Fuel cell flow channel structure, optimization methods and fuel cells
By setting trapezoidal baffles in the fuel cell flow channel and combining them with a multi-objective optimization method, the problems of uneven gas diffusion and insufficient water management in the flow channel structure were solved, thereby improving the output performance and lifespan of the fuel cell and achieving comprehensive performance optimization of the flow channel structure.
Patent Information
- Application Number
- CN202611628249.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-21
- Publication Date
- 2026-11-17
AI Technical Summary
Existing proton exchange membrane fuel cell flow channel structures suffer from problems such as insufficient gas diffusion, uneven hydrogen and oxygen transport, and failure to remove water in a timely manner. These issues make it difficult to meet the gas transfer requirements under high-speed convection conditions, and there are contradictions among the various performance indicators, making it difficult to optimize them to achieve the best overall performance.
A meandering flow channel structure with trapezoidal baffles is adopted. Multi-objective optimization is performed by combining the radial basis function surrogate model and the NSGA-II algorithm. The NSGA-II algorithm is used for multi-objective optimization search, and the TOPSIS multi-attribute decision method is used for comprehensive evaluation. The flow channel structure is optimized to enhance gas diffusion and water management capabilities.
It significantly improves the gas distribution and water management capabilities of the flow channel, alleviates local hypoxia and flooding problems, enhances the output performance and lifespan of the fuel cell, optimizes speed and iteration efficiency, and ensures the accuracy of results.
Smart Images

Figure CN123120771A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the technical field of battery flow channel structure design optimization, specifically relating to a fuel cell flow channel structure, optimization method, and fuel cell. Background Technology
[0002] Proton exchange membrane fuel cells are energy conversion devices that directly convert the chemical energy in fuel into electrical energy through electrochemical reactions without combustion. They have outstanding advantages such as high energy conversion efficiency, zero pollution emissions, convenient operation, and mild operating conditions. They have shown broad application prospects in fields such as vehicle power, portable power sources, and distributed power generation, and have become one of the core research directions in the current clean energy field.
[0003] Bipolar plates are one of the core components of proton exchange membrane fuel cells (PEMFCs). The flow channels arranged on their surface play a crucial role in transporting reactant gases (hydrogen and oxygen), discharging electrochemical reaction products (water), conducting current, and dissipating reaction heat. The rationality of the bipolar plate flow channel design directly determines the effectiveness of hydrothermal management and the uniformity of reactant gas distribution within the PEMFC, thus significantly affecting the overall output power, operational stability, and lifespan of the cell. Practice has shown that optimizing the flow channel structure parameters can effectively improve the uniformity of reactant gas distribution within the cell, improve the internal water balance, reduce concentration polarization and ohmic losses, and prevent flooding / membrane dryness failures, thereby further enhancing the durability and reliability of PEMFCs.
[0004] The overall performance of proton exchange membrane fuel cells is closely related to the flow channel configuration parameters. Different flow channel configuration parameters (including block structure parameters, flow channel geometric parameters, etc.) correspond to significant differences in the overall performance of the cells. At the same time, there are contradictory relationships between various key performance indicators of the cells. For example, the increase in cell output power is often accompanied by an increase in the pressure drop at the inlet and outlet of the flow channel, which in turn increases parasitic power consumption.
[0005] In existing technologies, traditional flow channel structures suffer from problems such as weak convection intensity, insufficient gas diffusion, and uneven hydrogen and oxygen transport, making it difficult to meet the gas transfer requirements under high-speed convection conditions in fuel cells. At the same time, if the water generated by the reaction on the cathode side is not removed in time, it will cause blockage of the gas diffusion layer, preventing oxygen from reaching the catalyst layer, and the concentration overpotential will increase rapidly, reducing power generation efficiency. The question is how to balance the contradictions between various performance indicators and obtain the flow channel structure with optimal overall performance by optimizing the flow channel configuration parameters. Summary of the Invention
[0006] To address the aforementioned technical problems, this invention provides a fuel cell flow channel structure, optimization method, and fuel cell, thereby overcoming the shortcomings of the prior art.
