter space. Several initial sampling points of each decision space are randomly arranged to break the correlation. Finally, one point is selected from several initial sampling points in each space to form a multidimensional decision variable parameter combination that meets the sample number, namely, a decision sample set.

[0027] Furthermore, the aerodynamic performance of the counter-rotating axial flow fan corresponding to the decision sample set is solved by numerical simulation;

[0028] The settings of the numerical simulation include:

[0029] The calculation area of ​​the numerical simulation includes the collector, two rows of counter-rotating impellers and diffuser of the counter-rotating axial flow fan;

[0030] The motion coordinate system and mixing plane model are used to describe the motion characteristics of the flow field inside the counter-rotating axial flow fan.

[0031] Use the NS governing equations, combined with the Reynolds average method and the RNG k-ε turbulence model to simulate the behavior of the turbulent flow field;

[0032] The boundary condition of the collector inlet section is mass flow inlet, and the boundary condition of the diffuser outlet section is pressure outlet.

[0033] The present invention also discloses a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan as described above are implemented.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] The present invention discloses a multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan. The optimization target is determined according to the specific needs of users in view of the performance deficiencies of the original counter-rotating axial flow fan, and parameterized modeling is performed on the multiple components of the counter-rotating axial flow fan. The geometric parameter combination of each component is selected as a multi-dimensional decision variable and the multi-dimensional decision space is further determined; integrated Latin hypercube sampling is performed in the multi-dimensional decision space to obtain a decision sample set, and multiple aerodynamic performances of the counter-rotating axial flow fan corresponding to different decision sample sets are solved by numerical simulation to form a multi-objective aerodynamic sample set; the decision sample set, i.e., the geometric parameter combination of each component of the counter-rotating axial flow fan, is used as the input layer, and the corresponding aerodynamic sample set, i.e., the multi-objective aerodynamic performance of the counter-rotating axial flow fan, is used as the output layer, and a genetic algorithm is used to optimize the weights and thresholds of the neural network to determine the hidden layer structure; the optimal multi-objective aerodynamic performance of the counter-rotating axial flow fan is used as the optimization target, and a global search is performed by a genetic algorithm based on a neural network proxy model. and nonlinear optimization, and output the Pareto optimal solution set; determine the global optimal solution and perform numerical simulation verification to obtain a combination of geometric parameters of each component that is better than the original counter-rotating axial flow fan in multiple optimization objectives, thereby obtaining the geometric model of the optimized counter-rotating axial flow fan; this optimization method has a complete and intelligent optimization system, and parametric modeling converts the complex geometric structure of the counter-rotating axial flow fan into a series of easy-to-operate parameters, and can achieve precise control of its key features, making the intelligent optimization method more feasible. The neural network can learn the complex nonlinear relationship between the performance of the counter-rotating axial flow fan and the decision variables through a large amount of training data, thereby guiding the genetic algorithm to perform more effective optimization search. As a global optimization algorithm, the genetic algorithm allows the exploration of non-traditional design schemes and helps find the optimal improvement scheme, while improving the efficiency of the counter-rotating axial flow fan, broadening the range of stable operating conditions, and achieving multi-objective optimization, saving time and cost.

[0036] Furthermore, taking the decision sample set, i.e., the geometric parameter combination of each component of the counter-rotating axial flow fan, as the input layer, and the corresponding aerodynamic sample set, i.e., the multi-objective aerodynamic performance of the counter-rotating axial flow fan, as the output layer, the neural network proxy model is trained in combination with the genetic algorithm to optimize the weights and thresholds, which can significantly improve the accuracy and efficiency of the model in predicting the multi-objective aerodynamic performance of the counter-rotating axial flow fan. The genetic algorithm reduces the risk of falling into local optimality through the global search mechanism, so that the neural network can better adapt to the complex flow field characteristics of the fan. The optimization process has strong adaptability and exploration ability, and can effectively find the geometric parameter combination of each component that is better than the original counter-rotating axial flow fan in multiple optimization objectives, thereby improving the prediction performance and overall robustness of the proxy model, while reducing training time and computational cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a flow chart of the multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan of the present invention;

[0038] Figure 2 A schematic diagram of decision variables selected after parametric modeling of the original counter-rotating axial flow fan blades in an embodiment of the present invention;

[0039] Figure 3 Schematic diagram of performance comparison of counter-rotating axial flow fans before and after optimization by the present invention in an embodiment of the present invention. DETAILED DESCRIPTION

[0040] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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 should fall within the scope of protection of the present invention.

