Method and system for optimizing fuel cell simulation based on sensitivity and genetic analysis

The method optimizes fuel cell simulations using sensitivity analysis and genetic algorithms to address computational inefficiencies by prioritizing high-sensitivity parameters, achieving a 20% increase in efficiency.

JP2024164809AActive Publication Date: 2024-11-27CATARC NEW ENERGY VEHICLE TEST CENT (TIANJIN) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
JP2024063928
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-05-15
Filing Date
2024-04-11
Publication Date
2024-11-27
Estimated Expiration
2044-04-11

AI Technical Summary

Technical Problem

Conventional fuel cell simulation optimization methods face significant computational challenges due to the large number of parameter combinations and the lack of consideration for the varying sensitivities of operating conditions, leading to inefficient calculations and high simulation times.

Method used

A method and system utilizing sensitivity analysis and genetic algorithms to optimize fuel cell simulations by quantifying parameter sensitivities, constructing mapping tables, and performing genetic operations to identify optimal parameter combinations efficiently.

Benefits of technology

This approach reduces simulation time and enhances efficiency by focusing on high-sensitivity parameters with precise grids while maintaining accuracy for low-sensitivity parameters, resulting in a 20% increase in computational efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024164809000001_ABST
    Figure 2024164809000001_ABST
Patent Text Reader

Abstract

To provide a method and a system that optimize fuel cell simulation based on, especially, sensitivity and genetic analysis.SOLUTION: A method comprises steps of: constructing a binary mapping table of a sensitivity value through sensitivity analysis of an operation parameter of a fuel cell; randomly coding an operation status parameter to acquire the value within a range of corresponding a code digit according to size of a population formed by the operation status parameter; acquiring a first generation random population code, and performing selection, crossover and mutation operations thereon to acquire a second generation random population code to decode it; converting and acquiring random parameter combination of the operation condition parameter, a power value, and an operation efficiency value; acquiring the corresponding operation condition parameter as a third generation random population through multi-objective optimization processing; continuing iteration until the number of iteration satisfies a stopping criterion; and if satisfied, recording a third generation random population code.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to the field of electrochemical fuel cells, and more particularly to a method and system for optimizing fuel cell simulation based on sensitivity and genetic analysis. [Background technology]

[0002] The operating conditions of a proton exchange membrane fuel cell include many parameters that need to be optimized, such as relative humidity, stoichiometric ratio, operating temperature and pressure, etc. The actual test method of the fuel cell has the problems of long test cycles and high costs, whereas the simulation test is a method with short cycles and low costs.

[0003] When performing simulation optimization, each operating situation parameter needs to be divided in detail to determine the combination of operating situation parameters for optimization. For example, if five values ​​are selected for one operating situation parameter, four operating situation parameters will generate 625 sets of operating situation parameter combinations, which will result in a huge amount of calculation for the simulation test. This method ignores the problem that different operating situation parameters have different degrees of influence on fuel cell performance, and if an operating situation parameter with a small degree of influence on fuel cell performance can be identified, the number of parameter values ​​can be reduced. This greatly reduces the amount of calculation for simulation and further improves the efficiency of simulation optimization.

[0004] In the prior art, the fuel cell operating parameters are generally combined to perform simulation optimization, and the various sensitivities of different operating parameters are ignored during the simulation process. In fact, if the sensitivity of the operating parameters to be analyzed is very low, that is, the impact of the change in the operating parameters on the performance of the fuel cell is small, there is no need to calculate with such high grid accuracy. Otherwise, a lot of simulation time will be wasted in searching for data and completing calculations. Summary of the Invention [Problem to be solved by the invention]

[0005] To solve the problem of huge computational complexity of simulating and optimizing various operating parameters of a conventional fuel cell, the present invention claims the protection of a fuel cell simulation optimization method and system based on sensitivity and genetic analysis. [Means for solving the problem]

[0006] According to a first aspect of the invention, the invention claims protection for a fuel cell simulation optimization method based on sensitivity and genetic analysis, the method comprising the steps of: Performing a sensitivity analysis of the operating status parameters of the fuel cell using a fundamental effect method to obtain a sensitivity value of the operating status parameters of the fuel cell; Building a code number mapping table for the sensitivity value of the operating status parameter of the fuel cell, and forming a mapping relationship between different sensitivity values ​​of the operating status parameter of the fuel cell and the corresponding code number used; According to the sensitivity value of the operating status parameter of the fuel cell and the code number of the sensitivity value mapping table, adopting the code number corresponding to the sensitivity value; According to the size of the population formed by the operating status parameters of the fuel cell, the operating status parameters of the fuel cell are randomly coded within a range of code digits to obtain a first generation random population code; performing selection, crossover, and mutation operations on the first generation random population code to obtain a second generation random population code; Decoding and further converting the second generation random population code to obtain a random parameter combination of the operating status parameters of the fuel cell; sequentially inputting the random parameter combinations into the fuel cell model to obtain power and operating efficiency values ​​for each set of random parameter combinations; Performing a fast non-dominated sorting and crowding calculation of the power value and the operating efficiency value, selecting a population combination of the power value and the operating efficiency whose fitness value is greater than a preset threshold, and making the combination of the operating status of the fuel cell corresponding to the population combination of the power value and the operating efficiency a third generation random population; Randomly coding the third generation random population according to the code digit to obtain a third generation random population code; Judge whether the number of iterations meets a stopping condition, if so, terminate the simulation and record the third generation random population code, if not, select the third generation random population code, and perform the operations of selection, crossover, and mutation again; Includes.

