CN117669139A - Optimization method of proton exchange membrane fuel cell flow channel based on improved genetic algorithm - Google Patents

Abstract

The invention discloses a proton exchange membrane fuel cell flow channel optimization method based on an improved genetic algorithm, which includes the steps of: establishing a three-dimensional model of a single straight flow channel PEMFC; aiming at the uneven population distribution and easy falling into localization in the optimization solution process of the genetic algorithm itself. For optimal solution problems, an improved genetic algorithm is proposed based on three improvement strategies. Then, combined with the improved genetic algorithm, the inlet and outlet cross-sectional dimensions of the PEMFC flow channel are simulated and optimized under the premise of comprehensively considering the loss power and output power, so as to obtain the optimal size configuration. The invention can not only determine the optimal size configuration of the flow channel according to the objective function, but also improve the optimization ability and convergence speed of the genetic algorithm to a certain extent, and help improve the design method of the proton exchange membrane fuel cell.

Description

Optimization method of proton exchange membrane fuel cell flow channel based on improved genetic algorithm

Technical field

The present invention relates to the field of fuel cells, and more specifically, to a proton exchange membrane fuel cell flow channel optimization method based on an improved genetic algorithm.

Background technique

In recent years, proton exchange membrane fuel cells (PEMFC), as an important component of fuel cells, have broad application prospects in the automotive field and portable electronic devices due to their high specific power, fast startup, and good environmental characteristics. The flow channel is one of the important structures that make up the proton exchange membrane fuel cell (PEMFC). It plays an important role in the transportation and distribution of reactants and is a key factor affecting battery performance. Therefore, changing the geometric configuration of the flow channel to obtain the optimal flow channel form is an important aspect to improve the performance of PEMFC.

Generally speaking, conventional flow channel structures mainly include parallel flow channels, serpentine flow channels, interdigitated flow channels and point flow channels. Among them, parallel flow channels have great advantages due to their convenient processing, small pressure loss, and uniform current density distribution. popular. Many researchers have studied the cross-sectional shape and size of parallel flow channels in order to obtain the best cross-sectional shape that is most beneficial to fuel cell operation. However, current research methods are limited to qualitative analysis of simple flow channels with different cross-sectional shapes, and cannot obtain an accurate quantitative relationship between performance output and cross-sectional geometric dimensions. Therefore, it is necessary to develop and design an accurate and efficient flow channel optimization method for proton exchange membrane fuel cells.

Chinese patent document CN112582635A provides a method for optimizing the flow channel cross section of a PEMFC bipolar plate and a three-dimensional proton exchange membrane fuel cell. The impact of different flow channel cross sections on the performance of PEMFC for the same flow channel structure was studied, and through optimization The PEMFC bipolar plate flow channel structure can effectively improve the uniformity of gas in the catalytic layer, thereby reducing cold spots, hot spots and flooding, ultimately achieving the purpose of reducing costs and extending life. However, this optimization method only stays on the qualitative study of "constant cross-section" flow channels, does not obtain the precise size configuration of the optimal cross-section, and does not take into account the influence of pressure losses caused by different cross-sections during optimization. At present, there is still a lack of an accurate, efficient and comprehensive method to optimize the flow channel of proton exchange membrane fuel cells.

Contents of the invention

The present invention provides a proton exchange membrane fuel cell flow channel optimization method based on an improved genetic algorithm. By establishing a three-dimensional model of a single DC channel PEMFC, combined with the improved genetic algorithm, the PEMFC flow channel is optimized under the premise of comprehensively considering the loss power and output power. The inlet and outlet cross-sectional dimensions are simulated and optimized to obtain the best size configuration to overcome at least one of the shortcomings of the existing technology.

In order to achieve the above technical objectives, the present invention provides a proton exchange membrane fuel cell flow channel optimization method based on an improved genetic algorithm, which includes the following steps:

The simulation software COMSOL Multiphysic is used to establish a single DC channel PEMFC three-dimensional model including cathode and anode flow channels, gas diffusion layer (GDL), catalytic layer (CL) and proton exchange membrane (PEM) and other computational domains. The size parameters can be set according to needs. The control equations of this model mainly include mass conservation equation, momentum conservation equation, energy conservation equation, component conservation equation, charge conservation equation and boundary equation;

Mesh the single DC channel PEMFC three-dimensional model. The number of grids is set between 70000-95000. When the battery operating voltage is a fixed value, within the range where the difference in current density being solved is less than 0.05%, select a grid with a smaller number of grids;