[0041] The present invention discloses a multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan, comprising the following steps:

[0042] Step 1: Determine multiple optimization targets based on the specific needs of users in view of the performance deficiencies of the counter-rotating axial flow fan;

[0043] Step 2, parametric modeling is performed on multiple components such as two rows of blades, inlet collector, outlet diffuser, etc. of the counter-rotating axial flow fan, and the geometric parameter combination of each component is selected as a multidimensional decision variable and the multidimensional decision space is further determined;

[0044] Step 3, performing integrated sampling in the multidimensional decision space to obtain a decision sample set, solving multiple aerodynamic performances of the counter-rotating axial flow fan corresponding to different decision sample sets through numerical simulation, and forming a multi-objective aerodynamic sample set;

[0045] Step 4, using the decision sample set as input and the corresponding multi-objective aerodynamic sample set as output, the neural network proxy model is trained by combining the genetic algorithm to optimize the weights and thresholds;

[0046] Step 5, taking the optimal multi-objective aerodynamic performance of the counter-rotating axial flow fan as the optimization target, a global search and nonlinear optimization are performed through a genetic algorithm based on a neural network agent model to output a Pareto optimal solution set;

[0047] Step 6, determine the global optimal solution and perform numerical simulation verification to obtain a combination of geometric parameters of each component that is better than the original counter-rotating axial flow fan in multiple optimization objectives, thereby obtaining a geometric model of the optimized counter-rotating axial flow fan.

[0048] Preferably, in step 2, parameterizing the two rows of counter-rotating blades includes the following aspects: end wall definition, flow surface definition, stacking and sweep law definition, and blade profile definition.

[0049] Preferably, in step 2, parameterizing the inlet collector includes: in the meridian flow surface, selecting a straight line, Bezier curve or B-spline curve to fit the collector flow channel shape; or using a composite line composed of a straight line, Bezier curve or B-spline curve to perform segmented fitting on the complex collector flow channel shape.

[0050] Preferably, in step 2, parameterizing the outlet diffuser includes: in the meridian flow surface, selecting a straight line, a Bezier curve or a B-spline curve to fit the diffuser flow path shape; or using a composite line shape composed of a straight line, a Bezier curve or a B-spline curve to perform segmented fitting on the complex diffuser flow path line shape.

[0051] Preferably, the two rows of moving blades are optimized integrally or separately, and the inlet collector and the outlet diffuser are optimized integrally or separately; or the two rows of moving blades, the inlet collector, and the outlet diffuser are optimized integrally.

[0052] Preferably, in step 3, Latin hypercube sampling is performed in the multidimensional decision space to obtain a decision sample set, and the number of decision sample sets is generally not less than 20 times the number of decision variables; multiple aerodynamic performances of counter-rotating axial flow fans corresponding to different decision sample sets are solved by numerical simulation to form a multi-objective aerodynamic sample set.

[0053] Preferably, in step 4, in the process of training the neural network proxy model, the decision sample set, i.e., the combination of geometric parameters of the components of the counter-rotating axial flow fan, is used as the input layer, and the corresponding aerodynamic sample set, i.e., the multi-objective aerodynamic performance of the counter-rotating axial flow fan, is used as the output layer. A genetic algorithm is used to optimize the weights and thresholds of the neural network to determine the hidden layer structure.

[0054] Preferably, in step 5, the optimal multi-objective aerodynamic performance of the counter-rotating axial flow fan can be the best aerodynamic performance under multiple working conditions or the best multiple aerodynamic performance under the same working condition. When the aerodynamic performance of the counter-rotating axial flow fan is improved by optimizing the blades of the counter-rotating axial flow fan, it usually refers to the former.

[0055] Preferably, in steps 5 and 6, a global search and nonlinear optimization are performed through a genetic algorithm based on a neural network proxy model, and a Pareto optimal solution set is output. The global optimal solution is determined and numerical simulation verification is performed to obtain a combination of geometric parameters of each component that is better than the original counter-rotating axial flow fan in multiple optimization objectives, thereby obtaining the geometric model parameters of the optimized counter-rotating axial flow fan.