[0007] According to a second aspect of the present invention, a fuel cell simulation optimization system based on sensitivity and genetic analysis is claimed, the system comprising: a sensitivity analysis module for performing sensitivity analysis of the operating status parameters of the fuel cell using a fundamental effect method to obtain sensitivity values ​​of the operating status parameters of the fuel cell; A mapping module for constructing a digit mapping table of sensitivity values ​​of the operating status parameters of the fuel cell, and forming a mapping relationship between different sensitivity values ​​of the operating status parameters of the fuel cell and correspondingly used code digits; a decimal code module for adopting a code decimal number corresponding to the sensitivity value according to the sensitivity value of the operating status parameter of the fuel cell and the decimal number mapping table of the sensitivity value; a random code module for randomly coding the operating status parameters of the fuel cell within a code digit range according to the size of the population formed by the operating status parameters of the fuel cell to obtain a first generation random population code; a genetic analysis module for performing selection, crossover, and mutation operations on the first generation random population code to obtain a second generation random population code; a decoding module for decoding and further converting the second generation random population code to obtain a random parameter combination of the operating status parameters of the fuel cell; a multi-objective optimization module for sequentially inputting random parameter combinations into the fuel cell model to obtain power values ​​and operating efficiency values ​​for each set of random parameter combinations, performing fast non-dominated sorting and crowding calculation of the power values ​​and operating efficiency values, selecting a population combination of power values ​​and operating efficiency whose fitness value is greater than a preset threshold, and taking the combination of the operating status of the fuel cell corresponding to the population combination of power values ​​and operating efficiency as a third-generation random population; The random code module randomly codes the third generation random population according to the code digit to obtain a third generation random population code; a loop judgment module for judging whether the iteration number meets the stopping condition, and if so, terminating the simulation and recording the third generation random population code, and if not, selecting, crossing and mutating the third generation random population code again; Equipped with.

[0008] The present invention claims the protection of a fuel cell simulation optimization method and system based on sensitivity and genetic analysis. Through the sensitivity analysis of the operating status parameters of the fuel cell, a digit mapping table of sensitivity values ​​is constructed, and according to the size of the population formed by the operating status parameters, within the corresponding code digit range, the operating status parameters are randomly coded to obtain values, and a random population code is obtained. After the operations of selection, crossover and mutation, the next generation random population code is obtained and decoded, and the random parameter combination of the operating status parameters and the power value and operating efficiency value under it are converted and obtained, and the corresponding operating status parameters are obtained as the next generation random population through multi-objective optimization processing, and the iteration is continued until the iteration number meets the stopping condition, and the result is recorded. This solution combines the sensitivity quantitative value of the operating status parameters of the fuel cell with the 5-digit genetic code of the genetic algorithm, so that the grid of high sensitivity parameters is refined and the search accuracy is high, and the grid of low sensitivity parameters is sparse and the search accuracy is low, so as to achieve higher simulation efficiency. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is an operational flow diagram of the fuel cell simulation optimization method based on sensitivity and genetic analysis claimed in the present invention. [Diagram 2] FIG. 2 is a second operational flow diagram of the fuel cell simulation optimization method based on sensitivity and genetic analysis claimed in the present invention. [Diagram 3] FIG. 3 is a third operational flow diagram of the fuel cell simulation optimization method based on sensitivity and genetic analysis claimed in the present invention. [Figure 4] FIG. 4 is a fourth operational flow diagram of the fuel cell simulation optimization method based on sensitivity and genetic analysis claimed by the present invention. [Diagram 5]FIG. 5 is a sample spatial distribution diagram of the first generation population before optimization of the fuel cell simulation optimization method based on sensitivity and genetic analysis claimed by the present invention. [Figure 6] FIG. 6 is a sample space distribution diagram of the final generation before optimization of the fuel cell simulation optimization method based on sensitivity and genetic analysis claimed by the present invention. [Figure 7] FIG. 7 is a sample spatial distribution diagram of the optimized first generation population of the fuel cell simulation optimization method based on sensitivity and genetic analysis claimed by the present invention. [Figure 8] FIG. 8 is a sample space distribution diagram of the optimized final generation population of the fuel cell simulation optimization method based on sensitivity and genetic analysis claimed by the present invention. [Figure 9] FIG. 9 is a schematic structural diagram of a fuel cell simulation optimization system based on sensitivity and genetic analysis claimed by the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Hereinafter, the embodiments of the present disclosure will be described in more detail with reference to the drawings. Terms such as "first", "second" and the like used in this application may be used to describe various elements in this specification, but it can be understood that these elements are not limited by these terms unless otherwise specified. These terms are used only to distinguish the first element from another element. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms without being limited by the embodiments described herein. Instead, these embodiments are provided to more fully understand the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0011] Hereinafter, embodiments of the present application will be described in more detail with reference to the accompanying drawings.