The size parameters of the cathode and anode inlet and outlet sections of the single DC channel PEMFC three-dimensional model are set as optimization variables, which are automatically generated in MATLAB according to certain rules by an improved genetic algorithm and transferred to COMSOL Multiphysics. Each set of variables corresponds to a A specific flow channel structure, and through simulation calculations, the required parameters after reaching convergence conditions are obtained;

Under the condition that the battery operating voltage is a certain fixed value, the current density and voltage drop calculated by the simulation are input into MATLAB to calculate the objective function f. If the objective function value does not reach the minimum value, then the individual gene will It will continue to change through inheritance, crossover, mutation, etc., and the corresponding geometric parameters of the flow channel will also change;

According to the above cycle process, under the condition that the battery operating voltage is a certain fixed value, the improved genetic algorithm continuously iteratively searches for optimization, and the model structure is continuously reconstructed and regenerated, thereby obtaining the optimal size configuration of the flow channel.

Further, it also includes steps: after establishing the model, compare the PEMFC battery polarization curves calculated by multiple sets of simulations with real experimental data, and verify the effectiveness of the model through the relative error between the two.

Furthermore, compared with the traditional genetic algorithm, the improved genetic algorithm makes the following three improvements:

Improvement 1. The improved genetic algorithm selects Circle chaos mapping as the method to generate the initial population. This method has strong randomness and will better travel the entire search range, making the initial population more evenly distributed, thereby effectively increasing the population. diversity. The specific mathematical expression is as follows:

Among them, mod is the remainder function.

Improvement 2. The improved genetic algorithm uses adaptive crossover probability P _c and adaptive mutation probability P _m during iterative calculation to improve the optimization performance of the algorithm. The specific calculation formula is as follows:

Among them, f _max is the maximum fitness in the group, f is the larger fitness of the two individuals that need to be crossed, f _avg is the average fitness of the group, and f′ is the fitness of the individual to be mutated.

Improvement 3. The improved genetic algorithm uses the Cauchy-Gaussian mutation strategy to ensure that the optimization result is the global optimal solution. The specific mathematical expression is as follows:

X _CG =X _best [1+λ ₁ cauchy(0, σ ² )+λ ₂ Gauss(0, σ ² )]

^Among _them ^{, ​} Random variables satisfying Gaussian distribution, λ ₁ and λ ₂ are dynamic parameters that are adaptively adjusted with the number of iterations.

The objective function f of the improved genetic algorithm can be set as needed. For different objective functions f, a set of corresponding optimal size configurations will be generated.

The advantages of the present invention compared with the prior art are:

1. In view of the problems that the genetic algorithm itself has in the optimization process, such as uneven population distribution and easy falling into local optimal solutions, the present invention proposes an improved genetic algorithm based on three improvement strategies, which improves the performance of the genetic algorithm to a certain extent. Optimization ability and convergence speed;

2. The present invention uses an improved genetic algorithm to simultaneously optimize the cross-sections of the cathode and anode inlet and outlet of the proton exchange membrane fuel cell flow channel, so that the optimization does not only stay at the level of "constant cross-section" optimization, but has a wider scope of application and more applications. widely;

3. The present invention uses an improved genetic algorithm to optimize the proton exchange membrane fuel cell flow channel. The objective function f can be set according to different optimization needs. In the end, an accurate size configuration that meets the conditions can be obtained, and the optimization process is more accurate, efficient and comprehensive. .

Description of drawings

Figure 1 is a flow chart of a proton exchange membrane fuel cell flow channel optimization method based on an improved genetic algorithm provided by an embodiment of the present invention;

Figure 2 is a schematic diagram of a three-dimensional model of a single DC channel PEMFC in an embodiment of the present invention;

In the figure: 1-cathode flow channel; 2-cathode gas diffusion layer; 3-cathode catalytic layer; 4-proton exchange membrane; 5-anode catalytic layer; 6-anode gas diffusion layer; 7-cathode flow channel.