[0007] In a first aspect, the present invention provides a method for optimizing the flow channel structure of a fuel cell, the method comprising: A parametric geometric model of the meandering flow channel is established, and several design variables of the meandering flow channel are adjusted to generate a three-dimensional model of the meandering flow channel with the corresponding structure. A three-dimensional numerical simulation model of a proton exchange membrane fuel cell with a meandering flow channel structure featuring trapezoidal baffles was established, and the three-dimensional numerical simulation model was verified. Determine the design variable space, and determine the upper and lower bound vectors based on the design variable space; A multi-objective optimization method is determined based on a preset objective, and the multi-objective optimization method is constrained by preset constraints. The Latin hypercube sampling method is used to generate initial sample points to be simulated in the design variable space, and the corresponding target matrix and pressure drop vector are obtained through numerical simulation to construct the initial sample dataset. Using the geometric parameters of the sample as input and the preset target performance index as output, a radial basis function surrogate model is constructed, and a mapping relationship between the design parameters and the radial basis function surrogate model is established. Based on the radial basis function surrogate model, fitness calculation is performed using the NSGA-II algorithm for multi-objective optimization search, in order to find the best solution in the design variable space, and guided by the preset objective of co-optimization, to output the predicted candidate Pareto solution set. The candidate Pareto solution set is sorted according to the predicted pressure drop and the target value, so as to select the optimal solution set sample for simulation calculation, and to supplement and update the sample database and iteratively update the radial basis function surrogate model and the multi-objective optimization method until the preset convergence termination condition is met. The TOPSIS multi-attribute decision method is used to comprehensively evaluate the candidate Pareto solution set. The weight coefficient of the preset target is set, and the relative closeness between each candidate Pareto solution set and the ideal optimal solution and the negative ideal solution is calculated. The comprehensive optimal flow channel structure is obtained based on the solution with the largest relative closeness.
[0008] Furthermore, the relationship between the upper bound vector and the lower bound vector is expressed as follows: ; In the formula, Denotes the lower bound vector. Indicates design variables, This represents the upper bound vector.
[0009] Furthermore, the preset targets include net power density, uniform oxygen mass fraction, and average water saturation at the membrane-catalyst layer interface, defined by the following expressions: ; ; ; In the formula, , , These represent net power density, uniformity of oxygen mass fraction distribution at the interface between the cathode gas diffusion layer and the catalyst layer, and average water saturation at the interface between the membrane and the catalyst layer, respectively. Indicates current density, Indicates battery voltage. This indicates the gas pressure loss at the cathode inlet and outlet. This indicates the volumetric flow rate of the cathode air. This represents the standard deviation of the oxygen mass fraction at the interface. This represents the average mass fraction of oxygen at the interface. Indicates the effective area of the battery. This indicates the water saturation at the interface between the membrane and the catalyst layer.
[0010] Furthermore, the step of determining a multi-objective optimization method based on a preset objective and constraining the multi-objective optimization method through preset constraints includes: A multi-objective optimization method was determined with net power density, oxygen mass fraction uniformity, and water saturation as optimization objectives. The multi-objective optimization method is constrained by the flow channel pressure drop.
[0011] Furthermore, the expression for the radial basis function surrogate model is: ; In the formula, This indicates the performance metrics of the output. Indicates design variables, Indicates the first The design variable vector corresponding to each input sample Indicates the weighting coefficient. This represents the Euclidean distance between the sample point and the point to be predicted. For radial basis kernel functions, Indicates the sample number. This represents the total number of samples in the radial basis function surrogate model.
[0012] Furthermore, the steps for performing multi-objective optimization search using the NSGA-II algorithm include: Based on the Pareto optimality theory, the NSGA-II algorithm achieves collaborative optimization of multiple conflicting objectives through non-dominated sorting and crowding distance.
[0013] Furthermore, the step of using the TOPSIS multi-attribute decision method to comprehensively evaluate the candidate Pareto solution set, setting the weight coefficient of the preset objective, and calculating the relative closeness of each candidate Pareto solution set to the ideal optimal solution and the negative ideal solution includes: Construct an initial decision matrix, and normalize the initial decision matrix and introduce weight coefficients to construct a weighted normalized matrix; Calculate the Euclidean distance from each of the candidate Pareto solution sets to the ideal optimal solution and the negative ideal solution, and calculate the relative proximity.
[0014] Furthermore, the expressions for calculating the Euclidean distance from each of the candidate Pareto solution sets to the ideal optimal solution and the negative ideal solution are as follows: ; ; In the formula, Indicates the first The Euclidean distance from each solution to the ideal optimal solution. Indicates the first Euclidean distance from the ideal solution to the negative ideal solution. Indicates the target dimension number. Indicates the number of optimization targets. Indicates the first The candidate solution corresponds to the th... The weighted normalized index value of the evaluation objective. Indicates the first The components of the ideal optimal solution corresponding to the project target. Indicates the first The negative ideal solution component corresponding to the project target; The expression for calculating the relative closeness is: ; In the formula, Indicates the first The comprehensive relative fit of each Pareto candidate solution.