[0012] According to a first aspect of the present invention, referring to FIG. 1, the present invention claims a fuel cell simulation optimization method based on sensitivity and genetic analysis, the method comprising: S101: Performing sensitivity analysis of the operating status parameters of the fuel cell using a fundamental effect method to obtain sensitivity values ​​of the operating status parameters of the fuel cell; S201: constructing a code number mapping table of sensitivity values ​​of the operating status parameters of the fuel cell, and forming a mapping relationship between different sensitivity values ​​of the operating status parameters of the fuel cell and the corresponding code number used; S301: According to a sensitivity value of the operating status parameter of the fuel cell and a code number mapping table of the sensitivity value, obtain a code number corresponding to the sensitivity value; S401: According to the size of the population formed by the operating status parameters of the fuel cell, the operating status parameters of the fuel cell are randomly coded within the range of the code digits to obtain a first generation random population code; S501: performing selection, crossover, and mutation operations on a first-generation random population code to obtain a second-generation random population code; S601: Decoding and further converting a second generation random population code to obtain a random parameter combination of operating status parameters of a fuel cell; S701: sequentially inputting random parameter combinations into a fuel cell model to obtain power values ​​and operating efficiency values ​​for each set of random parameter combinations; S801: Perform high-speed non-dominated sorting and crowding calculation of power values ​​and operating efficiency values, select a population combination of power values ​​and operating efficiency whose fitness value is greater than a preset threshold, and set the combination of fuel cell operating conditions corresponding to the population combination of power values ​​and operating efficiency as a third-generation random population; S901: Randomly coding a third-generation random population according to a code digit to obtain a third-generation random population code; S1001: Determine whether the number of iterations meets a stopping condition, and if so, terminate the simulation and record the third generation random population code; Includes. If not, select a third generation random population code, and perform crossover and mutation operations again. The stopping condition is a preset number of iterations.

[0013] Further, referring to FIG. 2, step S101 specifically includes the following steps: S102: setting a value range for an operating status parameter of the fuel cell; S103: standardizing the range of values ​​of the operating status parameters of the fuel cell and inputting them into a random matrix; S104: Obtaining a combination of the value ranges of the operating status parameters by calculating a random matrix, analyzing the combination of the value ranges and inputting it into the fuel cell model to obtain the output voltage of the fuel cell, and calculating the basic effect value; S105: Calculating sensitivity values ​​of operating status parameters of the fuel cell according to the basic effect values; Includes.

[0014] Among them, the principle effect method is a statistical analysis of the degree to which a change in an input parameter affects an output parameter.

[0015] Sensitivity analysis needs to be simulated jointly with the fuel cell model. The specific process is as follows: Setting a value range, a sample dimension, and a number of trajectories for an operating condition parameter of the fuel cell to be analyzed; Standardizing the range of values ​​of the operating parameters of the fuel cell to corresponding values ​​within 0-1 and incorporating them into the constructed random matrix; Generate different combinations of fuel cell operating status parameters in a random matrix, analyze them into actual parameter values ​​and sequentially import them into the fuel cell model, collect the calculated fuel cell voltage results, and calculate the basic effect; Includes.

[0016] The parameter point selection method of the basic effect method is to first extract a reference point from the k-dimensional parameter space, then move the range of the unit lattice length in all parameter dimensions, and finally obtain a parameter trajectory containing (k+1) parameter points. In a three-dimensional space (k=3), X1, X2, and X3 are the respective dimensions, and the reference point x (1) Randomly generate x and move it three times in any parameter dimension over the unit grid length range. (2) , x (3) , x (4) , i.e., the parameter trajectory.

[0017] The method for randomly generating sample trajectories is In a k-dimensional N-level grid space, we define a reference point x * A first step of randomly generating x * Increment Δ in the direction of the random parameter from the first sample point x (1) and a second step to obtain x (1) Increase or decrease the increment Δ in the direction of other parameters from x (2) That is, x (2) =x (1) ±e i Δ and e i is a unit direction vector, and the direction of change is the direction of parameter i; and x (2) From x (3) x (k+1) That is, we generate a randomly generated sequence of sample trajectories x (1) , x (2) , ..., x (k+1) and a fourth step of obtaining This can be summarized as including:

[0018] Since each parameter direction in each sample trajectory can only be used once, two adjacent points can be considered as adding or subtracting a unit length in a random parameter dimension.