Figure 3 is a grid independence verification diagram of the single DC channel PEMFC three-dimensional model in the embodiment of the present invention;

Figure 4 is a verification diagram of the accuracy of the single DC channel PEMFC three-dimensional model in the embodiment of the present invention;

Figure 5 is a cross-sectional view of a single DC channel PEMFC flow channel optimization in an embodiment of the present invention;

Figure 6 is a comparison chart of polarization curves before and after optimization of the single DC channel PEMFC flow channel in the embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiment is only a specific embodiment of the present invention, not all embodiments. . The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application or uses. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention. The relative arrangement of components and steps, expressions, and numerical values set forth in these examples do not limit the scope of the invention unless otherwise specifically stated. At the same time, it should be understood that, for convenience of description, the dimensions of various parts shown in the drawings are not drawn according to actual proportional relationships. Techniques, methods and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, such techniques, methods and devices should be considered part of the authorized specification. In all examples shown and discussed herein, any specific values are to be construed as illustrative only and not as limiting. Accordingly, other examples of the exemplary embodiments may have different values.

Examples of the present invention will be described in detail below.

Please refer to Figure 1. The present invention provides a proton exchange membrane fuel cell flow channel optimization method based on an improved genetic algorithm, which includes the following steps:

Step 1: Establish a three-dimensional model of a single DC channel PEMFC in the simulation software COMSOL Multiphysic.

In some embodiments of the present invention, a three-dimensional model of a single DC channel PEMFC is first established. This model can theoretically simulate almost all transmission conditions that can occur inside the battery. The schematic diagram of the model structure is shown in Figure 2, in which: 1 -Cathode flow channel, 2-cathode gas diffusion layer, 3-cathode catalytic layer, 4-proton exchange membrane, 5-anode catalytic layer, 6-anode gas diffusion layer, 7-anode flow channel. The calculation area includes the cathode flow channel, anode flow channel, gas diffusion layer (GDL), catalytic layer (CL) and proton exchange membrane (PEM). The gas diffusion layer, catalytic layer and proton exchange membrane are collectively called the membrane electrode assembly. (MEA). The size parameters of each calculation area can be set as needed. In this embodiment, the geometric sizes corresponding to each calculation area are as shown in Table 1.

Table 1 Dimensions of each calculation area of the PEMFC three-dimensional model

parameter name numerical value unit Runner length 20 mm runner height 1 mm Channel width 1 mm Floor width 1 mm GDL thickness 0.3 mm CL thickness 0.0129 mm PEM thickness 0.108 mm

The control equations of the single DC channel PEMFC three-dimensional model mainly include mass conservation equations, momentum conservation equations, energy conservation equations, component conservation equations, charge conservation equations and boundary equations.

In some embodiments of the present invention, in order to accurately solve the control equations, some boundary conditions need to be set. Among them, the concentration and humidity of oxygen and hydrogen introduced at the entrances of the cathode and anode can be set according to different needs. The temperature of the flow channel inlet and the surrounding wall temperature are equal to the operating temperature. According to the ideal gas law, the mass fraction of each component can be calculated from the inlet temperature, inlet pressure and relative humidity.

It is also necessary to preset initial conditions, which include working pressure, temperature, cathode and anode stoichiometric ratio, GDL porosity, GDL conductivity, and GDL permeability. Each parameter can be set according to different needs.

In some embodiments of the present invention, some of the electrochemical parameters involved in the single DC channel PEMFC three-dimensional model used are shown in Table 2.

Table 2 PEMFC three-dimensional model electrochemical parameter table

Step 2: Mesh the single DC channel PEMFC three-dimensional model.

The single DC channel PEMFC three-dimensional model was meshed and the mesh independence was verified. In some embodiments of the present invention, four grid models with different net numbers, 33124, 54912, 72192, and 92976, are used to verify the grid independence of the established model. As shown in Figure 3, when the battery operating voltage is 0.5V, the difference in current density solved by the two grid models with grid numbers of 72192 and 92976 is less than 0.05%, which is within the completely acceptable error range, so To shorten the solution time, the number of grids selected is 72192.

Step 3: Compare the PEMFC battery polarization curve calculated by simulation with experimental data to verify the effectiveness of the model.

In some embodiments of the present invention, in order to verify the accuracy of the PEMFC single-channel three-model, the PEMFC battery polarization curve calculated through simulation is compared with the experimental data. The comparison results are shown in Figure 4. The simulation results obtained through simulation calculation fit well with the experimental data. When the current density is low, the difference between the two is very small, and when the current density is high, the relative error between the two is within 5%. Therefore, this model can be applied as a primitive flow channel model.

Step 4: Set the size parameters of the cathode inlet and outlet cross-sections and the anode inlet and outlet cross-sections of the single DC channel PEMFC three-dimensional model as optimization variables, which are automatically generated in MATLAB according to certain rules by the improved genetic algorithm and transferred to COMSOL Multiphysics In , each set of variables corresponds to a specific flow channel structure, and the current density and voltage drop after reaching the convergence condition are obtained through simulation calculations.