[0015] Secondly, the present invention also provides a fuel cell flow channel structure, which is optimized by the above-mentioned fuel cell flow channel structure optimization method, including a plurality of cathode flow channels, wherein the cathode flow channels are meandering flow channels, the plurality of meandering flow channels are arranged in parallel, the plurality of meandering flow channels are a continuous meandering extension structure to form a plurality of periodically alternating waveform segments, and a trapezoidal baffle is provided at the center of each of the plurality of waveform segments.
[0016] Thirdly, the present invention also provides a fuel cell, comprising a cathode flow channel composed of the above-mentioned fuel cell flow channel structure, and further comprising an anode flow channel, a membrane electrode assembly, and a cathode plate. The anode flow channel is a conventional meandering flow channel without trapezoidal baffles. The membrane electrode assembly comprises an anode diffusion layer, an anode catalyst layer, a proton exchange membrane, a cathode catalyst layer, and a cathode diffusion layer stacked sequentially. An anode plate is provided on the lower surface of the anode diffusion layer. The cathode flow channel is disposed on the cathode plate, and the anode flow channel is disposed on the anode plate.
[0017] Compared with existing technologies, the beneficial effects of this invention are as follows: By setting trapezoidal baffles within the meandering flow channel, forced convection and turbulent disturbance within the channel are enhanced, strengthening the diffusion and transport capacity of multi-component gases such as hydrogen and oxygen, improving the gas distribution and water management capabilities of the flow channel, and alleviating flooding and localized hypoxia problems. Furthermore, by employing a radial basis function surrogate model combined with the NSGA-II multi-objective optimization method, the speed and iteration efficiency of structural optimization are improved while meeting multiple index requirements, further enhancing the overall performance of the flow channel. In addition, the final population and Pareto front are all derived from real simulation data; the surrogate model is only used for candidate selection and does not participate in the final result output, avoiding the problem of scheme failure caused by prediction errors and ensuring the accuracy of the results. Attached Figure Description
[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0019] Figure 1 This is a schematic diagram of the overall structure of the meandering flow channel structure of the hydrogen fuel cell with trapezoidal baffles in the first embodiment of the present invention. Figure 2 This is a partial structural schematic diagram of the meandering flow channel structure of a hydrogen fuel cell with trapezoidal baffles in the first embodiment of the present invention. Figure 3 This is a flowchart of the fuel cell flow channel structure optimization method in the second embodiment of the present invention; Figure 4 This is a comparison diagram of the polarization curve and power curve of the hydrogen fuel cell involved in the second embodiment of the present invention. Figure 5 This is a schematic diagram of a fuel cell with a flow channel in the third embodiment of the present invention; Figure 6 for Figure 5 Enlarged view of a portion of point A in the middle; Figure 7This is a schematic diagram of the multi-objective optimized flow channel structure in the third embodiment of the present invention.
[0020] Explanation of key component symbols: 1. Cathode flow channel; 2. Trapezoidal baffle; 30. Membrane electrode assembly; 3. Anode plate; 4. Anode diffusion layer; 5. Anode catalyst layer; 6. Proton exchange membrane; 7. Cathode catalyst layer; 8. Cathode diffusion layer; 9. Cathode plate.
[0021] The embodiments of the present invention will be further described below with reference to the accompanying drawings. Detailed Implementation
[0022] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
[0023] It should be noted that when a component is said to be "fixed to" another component, it can be directly on the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.
[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.
[0025] Example 1 Please see Figures 1 to 2 The diagram shows the fuel cell flow channel structure in an embodiment of the present invention, including several cathode flow channels 1, which are meandering flow channels. The meandering flow channels are arranged in parallel with each other and are continuous meandering extension structures to form several periodically alternating waveform segments. Each waveform segment has a trapezoidal baffle 2 at its center.