[0019] a fifth step of computing a base effect score for the sample trajectories in each dimension; The method also includes a sixth step of randomly generating r sample trajectories according to the above method, calculating r basic effect values ​​for each parameter, calculating basic effect average values ​​and basic effect variances of the r basic effect values, and calculating sensitivity values ​​for each parameter.

[0020] According to the calculation method of the basic effect, the parameter trajectory has (k+1) parameter points, and each parameter can calculate the basic effect value once in each dimension. Based on this, r parameter trajectories are set, and each parameter can calculate the basic effect value r times in each dimension. Therefore, the sensitivity of any parameter in the parameter dimension to any parameter can be expressed by the following formula:

number

[0021] Here, E.E. i j represents the basic effect value of parameter i on the jth parameter trajectory, r represents the number of parameter trajectories, and Σ j=1 r E.E. i j is the basic effect value EE from the 1st parameter trajectory to the rth parameter trajectory. i j represents the addition of u i represents the basic effect mean value, and u i * represents the average absolute value of the basic effect score, and σ i 2 represents the basic effect variance. The basic effect value of the parameters of the part can be greater than 0 or less than 0, so the average value of the basic effect will be smaller or negative. This can lead to a decrease in the sensitivity of the parameter or an insensitive parameter. This is the second type of error in statistics. Therefore, we need to first use the absolute value for the basic effect value, then perform the averaging process to obtain the average of the absolute value of the basic effect value, and complete the correction of the sensitivity value.

[0022] Further, referring to FIG. 3, step S201 specifically includes the following steps: S202: Obtaining maximum and minimum sensitivity values ​​of the operating status parameters of the fuel cell; S203: Dividing a sensitivity value interval according to the number of code digits to obtain a plurality of sensitivity value intervals; S204: The code number is distributed according to the range of the sensitivity value, and a mapping relationship is formed between the distributed code number and the range of the sensitivity value, and a sensitivity value number mapping table is obtained; Includes.

[0023] In this embodiment, in the installed genetic algorithm, the crossover and compilation of genes are performed based on a 5-digit decimal code, which means that the actual parameters need to be coded into a 5-digit decimal code during the process of entering the genetic algorithm, and after the calculation is completed, they need to be decoded into the actual parameters and substituted into the fuel cell model to participate in the calculation.

[0024] The value range of the 5-digit decimal code is 0 to 99999. If x is one of the parameters to be analyzed and y is a randomly generated 5-digit decimal code, the conversion relationship is as follows:

number

[0025] If we select relative humidity as the operating condition to be analyzed and take the range of relative humidity values ​​(x∈[0.1,0.9]) as an example, when one of the relative humidity values ​​is converted to a decimal number and becomes 56234, the actual value of this parameter will be approximately 0.550.

[0026] Although the grid precision coded with five decimal digits is 100,000, if the sensitivity of the operating parameters of the actual analyzed fuel cell is very low, that is, if the change in the operating parameters of the fuel cell has a very small effect on the performance of the fuel cell, it is not necessary to calculate with such high grid precision, otherwise a lot of simulation time will be wasted in searching for data and completing the calculation.

[0027] In this embodiment, the sensitivity quantification values ​​of the operating status parameters of the fuel cell are combined with the five-digit genetic code of the genetic algorithm, so that the grid of high sensitivity parameters is refined and the search accuracy is high, and the grid of low sensitivity parameters is sparse and the search accuracy is low, thereby achieving higher simulation efficiency.

[0028] This embodiment utilizes different hexadecimal code methods to change the grid size. In order to ensure the gene length of the genetic algorithm, five digits are selected as the number of genes corresponding to the operating status parameters of each fuel cell. The five-digit codes corresponding to different hexadecimal numbers and their actual values ​​converted to decimal numbers are shown in Table 1.

[0029] [Table 1]

[0030] Using this method, G 10 max is the maximum value of the randomly generated 5-digit n-ary code converted to a decimal number, and G n Assuming that ' is the decimal value of a randomly generated 5-digit n-ary code, the value of the operating status parameter x of the fuel cell under analysis is as follows:

number

[0031] Using this formula, we select a range of relative humidity values ​​(x∈[0.1,0.9]) as an example. One of the relative humidity values ​​is converted to ternary to become 11202, and by calculation, G n ' becomes 128, and the maximum value of a five-digit ternary number is G 10 max = 242, and the actual value of this parameter is approximately 0.523.

[0032] Based on the conversion relationship between the operating status parameters of the fuel cell to be analyzed and different base numbers, it is necessary to determine which parameters correspond to which level of grid accuracy and perform the analysis.