In this embodiment, the genetic algorithm is selected as the optimization algorithm for the shape and size of the flow channel, and is combined with simulation to obtain the optimal size configuration of the flow channel. The specific implementation process is realized through the commercial mathematics software MATLAB2021 and the simulation software COMSOL Multiphysics6.0. The two can be connected through interfaces to achieve automatic calculation, which can greatly improve the optimization efficiency.

Select the size of the inlet and outlet sections of the cathode and anode flow channels as the optimization target, as shown in Figure 5, the width and height of the cathode flow channel inlet, the width and height of the cathode flow channel outlet, the width and height of the anode flow channel inlet, the anode flow channel The outlet width and height are represented by a, b, c, d, e, f, g, h respectively. After the inlet and outlet cross-section dimensions of the cathode and anode flow channels are determined, the shape of the flow channel can be determined.

In order to ensure the consistency of the boundary conditions at the inlet and outlet of the flow channel during the optimization process, the restriction condition of the improved genetic algorithm is that the areas of the inlet and outlet cross-sections of the flow channel are equal. In some embodiments of the present invention, the inlet and outlet channels are The cross-sectional area is fixed at 1mm ² , and the cross-sectional area can be automatically adjusted according to the model size, and the range of each size parameter is specified as follows based on experimental and simulation experience:

0.3mm≤a,b,c,d,e,f,g,h≤1.7mm

Step 5: Under the condition that the battery operating voltage is a certain fixed value, input the current density and voltage drop calculated by the simulation into MATLAB to calculate the objective function f. If the objective function value does not reach the minimum value, then the individual The genes will continue to change through inheritance, crossover, mutation, etc., and the corresponding geometric parameters of the flow channel will also change.

During the operation of PEMFC, an air compressor is usually required to provide the required oxygen. However, different flow channel structures will produce different pressure drops, so that the energy consumed by the air compressor is also different. This part of energy is used in the optimization process. cannot be ignored. Therefore, under the premise of comprehensively considering the output power and loss power, the objective function f is defined as follows:

E _cell = V _cell i _ave A _m

Among them, k is a constant, set to 10 ⁵ , E _cell is the battery output power, E _con is the battery loss power, V _cell is the battery operating voltage, taken as 0.4V, i _ave is the battery average current density, and A _m is the battery activation area. , Δp _c is the battery cathode voltage drop, is the battery inlet flow rate,/> is the cross-sectional area of the battery inlet,. When the battery output power E _cell is larger and the power loss E _con is smaller, the fitness value (that is, the objective function value f) is smaller, indicating that the fitness of the individual is higher, the net power of the PEMFC is higher, and the performance of the battery is better. .

When the optimization variables, constraints, and objective functions of the genetic algorithm are determined one after another, the optimization work can begin.

The optimization variables described are automatically generated by genetic algorithms in MATLAB according to certain rules, and each set of variables is a gene-carrying individual in the population.

The above-mentioned optimization variables are input into COMSOL Multiphysics. Each set of optimization variables corresponds to a specific flow channel structure, and the current density and voltage drop after reaching the convergence condition are obtained through simulation calculation, and the objective function value is calculated based on this. If the objective function value does not reach the minimum value, then the individual genes will continue to change through inheritance, crossover, mutation, etc., and the corresponding geometric parameters of the flow channel will also change.

According to the above cycle process, under the condition that the battery operating voltage is 0.4V, the genetic algorithm continues to iterate and optimize, and the model structure is continuously reconstructed and regenerated, thereby obtaining the optimal size configuration of the flow channel.

In view of the problems that the genetic algorithm itself has in the optimization and solution process, such as uneven population distribution and easy falling into local optimal solutions, an improved genetic algorithm is proposed based on three improvement strategies.

Improvement 1. The genetic algorithm selects Circle chaos mapping as the method to generate the initial population. This method has strong randomness and will better travel the entire search range, making the initial population more evenly distributed, thereby effectively increasing the diversity of the population. sex. The specific mathematical expression is as follows:

Among them, mod is the remainder function, and a is the generated a-th individual.

Improvement 2. The improved genetic algorithm uses adaptive crossover probability P _c and adaptive mutation probability P _m during iterative calculation to improve the optimization performance of the algorithm. The specific calculation formula is as follows:

Among them, f _max is the maximum fitness in the group, f is the larger adaptation of the two individuals that need to be crossed, f _avg is the average fitness of the group, and f′ is the fitness of the individual to be mutated.