[0026] In this embodiment, the geometric parameters of the meandering channel include channel width W1, channel depth H, rib width W2, transverse dimension X of the waveform section, longitudinal dimension Y of the waveform section, transition fillet radius R, number of meandering unit groups N, trapezoidal block height h, short side length L1 of the trapezoidal block, and long side length L2 of the trapezoidal block, totaling 10 design parameters, denoted as the design variable vector. The relational expression is: ; It should be noted that each meandering channel extends continuously, forming multiple alternating rising and falling sections. The trapezoidal baffle is a trapezoidal body with the upper side shorter than the lower side and the height lower than the channel depth. The trapezoidal baffle is fixedly set at the center of each section of the meandering channel.
[0027] Example 2 Please see Figure 3 The image shows a fuel cell flow channel structure optimization method provided in the second embodiment of the present invention, used to optimize the fuel cell flow channel structure in Embodiment 1. The method includes steps S1 to S9: S1. Establish a parametric geometric model of the meandering flow channel and adjust several design variables of the meandering flow channel to generate a three-dimensional model of the meandering flow channel with the corresponding structure. It should be noted that by adjusting ten design variables, including the channel width W1, channel depth H, rib width W2, transverse dimension X of the waveform section, longitudinal dimension Y of the waveform section, transition fillet radius R, number of meandering unit groups N, height h of the trapezoidal stop, length L1 of the short side of the trapezoidal stop, and length L2 of the long side of the trapezoidal stop, a three-dimensional model of the meandering channel with the corresponding structure can be automatically generated. In this embodiment, the original values of each design variable before optimization are shown in Table 1. Table 1
[0028] S2. Establish a three-dimensional numerical simulation model of a proton exchange membrane fuel cell with a meandering flow channel structure featuring trapezoidal baffles, and verify the three-dimensional numerical simulation model. It should be noted that a three-dimensional numerical simulation model of a proton exchange membrane fuel cell with a meandering flow channel structure featuring trapezoidal baffles was established in Fluent, and the model was validated. The main simulation parameters of the model are shown in Table 2. Table 2
[0029] S3, determine the design variable space, and determine the upper bound vector and lower bound vector based on the design variable space; In this embodiment, the relationship between the upper bound vector and the lower bound vector is expressed as follows: ; In the formula, Denotes the lower bound vector. Indicates design variables, This represents the upper bound vector.
[0030] It should be noted that the design variables are determined and the value range of each variable is set. At the same time, the geometric constraint relationship between the variables is clarified to ensure that the flow channel structure can be generated reasonably. The design variables are the above 10 parameterized geometric parameters. In this embodiment, the constraint ranges for each design variable are: 0.5mm≤W1≤1.5mm; 0.5mm≤W2≤1.5mm; 2mm≤X≤10mm; 1mm≤Y≤10mm; 4≤N≤20mm; 0.4mm≤R≤0.6mm; 0.2mm≤H≤0.5mm; 0.1mm≤h≤0.5mm; 0.2mm≤L1≤1.2mm; 0.2mm≤L2≤1.2mm; To ensure that the flow channels corresponding to the generated parameter sets can be generated normally and appropriately, certain constraints need to be met between the design variables. The constraint relationships between the design variables are: 18≤10W1+9W2+Y≤20; 88≤2NX≤92; L2-L1≥0.2; h≤H; XY≥1. The constraints are to ensure that the flow channels are generated normally in the electrode plate in the horizontal and vertical directions, to ensure that the generated resistance blocks are trapezoidal, to prevent the resistance blocks from completely blocking the flow channels, and to prevent interference between multiple flow channels.
[0031] S4, determine a multi-objective optimization method based on a preset objective, and constrain the multi-objective optimization method by preset constraints; Specifically, step S4 includes steps S41 to S42: S41, a multi-objective optimization method is determined with net power density, oxygen mass fraction uniformity and water saturation as optimization objectives; S42, the multi-objective optimization method is constrained by the flow channel pressure drop.
[0032] In this embodiment, the preset targets include net power density, uniform oxygen mass fraction, and average water saturation at the membrane-catalyst interface, defined by the following expressions: ; ; ; In the formula, , , These represent net power density, uniformity of oxygen mass fraction distribution at the interface between the cathode gas diffusion layer and the catalyst layer, and average water saturation at the interface between the membrane and the catalyst layer, respectively. Indicates current density, Indicates battery voltage. This indicates the gas pressure loss at the cathode inlet and outlet. This indicates the volumetric flow rate of the cathode air. This represents the standard deviation of the oxygen mass fraction at the interface. This represents the average mass fraction of oxygen at the interface. Indicates the effective area of the battery. This indicates the water saturation at the interface between the membrane and the catalyst layer.