[0033] According to the sensitivity quantification results of the fuel cell operating status parameters, the sensitivity quantification value range of the fuel cell operating status parameters is between 0.01 and 0.15, and within this range, the code and decode digit levels corresponding to different fuel cell operating status parameters can be defined, as shown in Table 2, which are used in the process of running the genetic algorithm.

[0034] [Table 2]

[0035] Furthermore, step S601 specifically includes the following steps: Decoding the second generation random population code using the formula:

number

[0036] Further, referring to FIG. 4, step S1001 specifically includes the following steps: S1002: According to the decoding result of the second generation random population code, a combination of the fuel cell operating status parameters of power and operating efficiency is selected, and the corresponding combination of the fuel cell operating status parameters is coded according to the digit number of sensitivity to obtain a third generation random population, and all individuals of the third generation random population are sorted non-dominated; S1003: generating a fourth generation random population according to a genetic algorithm operation; S1004: Merging the individuals of the third generation random population and the fourth generation random population as a parent population; S1005: Re-sort the parent population into non-dominated groups, calculate crowding distances, and match the population to the initial state value through survival of the fittest; S1006: Re-running the genetic algorithm operation to generate a new generation of child population; S1007: Determine whether the number of iterations is less than the maximum algebra, and if so, return to the second step; if not, record the data and end the simulation; Includes.

[0037] In this embodiment, in order to better maintain the diversity of individuals in the population, it is necessary to retain individuals with good performance and relatively small density in the population and participate in the next generation of genetic repetition.

[0038] We need to determine the crowding distance between different individuals to decide which superior individual can be preserved. Here, we draw a rectangle around individual i with its crowding distance i d It is necessary to calculate this i d is defined as the sum of the differences of the objective function in different dimensions of two sample points around individual i.

[0039] i dA small value means that there are many individuals around the individual. Therefore, for population diversity in the optimal solution, a larger i d It is necessary to inherit the individuals with this to the next generation and perform the iterative calculation. Next, the rule of population iteration is that when performing population iteration, the individuals with the series of non-dominant individuals larger than a preset threshold must be inherited to the next generation first, and among the individuals with the same series, the individuals with higher crowding degree have the iteration priority.

[0040] To better indicate the output performance and fuel utilization rate of the fuel cell, it is set as a parameter to optimize the power and operation efficiency of the fuel cell.

number

[0041] where P is the fuel cell power (W) and V out is the fuel cell output voltage (V) and I is the current density (A cm -2 ) and A is the effective reaction area (cm -2 ), η is the operating efficiency, and LHV H2 is the low calorific value of hydrogen (MJ kg -1 ), and m H2 is the hydrogen mass flow rate (kg s -1 ) and ST a is the anode stoichiometric ratio, and M H2 is the molar mass of hydrogen (g mol -1 ), and F is the Faraday constant (C mol -1 ) is the case.

[0042] Based on the above, the final multi-objective optimization function is determined as follows:

number

[0043] where f(x) is a multi-objective optimization function, T stack is the fuel cell operating temperature (K) and RH a is the anode relative humidity, RHc is the cathode relative humidity, and ST a is the anode stoichiometric ratio, and ST c is the cathode stoichiometric ratio, and P a is the anode pressure (atm), P c is the cathode pressure (atm).

[0044] In order to show the optimization effect of multi-objective optimization, the evolution algebra needs to be set to a larger value. In order to more intuitively show the excellent computation effect after optimization, we compared the sample space distribution of the 1st generation and 80th generation populations before and after optimization.

[0045] Referring to Figure 5-8, 1.0A / cm -2 The change in the population values ​​during the iterations at a current density of 10 ...

[0046] By continuing to iterate on the sample data before optimization, calculations are performed up to the 96th generation, and the state-of-the-art population size is similar to that of the 80th generation after optimization, and the computational efficiency after optimization is increased by 20% compared to before optimization.

[0047] According to a second embodiment of the present invention, referring to FIG. 9, the present invention claims a fuel cell simulation optimization system based on sensitivity and genetic analysis, the system comprising: a sensitivity analysis module for performing sensitivity analysis of the operating status parameters of the fuel cell using a fundamental effect method to obtain sensitivity values ​​of the operating status parameters of the fuel cell; A mapping module includes a code number mapping table of sensitivity values ​​of the operating status parameters of the fuel cell, and forms a mapping relationship between different sensitivity values ​​of the operating status parameters of the fuel cell and the code number number used correspondingly; a decimal code module for adopting a code decimal number corresponding to the sensitivity value according to the sensitivity value of the operating status parameter of the fuel cell and the decimal number mapping table of the sensitivity value; a random code module for randomly coding the operating status parameters of the fuel cell within a code digit range according to the size of the population formed by the operating status parameters of the fuel cell to obtain a first generation random population code; a genetic analysis module for performing selection, crossover, and mutation operations on the first generation random population code to obtain a second generation random population code; a decoding module for decoding and further converting the second generation random population code to obtain a random parameter combination of the operating status parameters of the fuel cell; a multi-objective optimization module for sequentially inputting random parameter combinations into the fuel cell model to obtain power values ​​and operating efficiency values ​​for each set of random parameter combinations, performing fast non-dominated sorting and crowding calculation of the power values ​​and operating efficiency values, selecting a population combination of power values ​​and operating efficiency whose fitness value is greater than a preset threshold, and taking the combination of the operating status of the fuel cell corresponding to the population combination of power values ​​and operating efficiency as a third-generation random population; The random code module randomly codes the third generation random population according to the code digit to obtain a third generation random population code; a loop judgment module for judging whether the iteration number meets the stopping condition, and if so, terminating the simulation and recording the third generation random population code, and if not, selecting, crossing and mutating the third generation random population code again; Equipped with.