[0056] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0057] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0058] like Figure 1 As shown, the present invention discloses a multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan, comprising the following steps:

[0059] Step 1: Determine the optimization target (such as total pressure efficiency, blade channel efficiency, etc.) according to the specific needs of users in view of the performance deficiencies of the counter-rotating axial flow fan;

[0060] Step 2: Parametric modeling is performed on multiple components such as two rows of blades, inlet collector, and outlet diffuser of the counter-rotating axial flow fan, and the geometric parameter combination of each component is selected as a multidimensional decision variable and the multidimensional decision space is further determined;

[0061] Step 3: Perform integrated sampling in the multidimensional decision space to obtain a decision sample set. The number of decision samples is generally not less than 20 times the number of decision variables. Through numerical simulation, multiple aerodynamic performances of counter-rotating axial flow fans corresponding to different decision sample sets are obtained to form a multi-objective aerodynamic sample set.

[0062] Step 4, using the decision sample set, i.e., the geometric parameter combination of each component of the counter-rotating axial flow fan, as the input layer, and the corresponding aerodynamic sample set, i.e., the multi-objective aerodynamic performance of the counter-rotating axial flow fan, as the output layer, using a genetic algorithm to optimize the weights and thresholds of the neural network to determine the hidden layer structure;

[0063] Step 5, taking the optimal multi-objective aerodynamic performance of the counter-rotating axial flow fan as the optimization target (i.e., fitness function), a global search and nonlinear optimization are performed through a genetic algorithm based on a neural network agent model to output a Pareto optimal solution set;

[0064] Step 6, determine the global optimal solution and perform numerical simulation verification to obtain a combination of geometric parameters of each component that is better than the original counter-rotating axial flow fan in multiple optimization objectives, thereby obtaining new geometric parameters of the components to be optimized of the optimized counter-rotating axial flow fan, and realizing multi-component and multi-objective integrated optimization of the counter-rotating axial flow fan, thereby improving the efficiency of the counter-rotating axial flow fan, broadening the range of stable operating conditions, and saving time and cost.

[0065] Preferably, the main process of parameterizing two rows of counter-rotating blades is as follows:

[0066] S1: End wall definition

[0067] The number of control points can be customized by fitting the wheel disc and hub profile through Bezier curves, and each control point is represented by Z and R coordinates;

[0068] S2: Stream surface definition

[0069] Define the section from the blade root to the blade tip, and you can customize the section position, such as 5%, 25%, 50%, 75%, 99.5% of the blade height;

[0070] S3: Definition of stacking and bending rules

[0071] Leading edge stacking, trailing edge stacking, centroid stacking, etc. can be used, and the position and shape of the leading and trailing edge lines can be defined by Bezier curves;

[0072] S4: Leaf Type Definition

[0073] There are two ways to define the blade profile: one is to give the thickness distribution value of the blade from the leading edge to the trailing edge, combined with the shape of the mid-arc line, to generate a blade cross-section; the other is to use control points to control the pressure surface and the suction surface respectively, combined with the shape of the mid-arc line, to generate a blade cross-section; usually an n-order Bezier curve with n+1 control points is used to fit the mid-arc line of each blade cross-section of the counter-rotating axial flow fan blade.

[0074] Preferably, after the parametric modeling of the two rows of counter-rotating axial flow fan blades is completed, the geometric parameters that have a greater impact on the optimization target can be selected as decision variables. For example, the mid-arc lines of the five cross-sections taken along the blade height of the two rows of counter-rotating blades are fitted respectively through the four control points of the third-order Bezier curve, and the three parameters of the distances H1 and H2 from the second and third control points of each cross-section to the chord length and the installation angle Gama are obtained. A total of 30 geometric parameters are used as decision variables, and the decision space of the 30 decision variables is determined.

[0075] Preferably, Latin hypercube sampling is performed in a 30-dimensional decision space to obtain a decision sample set containing 600 decision samples.