[0033] It should be noted that the objective function is determined, which includes the net power density of the fuel cell, the uniformity of oxygen mass fraction at the interface between the cathode catalyst layer and the gas diffusion layer, and the average water saturation at the interface between the membrane and the catalyst layer. Furthermore, the flow channel pressure drop is used as a metric. As an optimization constraint, the flow channel pressure drop must be lower than the set pressure drop threshold.
[0034] Understandably, the three objectives represent battery output performance, gas mass transfer uniformity, and water management capability, respectively. This forms the basis for a multi-objective optimization system to obtain the flow channel structure with optimal overall performance. The optimization objective function is uniformly expressed as a minimum optimization problem: ; in: ; ; ; In the formula, The objective function representing the final multi-objective goal. , , These represent the objective functions of three different sub-objectives. In this embodiment, the pressure drop threshold is set to 8000Pa to reduce pressure drop loss and prevent the generated flow channel interface cross-section from being too small and difficult to process.
[0035] S5. The Latin hypercube sampling method is used to generate initial sample points to be simulated in the design variable space, and the corresponding target matrix and pressure drop vector are obtained through numerical simulation to construct the initial sample dataset. It should be noted that the Latin hypercube sampling method is used to generate the initial sample points to be simulated in the design variable space, and the corresponding target matrix is obtained through numerical simulation. With pressure drop vector Build the initial sample dataset .
[0036] S6. Using the geometric parameters of the sample as input and the preset target performance index as output, construct a radial basis function surrogate model and establish a mapping relationship between the design parameters and the radial basis function surrogate model. It should be noted that, using the geometric parameters of the sample as input and the three target performance indicators as output, a radial basis function surrogate model is constructed to establish the mapping relationship between the design parameters and the objective function. Radial basis functions are a class of interpolation functions that use spatial Euclidean distance as the independent variable. The expression for the radial basis function surrogate model is: ; In the formula, This indicates the performance metrics of the output. Indicates design variables, Indicates the first The design variable vector corresponding to each input sample Indicates the weighting coefficient. This represents the Euclidean distance between the sample point and the point to be predicted. For radial basis kernel functions, Indicates the sample number. This represents the total number of samples in the radial basis function surrogate model. In this embodiment, a Gaussian kernel function is selected.
[0037] S7. Using the radial basis function surrogate model as the basis for fitness calculation, the NSGA-II algorithm is used to perform multi-objective optimization search to find the best solution in the design variable space, and guided by the preset objective of co-optimization, to output the predicted candidate Pareto solution set. Specifically, step S7 includes step S71: S71, based on the NSGA-II algorithm with Pareto optimality as the theoretical basis, achieves collaborative optimization of multiple conflicting objectives through non-dominated sorting and crowding distance.
[0038] It should be noted that the Radial Basis Function (RBF) surrogate model is used as the basis for fitness calculation, and the NSGA-II algorithm is employed for multi-objective optimization search. Global optimization is performed within the design space, guided by the goal of achieving co-optimization of the three objectives, and outputting a candidate Pareto solution set predicted by the surrogate model. The NSGA-II algorithm, based on Pareto optimality theory, achieves co-optimization of multiple conflicting objectives through non-dominated ranking and crowding distance.
[0039] For any two solutions and ,like It is no worse than in all optimization objectives. If it is better than at least one objective, then it is called... Dominate , denoted as: ; Individuals in the population are ranked and sorted according to dominance relationships, with priority given to individuals at the non-dominant level. Crowding distance is also incorporated. The formula for evaluating the uniformity of solution distribution is as follows: ; In the formula, Indicates the number of optimization targets. Indicates the target dimension number. , Indicates the maximum and minimum values of the corresponding target; , They represent the first time. Under the project headings, the next item in the sorting list is... The objective function value on the right side of each individual, at the th Under the project headings, the next item in the sorting list is... The objective function value on the left side of each individual; Through iterative evolution involving selection, crossover, and mutation, NSGA-II can obtain candidate Pareto solutions with good convergence and uniformity. .