[0048] Furthermore, a sensitivity analysis module for performing sensitivity analysis of the operating status parameters of the fuel cell using the basic effect method and obtaining sensitivity values ​​of the operating status parameters of the fuel cell includes: Setting a range of values ​​for an operating condition parameter of the fuel cell; Standardizing a range of values ​​of the operating parameters of the fuel cell and inputting the range into a random matrix; Calculating the value range combination of the operating status parameters of the fuel cell using a random matrix to obtain the combination of the value range, analyzing and inputting the combination of the value range into the fuel cell model to obtain the output voltage of the fuel cell, and calculating the basic effect value; Calculating sensitivity values ​​of operating status parameters of the fuel cell based on the basic effect values; Specifically includes:

[0049] A mapping module includes a code number mapping table for the sensitivity value of the operating status parameter of the fuel cell, and forms a mapping relationship between different sensitivity values ​​of the operating status parameter of the fuel cell and the corresponding code number used; Obtaining maximum and minimum sensitivity values ​​of the operating status parameters of the fuel cell; Dividing a sensitivity value interval according to the number of code digits to obtain a plurality of sensitivity value intervals; The code number is distributed according to the range of the sensitivity value, and a mapping relationship is formed between the distributed code number and the range of the sensitivity value, and a sensitivity value number mapping table is obtained; Specifically includes:

[0050] Further, a decoding module for decoding and further converting the second generation random population code to obtain a random parameter combination of the operating status parameters of the fuel cell includes: Specifically, this involves decoding the second generation random population code using the following formula:

number

[0051] A loop judgment module judges whether the iteration number meets the stopping condition, and if so, terminates the simulation and records the third generation random population code; if not, selects the third generation random population code, and performs the crossover and mutation operations again; According to the decoding result of the second generation random population code, a combination of the fuel cell operating status parameters of power and operating efficiency is selected, and the corresponding combination of the fuel cell operating status parameters is coded according to the digit number of sensitivity to obtain a third generation random population, and all the individuals of the third generation random population are sorted in a non-dominant manner; generating a fourth generation random population according to a genetic algorithm operation; Merging the individuals of the third generation random population and the fourth generation random population as a parent population; and Re-sort the parent population into non-dominated groups, calculate crowding distances, and match the population to the initial state value through survival of the fittest; re-running the genetic algorithm operation to generate a new generation of offspring population; determining whether the number of iterations is less than the maximum algebra, and if so, returning to the second step, and if not, recording the data and terminating the simulation; Specifically includes:

[0052] In addition, the multi-objective optimization module: Specifically, the value of the multi-objective optimization function is determined based on the power value and the operating efficiency value.

[0053] Those skilled in the art will appreciate that the disclosed subject matter may be modified or improved in various ways. For example, the various devices or components described above may be realized by hardware, software, firmware, or a combination thereof.

[0054] The flow charts of the present invention are used to explain the steps of the method according to the embodiments of the present invention. It should be noted that the steps do not necessarily follow the exact order, but rather various steps may be performed in reverse order or simultaneously, and other operations may be added to these processes.

[0055] Those skilled in the art will understand that all or part of the steps of the above-mentioned method may be completed by instructing relevant hardware by a computer program, and the program may be stored in a computer-readable storage medium such as a read-only memory, a magnetic disk, or an optical disk. Alternatively, all or part of the steps of the above-mentioned embodiments may be realized using one or more integrated circuits. In addition, each module / unit in the above-mentioned embodiments may be realized by hardware or a software function module. The present disclosure is not limited to a specific form of combination of hardware and software.

[0056] All terms used herein, unless otherwise specified, have the same meaning as commonly understood by a person of ordinary skill in the art to which this disclosure pertains. It should also be understood that terms defined in common dictionaries should not be interpreted in an idealized or overly formal sense unless expressly defined herein, but should be interpreted as having a meaning consistent with the meaning in the context of the relevant art.