[0076] Preferably, the total pressure efficiency of the counter-rotating axial flow fan under three working conditions of small flow, design flow and large flow corresponding to 600 decision samples is obtained through numerical simulation to form an aerodynamic sample set. The settings of the simulation solution can be referred to as follows: the single flow channel of the counter-rotating axial flow fan is simulated, and the calculation area mainly includes the collector, the motor part, two rows of counter-rotating impellers and the diffuser. The impeller is the rotating area, and the other areas are stationary; the motion coordinate system and the mixed plane model are used to describe the motion characteristics of the internal flow field of the fan; the NS control equation group is used, combined with the Reynolds average method (RANS) and the RNG k-ε turbulence model, to accurately simulate the behavior of the turbulent flow field; the SIMPLE separation algorithm is used for iterative solution to deal with the velocity and pressure coupling problems in the flow field; the boundary condition of the collector inlet section is the mass flow inlet, and the boundary condition of the diffuser outlet section is the pressure outlet.

[0077] Preferably, in the process of training the neural network proxy model, the decision sample set, i.e., the combination of geometric parameters of the components of the counter-rotating axial flow fan, is used as the input layer, and the corresponding aerodynamic sample set, i.e., the multi-objective aerodynamic performance of the counter-rotating axial flow fan, is used as the output layer. A mapping relationship between the input and output is established, and a genetic algorithm is used to optimize the weights and thresholds of the neural network, determine the hidden layer structure, and improve the generalization prediction ability of the neural network model.

[0078] Preferably, a sample containing 30 decision variables is encoded into individuals of a genetic algorithm to generate an initial population. Each individual represents a combination of a set of blade geometric parameters. The total pressure efficiency of the counter-rotating axial flow fan under three working conditions corresponding to each individual is predicted through a trained neural network proxy model. This is used as the value of the fitness function. Through a series of genetic operations such as selection, crossover, and mutation, the old generation of individuals is continuously replaced by the new generation of individuals to achieve the maximum number of iterations or the fitness value change is less than a certain value as the termination condition, and the Pareto optimal solution set is output.

[0079] Preferably, the global optimal solution is determined from the Pareto solution set and verified by numerical simulation, and a set of 30 decision variables are obtained whose total pressure efficiency of the counter-rotating axial flow fan is better than that of the original counter-rotating axial flow fan under three working conditions of small flow rate, design flow rate and large flow rate, including H1, H2 and Gama of different cross-sections of the two rows of counter-rotating blades after optimization, that is, the new two rows of blades of the optimized counter-rotating axial flow fan are obtained, realizing multi-objective optimization of the counter-rotating axial flow fan.

[0080] Example 1

[0081] This embodiment uses the above optimization method to optimize the original counter-rotating axial flow fan, and the specific process is as follows:

[0082] Given a C4 airfoil counter-rotating axial flow fan, the working medium is air, and its geometric and aerodynamic parameters are as follows: Design flow rate 44000m 3 / h, design speed ±2950r / min, outlet static pressure 6400Pa, impeller outer diameter 800mm, hub ratio 0.7, the number of blades in the first and second rows are 15 and 10 respectively, and the total pressure efficiency under the design condition is 84.91%; the two rows of blades of the original counter-rotating axial flow fan are now subjected to multi-objective optimization according to the above method, with the maximum total pressure efficiency under small flow, design flow and large flow as the optimization target;

[0083] Figure 2 Schematic diagram of the decision variables selected after parametric modeling of the original counter-rotating axial flow fan blades. Figure 3 The performance comparison curve of the counter-rotating axial flow fan before and after optimization. It can be seen that the total pressure efficiency of the counter-rotating axial flow fan at small flow, design flow and large flow after optimization is increased by 0.59%, 0.88% and 1.21% respectively, and the stable operating range is significantly widened. The multi-objective optimization effect of the counter-rotating axial flow fan is remarkable.

[0084] A terminal device is provided in one embodiment of the present invention. The terminal device of this embodiment includes: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps in the above-mentioned method embodiments are implemented. Alternatively, when the processor executes the computer program, the functions of the modules / units in the above-mentioned device embodiments are implemented.

[0085] The computer program may be divided into one or more modules / units, and the one or more modules / units are stored in the memory and executed by the processor to accomplish the present invention.

[0086] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0087] The processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0088] The memory may be used to store the computer programs and / or modules, and the processor implements various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory.