Improvement 3. The improved genetic algorithm uses the Cauchy-Gaussian mutation strategy to ensure that the optimization result is the global optimal solution. The specific mathematical expression is as follows:

X _CG =X _best [1+λ ₁ cauchy(0, σ ² )+λ ₂ Gauss(0, σ ² )]

Among them, X _CG represents the position of the optimal individual after ^mutation , _{X best} is the position of the optimal individual before mutation _, is the maximum number of iterations, t is the current number of iterations, cauchy(0, σ ² ) is a random variable that satisfies the Cauchy distribution, Gauss(0, σ ² ) is a random variable that satisfies the Gaussian distribution, λ ₁ , λ ₂ are random variables with iteration Dynamic parameters for adaptive adjustment of times.

In some embodiments of the present invention, the original flow channel model is optimized through a genetic algorithm and an improved genetic algorithm respectively. After 30 iterative calculations, when the average fitness is the same as the best fitness, the iterative optimization work is completed. , the optimal solution can be obtained.

After the optimization iteration of the genetic algorithm to 18 generations, the calculation results began to converge, and the best fitness value after convergence was 3.1169; after the improved genetic algorithm iteration to 18 generations, the calculation results began to converge, and the best fitness value after convergence The value is 2.9673. Compared with the genetic algorithm, the improved genetic algorithm converges faster and saves about 33.3% of the convergence time.

The optimization variables corresponding to the best fitness obtained after the convergence of the genetic algorithm and the improved genetic algorithm are the optimal parameters for the shape and size of the flow channel, as shown in Table 3.

Table 3 Optimum size of flow channel after optimization by genetic algorithm and improved genetic algorithm

The shapes of the cathode and anode flow channels optimized by the genetic algorithm and the improved genetic algorithm are both tapered flow channels with variable cross-sections, and show a trend of decreasing depth and increasing width along the reaction gas transport direction.

The flow channel polarization curve optimized by genetic algorithm and improved genetic algorithm was studied and compared with the original model.

As shown in Figure 6, comparing the polarization curves before and after optimization, it can be seen that the performance of the PEMFC optimized by the method of the present invention has been significantly improved. When the current density is low, the difference in the polarization curves is very small. However, as the current density is lower, the difference in the polarization curves is very small. As the current density continues to increase, the performance advantages of optimized flow channels become more and more obvious.

In some embodiments of the present invention, at an operating voltage of 0.4V, compared with the original flow channel model, the flow channel current density optimized by the genetic algorithm and the improved genetic algorithm increased by 14.20% and 19.01% respectively, and the peak power The density is increased by 12.58% and 15.77%, and the optimization effect of the improved genetic algorithm is more significant.

The method provided by the foregoing embodiments of the present invention is jointly implemented by the simulation software COMSOL Multiphysic and the commercial mathematics software MATLAB. Combined with the improved genetic algorithm, the inlet and outlet cross-sectional dimensions of the PEMFC flow channel are simulated and optimized under the premise of comprehensively considering the loss power and output power, so as to obtain the optimal size configuration. The present invention can not only determine the optimal size configuration of the flow channel according to a specific objective function, but also improves the optimization ability and convergence speed of the genetic algorithm to a certain extent, helping to improve the design of proton exchange membrane fuel cells.

The scope of protection of the present invention is not limited to the above-mentioned embodiments. Those skilled in the art and scientific researchers can make various modifications to the present invention without departing from the scope of the present invention. If these modifications fall within the scope of the claims of the present invention and equivalent technologies, the present invention is intended to include these modifications.

Claims (10)

1. The proton exchange membrane fuel cell runner optimization method based on the improved genetic algorithm is characterized by comprising the following steps:

establishing a single direct current channel PEMFC three-dimensional model;

grid division is carried out on the single direct current channel PEMFC three-dimensional model;

setting size parameters of an inlet section and an outlet section of a cathode section and an anode section of a single direct current channel PEMFC three-dimensional model as optimized variables, automatically generating in mathematical software by an improved genetic algorithm, transmitting the optimized variables into simulation software, wherein each group of optimized variables corresponds to a channel structure, obtaining current density and voltage drop after reaching convergence conditions by simulation and analog calculation based on the improved genetic algorithm, wherein the improvement of the improved genetic algorithm comprises the steps of selecting Circle chaotic map as a method for generating an initial population, and adopting self-adaptive crossover probability P in iterative calculation _c And adaptive mutation probability P _m And using a cauchy-gaussian variation strategy;