[0057] The above is a description of the present invention, but is not intended to be limiting. Although some exemplary embodiments of the present disclosure have been described, those skilled in the art will readily appreciate that many modifications can be made to the exemplary embodiments without departing from the novel teachings and advantages of the present disclosure. Accordingly, such modifications are intended to be included within the scope of the present invention as set forth in the claims. The above is a description of the present invention, and should not be considered limited to the specific embodiments disclosed, and it is understood that modifications to the disclosed embodiments and other embodiments are included within the scope of the appended claims. The present disclosure is defined by the claims and their equivalents.

[0058] In the description herein, the terms "one embodiment," "some embodiments," "exemplary embodiments," "example," "particular example," or "some examples" mean that the particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. It should be noted that, in the present specification, generalized expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0059] The above description is merely a preferred embodiment of the present application, and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A fuel cell simulation optimization method based on sensitivity and genetic analysis, comprising: Performing sensitivity analysis of the operating status parameters of the fuel cell using a fundamental effect method to obtain sensitivity values ​​of the operating status parameters of the fuel cell; Building a code number mapping table of the sensitivity value of the operating status parameter of the fuel cell, and forming a mapping relationship between different sensitivity values ​​of the operating status parameter of the fuel cell and the corresponding code number used; According to a sensitivity value of the operating status parameter of the fuel cell and a sensitivity value decimal mapping table, adopting a code decimal number corresponding to the sensitivity value; According to the size of the population formed by the operating status parameters of the fuel cell, the operating status parameters of the fuel cell are randomly coded within the range of the code digits to obtain a first generation random population code; performing selection, crossover and mutation operations on the first generation random population code to obtain a second generation random population code; Decoding and further converting the second generation random population code to obtain a random parameter combination of operating status parameters of the fuel cell; sequentially inputting the random parameter combinations into a fuel cell model to obtain power and operating efficiency values ​​for each set of the random parameter combinations; Performing a fast non-dominated sorting and crowding calculation of the power value and the operating efficiency value, selecting a population combination of the power value and the operating efficiency whose fitness value is greater than a preset threshold, and making the combination of the operating status of the fuel cell corresponding to the population combination of the power value and the operating efficiency a third generation random population; Randomly coding the third generation random population according to the code digits to obtain a third generation random population code; Judge whether the number of iterations meets a stopping condition, if so, terminate the simulation, and record the third generation random population code; if not, perform the selection, crossover, and mutation operations of the third generation random population code again; Including, Performing a sensitivity analysis of the operating status parameters of the fuel cell using the fundamental effect method to obtain sensitivity values ​​of the operating status parameters of the fuel cell, setting a value range for an operating condition parameter of the fuel cell; standardizing the range of values ​​of the operating parameters of the fuel cell and inputting them into a random matrix; A combination of the value ranges of the operating parameters is obtained by calculating the random matrix, and the combination of the value ranges is analyzed and input into a fuel cell model to obtain an output voltage of the fuel cell, and calculate a basic effect value; Calculating sensitivity values ​​of operating status parameters of the fuel cell based on the basic effect values; Specifically including: Building a code number mapping table of the sensitivity value of the operating status parameter of the fuel cell, and forming a mapping relationship between different sensitivity values ​​of the operating status parameter of the fuel cell and the corresponding code number used; Obtaining maximum and minimum sensitivity values ​​of the operating status parameters of the fuel cell; Dividing the sensitivity value interval according to the number of the code digits to obtain a plurality of sensitivity value intervals; The code number is distributed according to the range of the sensitivity value, and a mapping relationship is formed between the distributed code number and the range of the sensitivity value, thereby obtaining a sensitivity value-number mapping table; Specifically including: A fuel cell simulation optimization method based on sensitivity and genetic analysis.

2. 2. The method for optimizing a fuel cell simulation based on sensitivity and genetic analysis according to claim 1, comprising: The decoding and further converting of the second generation random population code to obtain a random parameter combination of the operating status parameters of the fuel cell specifically includes: wherein the second generation random population code is decoded using the formula: [0080] G n max is a value obtained by converting a maximum 5-digit n-ary code corresponding to the operating status parameter of the fuel cell into a decimal number, and G n ' is the value obtained by converting the 5-digit n-ary code of the second generation random population code into a decimal number, and x min is the minimum value of the operating status parameter of the fuel cell, and x max is the maximum value of the operating condition parameter of the fuel cell, and x is the value of the decoded operating condition parameter; A fuel cell simulation optimization method based on sensitivity and genetic analysis.