[0040] S8. Sort the candidate Pareto solution set according to the predicted pressure drop and the target value, select the optimal solution set sample for simulation calculation, and supplement and update the sample database and iteratively update the radial basis function surrogate model and the multi-objective optimization method until the preset convergence termination condition is met. It should be noted that from the candidate Pareto solution set The simulation is performed by prioritizing sample points that satisfy the pressure drop constraint and have a good objective function, and the results are then incorporated into the sample database. Based on the updated database Retrain the surrogate model and repeat the multi-objective optimization operation; the iterative loop continues until the actual number of simulation samples is reached. The iteration loop terminates when the preset value is reached and the Pareto front distribution no longer changes significantly.
[0041] In this embodiment, in the actual simulation sample The iteration stopped after reaching 320 samples, and the Pareto front was obtained from these 320 simulation samples. The dataset contains 23 samples, and the objective function values for each sample are shown in Table 3. Table 3
[0042] Throughout the optimization process, the radial basis function surrogate model only serves to screen candidate solutions and predict preliminary performance, and is not used as the basis for final performance judgment; all Pareto optimal samples are recalculated through numerical simulation, and subsequent decision-making and parameter output are completed based on simulation and measured data.
[0043] S9, the TOPSIS multi-attribute decision method is used to comprehensively evaluate the candidate Pareto solution set, the weight coefficient of the preset target is set, and the relative closeness between each candidate Pareto solution set and the ideal optimal solution and the negative ideal solution is calculated, so as to obtain the comprehensive optimal flow channel structure based on the solution with the largest relative closeness. It should be noted that the TOPSIS multi-attribute decision method is used to comprehensively evaluate the Pareto solution set. The weight coefficients of each objective are set, and the solution with the highest degree of closeness to the ideal optimal solution and the negative ideal solution is determined by calculating the relative closeness of each solution to the ideal optimal solution and the negative ideal solution. The solution with the highest degree of closeness is determined as the comprehensive optimal flow channel structure scheme, and all geometric parameters of the scheme are output.
[0044] Specifically, step S9 includes steps S91 to S92: S91, construct an initial decision matrix, and normalize the initial decision matrix and introduce weight coefficients to construct a weighted normalized matrix; It should be noted that the expression for the initial decision matrix is: ; In the formula, Let be the initial decision matrix. Indicates the first The candidate solution at the th... Function values on each target Indicates the number of candidate solutions. This indicates the number of targets to be optimized.
[0045] To normalize the decision matrix, weight coefficients are introduced to construct a weighted normalized matrix, expressed as: ; ; ; In the formula, Represents the normalized decision matrix, Indicates the first The candidate solution corresponds to the th... The weighted normalized index value of the evaluation objective. This represents the weighted normalized matrix. Represents the normalized result The value, This represents the weighting coefficient.
[0046] S92, calculate the Euclidean distance from each of the candidate Pareto solution sets to the ideal optimal solution and the negative ideal solution, and calculate the relative proximity. It should be noted that the expressions for calculating the Euclidean distance from each candidate Pareto solution set to the ideal optimal solution and the negative ideal solution are as follows: ; ; In the formula, Indicates the first The Euclidean distance from each solution to the ideal optimal solution. Indicates the first Euclidean distance from the ideal solution to the negative ideal solution. Indicates the target dimension number. Indicates the number of optimization targets. Indicates the first The candidate solution corresponds to the th... The weighted normalized index value of the evaluation objective. Indicates the first The components of the ideal optimal solution corresponding to the project target. Indicates the first The negative ideal solution component corresponding to the project target; The expression for calculating the relative closeness is: ; In the formula, Indicates the first The comprehensive relative proximity of each Pareto candidate solution is used to select the solution with the highest relative proximity as the comprehensive optimal flow channel structure scheme.
[0047] In this embodiment, the weights of net power density, oxygen uniformity, and water saturation are set to 0.6, 0.2, and 0.2, respectively. After applying the TOPSIS multi-attribute decision method, the 16th Pareto front is selected. The resulting flow channel geometric parameters are shown in Table 4. Table 4
[0048] The original flow channel fuel cell before optimization and the fuel cell without added block were selected as the comparison group, and simulation analysis was performed on them under the same conditions as the optimized model; please refer to [link / reference]. Figure 4The polarization curves of fuel cells with three different flow channels were compared. The results showed that, compared with the conventional meandering flow channel without trapezoidal baffles, the meandering flow channel with trapezoidal baffles proposed in this invention can significantly improve the output performance of the fuel cell, especially under high current density conditions. The underlying mechanism is that the trapezoidal baffle structure effectively enhances the fluid disturbance and reactant gas transport capacity within the flow channel, improves the drainage efficiency and oxygen mass transfer conditions in the cathode region, thereby significantly alleviating the oxygen deficiency and flooding problems that easily occur under high current density conditions, ultimately improving battery performance. Through iterative optimization using the multi-objective optimization method proposed in this invention, the potential performance of the flow channel structure was further explored, and the overall comprehensive performance was further enhanced and optimized.