under the condition that the working voltage of the battery is a certain fixed value, the current density and the voltage drop obtained by simulation calculation are input into the mathematical software, so that an objective function f is obtained by calculation, if the objective function value does not reach the minimum value, the genes of an individual are continuously changed in a genetic, cross, mutation and other modes, and the corresponding geometric parameters of a flow channel are also changed;

according to the above-mentioned circulation process, under the condition that the working voltage of the battery is a certain preset fixed value, the improved genetic algorithm is continuously iterated and optimized, the model structure is continuously rebuilt and regenerated, and the optimum size configuration of the flow channel is obtained.

2. The improved genetic algorithm-based proton exchange membrane fuel cell flow channel optimization method as claimed in claim 1, wherein the calculation region of the single direct flow channel PEMFC three-dimensional model comprises a cathode flow channel, an anode flow channel, a gas diffusion layer, a catalytic layer and a proton exchange membrane.

3. The improved genetic algorithm-based proton exchange membrane fuel cell runner optimization method as claimed in claim 1, wherein the single direct current runner PEMFC three-dimensional model performs simulation calculation based on preset initial conditions and boundary conditions, wherein the initial conditions comprise working pressure, temperature, positive-negative polarization stoichiometric ratio, GDL porosity, GDL conductivity and GDL permeability, and the boundary conditions comprise oxygen concentration and humidity at an inlet of a cathode runner and hydrogen concentration and humidity at an inlet of an anode runner.

4. The improved genetic algorithm-based proton exchange membrane fuel cell runner optimization method as claimed in claim 1, wherein the dimensional parameters of the inlet and outlet sections of the single-direct-runner PEMFC three-dimensional model are set as optimization variables.

5. The method for optimizing a proton exchange membrane fuel cell flow channel based on an improved genetic algorithm as claimed in claim 4, wherein the limitation of the improved genetic algorithm is that the areas of the inlet and outlet sections of the flow channel are equal.

6. The improved genetic algorithm-based proton exchange membrane fuel cell runner optimization method as claimed in claim 1, wherein the initial population is generated by adopting Circle chaotic mapping, and the mathematical expression is X _a+1 The following are provided:

where mod is the remainder function and a is the generated a-th individual.

7. The method for optimizing a proton exchange membrane fuel cell flow channel based on an improved genetic algorithm according to claim 1, wherein the improved genetic algorithm adopts an adaptive crossover probability P in iterative calculation _c The calculation formula of (2) is as follows:

wherein f _max For maximum fitness in the population, f is the required crossoverIs large in adaptability of two bodies, f _avg Is population average fitness.

8. The improved genetic algorithm-based proton exchange membrane fuel cell flow channel optimization method as claimed in claim 1, wherein the improved genetic algorithm adopts an adaptive variation probability P in iterative calculation _m The calculation formula of (2) is as follows:

wherein f _max For maximum fitness in the population, f _avg For population average fitness, f' is the fitness of the individual to be mutated.

9. The improved genetic algorithm-based proton exchange membrane fuel cell flow channel optimization method as claimed in claim 1, wherein the improved genetic algorithm adopts the following mathematical expression of cauchy-gaussian variation strategy:

X _CG ＝X _best [1+λ ₁ cauchy(0，σ ² )+λ ₂ Gauss(0，σ ² )]

wherein X is _CG Represents the position after the mutation of the optimal individual, X _best For optimal position before individual variation, X _i For the location where the i-th individual is located,σ ² standard deviation, T, representing the cauchy-gaussian variation strategy _max For the maximum number of iterations, t is the current number of iterations, cauchy (0, σ) ² ) Is a random variable satisfying the Coxis distribution, gauss (0, σ ² ) Is a random variable satisfying a Gaussian distribution, lambda ₁ 、λ ₂ Is a dynamic parameter adaptively adjusted with the number of iterations.

10. A proton exchange membrane fuel cell flow channel optimization method based on an improved genetic algorithm as claimed in any one of claims 1-9, wherein said objective function f is

E _cell ＝V _cell i _ave A _m

Where k is a constant, E _cell For battery output power E _con For battery power loss, V _cell For battery operating voltage, i _ave For average current density of battery, A _m Δp for battery activation area _c For the cathode fall of a battery,for the battery inlet flow rate, ">Is the cross-sectional area of the inlet of the battery.