[0089] If the module / unit integrated in the terminal device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0090] The above contents are only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan, characterized in that: The following steps are involved: Obtain the optimization target of the counter-rotating axial flow fan; Parametric modeling is performed on multiple components of a counter-rotating axial flow fan, geometric parameter combinations of the multiple components are selected through parametric modeling, and a multidimensional decision space is determined based on the geometric parameter combinations of the multiple components as multidimensional decision variables; An integrated sampling is performed in the multi-dimensional decision space to obtain a decision sample set; then the aerodynamic performance of the counter-rotating axial flow fan corresponding to the decision sample set is solved to form a multi-objective aerodynamic sample set; With the decision sample set as the input layer and the aerodynamic sample set corresponding to the decision sample set as the output layer, the genetic algorithm is used to optimize the weights and thresholds of the neural network, determine the hidden layer structure, and obtain the optimized neural network proxy model; According to the determined optimization goal, the genetic algorithm is used to perform global search and nonlinear optimization in the multidimensional decision space based on the optimization neural network agent model, and the Pareto optimal solution set is output; The global optimal solution is determined from the Pareto optimal solution set and verified by numerical simulation, thereby obtaining the geometric model of the optimized counter-rotating axial flow fan.

2. The multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan according to claim 1, characterized in that: Components of a counter-rotating axial flow fan are parametrically modeled, and the components of the counter-rotating axial flow fan include two rows of blades, an inlet collector, and an outlet diffuser.

3. The multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan according to claim 2, characterized in that: Parametric modeling of two rows of blades for a counter-rotating axial fan includes the following steps: The end wall is defined by fitting the wheel disc and hub profiles of two rows of blades through Bezier curves; Define the cross section from the root to the tip of the two rows of blades to define the flow surface; The stacking and bending rules of the two rows of blades are defined, and then the blade shape is defined to complete the parametric modeling of the two rows of blades of the counter-rotating axial flow fan.

4. The multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan according to claim 2, characterized in that: The parametric modeling of the inlet collector of the counter-rotating axial flow fan includes the following steps: In the meridian flow surface, a straight line, a Bezier curve or a B-spline curve is selected to fit the inlet collector flow channel shape; or a composite line consisting of a straight line, a Bezier curve or a B-spline curve is used to perform segmented fitting on the complex inlet collector flow channel shape.

5. The multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan according to claim 2, characterized in that: The parametric modeling of the outlet diffuser of a counter-rotating axial flow fan includes the following steps: In the meridian flow surface, a straight line, Bezier curve or B-spline curve is selected to fit the outlet diffuser flow path shape; or a composite line consisting of a straight line, Bezier curve or B-spline curve is used to perform segmented fitting on the complex outlet diffuser flow path line shape.

6. The multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan according to claim 1, characterized in that: The method of obtaining the decision sample set according to the multidimensional decision space is to perform integrated Latin hypercube sampling in the multidimensional decision space to obtain the decision sample set.

7. A multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan according to claim 6, characterized in that: The number of the decision sample sets is not less than 20 times the number of decision variables.

8. The multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan according to claim 6, characterized in that: The specific steps of performing integrated Latin hypercube sampling in the multidimensional decision space to obtain a decision sample set are: Latin hypercube sampling is performed in the multidimensional decision space and the sampling number is determined. Latin hypercube sampling divides each decision space into several intervals with equal probability based on uniform distribution. A point is randomly selected from several intervals corresponding to each decision space as the initial sampling point to reflect the characteristics of the multidimensional parameter space. Several initial sampling points of each decision space are randomly arranged to break the correlation. Finally, one point is selected from several initial sampling points in each space to form a multidimensional decision variable parameter combination that meets the sample number, namely, a decision sample set.

9. The multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan according to claim 1, characterized in that: The aerodynamic performance of the counter-rotating axial flow fan corresponding to the decision sample set is solved by numerical simulation; The settings of the numerical simulation include: The calculation area of ​​the numerical simulation includes the collector, two rows of counter-rotating impellers and diffuser of the counter-rotating axial flow fan; The motion coordinate system and mixing plane model are used to describe the motion characteristics of the flow field inside the counter-rotating axial flow fan. Use the NS governing equations, combined with the Reynolds average method and the RNG k-ε turbulence model to simulate the behavior of the turbulent flow field; The boundary condition of the collector inlet section is mass flow inlet, and the boundary condition of the diffuser outlet section is pressure outlet.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the multi-objective and multi-component integrated optimization method for a counter-rotating axial flow fan as described in any one of claims 1 to 9 are implemented.

Images