[0049] As shown in Table 5, the comparison results of the three optimization objectives of three fuel cells with different flow channel structures under the same operating conditions show that: compared with the unoptimized flow channel, the net power density of the optimized new flow channel is 9737 W / m³. 2 The net power density increased from 0.747 to 0.768, a 2.8% increase; while the water saturation decreased from 0.0637 to 0.0571, a 10.36% decrease. This demonstrates that the novel flow channel structure and multi-objective optimization method proposed in this invention can achieve synergistic optimization of water saturation, net power density, and oxygen mass fraction uniformity.
[0050] Table 5
[0051] Example 3 Please see Figures 5 to 7 The figure shows a fuel cell in the third embodiment of the present invention, including a cathode flow channel 1 composed of the fuel cell flow channel structure in the first embodiment, and also includes an anode flow channel (not shown in the figure, located on the lower surface of the anode plate 3), a membrane electrode assembly 30, and a cathode plate 9. The anode flow channel is a normal meandering flow channel without trapezoidal baffles 2. The membrane electrode assembly 30 includes an anode diffusion layer 4, an anode catalyst layer 5, a proton exchange membrane 6, a cathode catalyst layer 7, and a cathode diffusion layer 8 stacked in sequence. The anode plate 3 is provided on the lower surface of the anode diffusion layer 4. The cathode flow channel 1 is disposed on the cathode plate 9, and the anode flow channel is disposed on the anode plate 3.
[0052] In summary, the fuel cell flow channel structure, optimization method, and fuel cell in the above embodiments of the present invention can significantly enhance convective disturbance, improve gas diffusion and hydrogen-oxygen transport capabilities, and alleviate local hypoxia and flooding problems. The proposed multi-objective optimization method can carry out synergistic optimization of convective characteristics, gas diffusion performance, and comprehensive output performance, and finally obtain the flow channel structure with optimal comprehensive performance, which is suitable for fuel cell application scenarios with high-speed convection and strong gas diffusion requirements.
[0053] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0054] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.
Claims
1. A method for optimizing the flow channel structure of a fuel cell, characterized in that, The method includes: A parametric geometric model of the meandering flow channel is established, and several design variables of the meandering flow channel are adjusted to generate a three-dimensional model of the meandering flow channel with the corresponding structure. A three-dimensional numerical simulation model of a proton exchange membrane fuel cell with a meandering flow channel structure featuring trapezoidal baffles was established, and the three-dimensional numerical simulation model was verified. Determine the design variable space, and determine the upper and lower bound vectors based on the design variable space; A multi-objective optimization method is determined based on a preset objective, and the multi-objective optimization method is constrained by preset constraints. The Latin hypercube sampling method is used to generate initial sample points to be simulated in the design variable space, and the corresponding target matrix and pressure drop vector are obtained through numerical simulation to construct the initial sample dataset. Using the geometric parameters of the sample as input and the preset target performance index as output, a radial basis function surrogate model is constructed, and a mapping relationship between the design parameters and the radial basis function surrogate model is established. Based on the radial basis function surrogate model, fitness calculation is performed using the NSGA-II algorithm for multi-objective optimization search, in order to find the best solution in the design variable space, and guided by the preset objective of co-optimization, to output the predicted candidate Pareto solution set. The candidate Pareto solution set is sorted according to the predicted pressure drop and the target value, so as to select the optimal solution set sample for simulation calculation, and to supplement and update the sample database and iteratively update the radial basis function surrogate model and the multi-objective optimization method until the preset convergence termination condition is met. The TOPSIS multi-attribute decision method is used to comprehensively evaluate the candidate Pareto solution set. The weight coefficient of the preset target is set, and the relative closeness between each candidate Pareto solution set and the ideal optimal solution and the negative ideal solution is calculated. The comprehensive optimal flow channel structure is obtained based on the solution with the largest relative closeness.
2. The fuel cell flow channel structure optimization method according to claim 1, characterized in that, The relationship between the upper bound vector and the lower bound vector is expressed as follows: ; In the formula, Denotes the lower bound vector. Indicates design variables, This represents the upper bound vector.
3. The fuel cell flow channel structure optimization method according to claim 1, characterized in that, The preset targets include net power density, uniform oxygen mass fraction, and average water saturation at the membrane-catalyst layer interface, defined by the following expressions: ; ; ; In the formula, , , These represent the net power density, the uniformity of oxygen mass fraction distribution at the interface between the cathode gas diffusion layer and the catalyst layer, and the average water saturation at the interface between the membrane and the catalyst layer, respectively. Indicates current density, Indicates battery voltage. This indicates the gas pressure loss at the cathode inlet and outlet. This indicates the volumetric flow rate of the cathode air. This represents the standard deviation of the oxygen mass fraction at the interface. This represents the average mass fraction of oxygen at the interface. Indicates the effective area of the battery. This indicates the water saturation at the interface between the membrane and the catalyst layer.
4. The fuel cell flow channel structure optimization method according to claim 1, characterized in that, The steps of determining a multi-objective optimization method based on a preset objective and constraining the multi-objective optimization method through preset constraints include: A multi-objective optimization method was determined with net power density, oxygen mass fraction uniformity, and water saturation as optimization objectives. The multi-objective optimization method is constrained by the flow channel pressure drop.
5. The fuel cell flow channel structure optimization method according to claim 1, characterized in that, The expression for the radial basis function surrogate model is: ; In the formula, This indicates the performance metrics of the output. Indicates design variables, Indicates the first The design variable vector corresponding to each input sample Indicates the weighting coefficient. This represents the Euclidean distance between the sample point and the point to be predicted. For radial basis kernel functions, Indicates the sample number. This represents the total number of samples in the radial basis function surrogate model.
6. The fuel cell flow channel structure optimization method according to claim 1, characterized in that, The steps for multi-objective optimization search using the NSGA-II algorithm include: Based on the Pareto optimality theory, the NSGA-II algorithm achieves collaborative optimization of multiple conflicting objectives through non-dominated sorting and crowding distance.
7. The fuel cell flow channel structure optimization method according to claim 1, characterized in that, The steps of comprehensively evaluating the candidate Pareto solution set using the TOPSIS multi-attribute decision method, setting the weight coefficients of the preset objective, and calculating the relative closeness of each candidate Pareto solution set to the ideal optimal solution and the negative ideal solution include: An initial decision matrix is constructed, and the initial decision matrix is normalized and weight coefficients are introduced to construct a weighted normalized matrix; Calculate the Euclidean distance from each of the candidate Pareto solution sets to the ideal optimal solution and the negative ideal solution, and calculate the relative proximity.
8. The fuel cell flow channel structure optimization method according to claim 7, characterized in that, The expressions for calculating the Euclidean distance from each of the candidate Pareto solution sets to the ideal optimal solution and the negative ideal solution are as follows: ; ; In the formula, Indicates the first The Euclidean distance from each solution to the ideal optimal solution. Indicates the first Euclidean distance from the ideal solution to the negative ideal solution. Indicates the target dimension number. Indicates the number of targets to be optimized. Indicates the first The candidate solution corresponds to the first... The weighted normalized index value of the evaluation objective. Indicates the first The components of the ideal optimal solution corresponding to the project target. Indicates the first The negative ideal solution component corresponding to the project target; The expression for calculating the relative closeness is: ; In the formula, Indicates the first The comprehensive relative fit of each Pareto candidate solution.
9. A fuel cell flow channel structure, optimized by the fuel cell flow channel structure optimization method according to any one of claims 1 to 8, characterized in that, It includes several cathode channels, which are meandering channels. The meandering channels are arranged in parallel with each other and are continuous meandering extensions to form several periodically alternating waveform segments. Each waveform segment has a trapezoidal baffle at its center.
10. A fuel cell, comprising a cathode flow channel formed by the fuel cell flow channel structure as described in claim 9, characterized in that, It also includes an anode flow channel, a membrane electrode assembly, and a cathode plate. The anode flow channel is a conventional meandering flow channel without trapezoidal baffles. The membrane electrode assembly includes an anode diffusion layer, an anode catalytic layer, a proton exchange membrane, a cathode catalytic layer, and a cathode diffusion layer stacked in sequence. An anode plate is provided on the lower surface of the anode diffusion layer. The cathode flow channel is disposed on the cathode plate, and the anode flow channel is disposed on the anode plate.