3. 2. The method for optimizing a fuel cell simulation based on sensitivity and genetic analysis according to claim 1, comprising: Determining whether the number of iterations satisfies a stopping condition, terminating the simulation if it satisfies the stopping condition, and recording the third generation random population code, and if it does not satisfy the stopping condition, performing the operations of selection, crossover, and mutation on the third generation random population code again. According to the decoding result of the second generation random population code, a combination of the fuel cell operating status parameters of power and operating efficiency is selected, and the corresponding combination of the fuel cell operating status parameters is coded according to the digit number of sensitivity to obtain the third generation random population, and all the individuals of the third generation random population are sorted in a non-dominant manner; generating a fourth generation random population according to a genetic algorithm operation; merging the individuals of the third generation random population and the fourth generation random population as a parent population; re-sorting the parent population into non-dominated groups, calculating crowding distances, and matching the population to the initial state value through survival of the fittest; re-running the genetic algorithm operation to generate a new generation of offspring population; determining whether the number of iterations is less than a maximum algebra, and if so, returning to the second step, otherwise recording the data and terminating the simulation; Specifically including: A fuel cell simulation optimization method based on sensitivity and genetic analysis.

4. 2. The method for optimizing a fuel cell simulation based on sensitivity and genetic analysis according to claim 1, comprising: performing a fast non-dominated sort and a crowding calculation of the power value and the operating efficiency value, selecting a population combination of the power value and the operating efficiency whose fitness value is greater than a preset threshold value, and making a combination of the operating status of the fuel cell corresponding to the population combination of the power value and the operating efficiency into a third generation random population; Specifically including determining a value of a multi-objective optimization function based on the power value and the operational efficiency value. A fuel cell simulation optimization method based on sensitivity and genetic analysis.

5. A fuel cell simulation optimization system based on sensitivity and genetic analysis, comprising: a sensitivity analysis module for performing sensitivity analysis of the operating status parameters of the fuel cell using a fundamental effect method to obtain sensitivity values ​​of the operating status parameters of the fuel cell; A mapping module includes a code number mapping table of the sensitivity value of the operating status parameter of the fuel cell, and forms a mapping relationship between different sensitivity values ​​of the operating status parameter of the fuel cell and the corresponding code number used; a decimal code module for adopting a code decimal number corresponding to the sensitivity value according to a decimal mapping table of the sensitivity value of the operating status parameter of the fuel cell and the decimal number of the sensitivity value; A random code module randomly codes the operating status parameters of the fuel cell according to the size of the population formed by the operating status parameters of the fuel cell within the range of the code digits to obtain a first generation random population code; A genetic analysis module performs selection, crossover and mutation operations on the first generation random population code to obtain a second generation random population code; a decoding module for decoding and further converting the second generation random population code to obtain a random parameter combination of operating status parameters of the fuel cell; a multi-objective optimization module for sequentially inputting the random parameter combinations into a fuel cell model to obtain power values ​​and operating efficiency values ​​for each set of the random parameter combinations, performing fast non-dominated sorting and crowding degree calculation of the power values ​​and operating efficiency values, selecting a population combination of power values ​​and operating efficiencies whose fitness value is greater than a preset threshold, and setting the combination of the operating status of the fuel cell corresponding to the population combination of power values ​​and operating efficiencies as a third generation random population; A random code module randomly codes the third generation random population according to the code digit to obtain a third generation random population code; A loop judgment module judges whether the number of iterations meets a stopping condition, and if so, ends the simulation and records the third generation random population code, and if not, selects, crosses and mutates the third generation random population code again; Equipped with The sensitivity analysis module is configured to perform a sensitivity analysis of the operating status parameters of the fuel cell using a fundamental effect method to obtain a sensitivity value of the operating status parameters of the fuel cell, setting a value range for an operating condition parameter of the fuel cell; standardizing the range of values ​​of the operating parameters of the fuel cell and inputting them into a random matrix; A combination of the value ranges of the operating status parameters of the fuel cell is obtained by calculating the random matrix, and the combination of the value ranges is analyzed and input into a fuel cell model to obtain an output voltage of the fuel cell, and calculate a basic effect value; Calculating sensitivity values ​​of operating status parameters of the fuel cell based on the basic effect values; Specifically including: The mapping module includes a code number mapping table of the sensitivity value of the operating status parameter of the fuel cell, and forms a mapping relationship between different sensitivity values ​​of the operating status parameter of the fuel cell and the corresponding code number used, Obtaining maximum and minimum sensitivity values ​​of the operating status parameters of the fuel cell; Dividing the sensitivity value interval according to the number of the code digits to obtain a plurality of sensitivity value intervals; The code number is distributed according to the range of the sensitivity value, and a mapping relationship is formed between the distributed code number and the range of the sensitivity value, thereby obtaining a sensitivity value-number mapping table; Specifically including: A fuel cell simulation optimization system based on sensitivity and genetic analysis.

6. 6. The fuel cell simulation optimization system based on sensitivity and genetic analysis according to claim 5, The decoding module for decoding and converting the second generation random population code to obtain a random parameter combination of the operating status parameters of the fuel cell is specifically configured to: decoding said second generation random population code using the formula: [0090] G n max is a value obtained by converting a maximum 5-digit n-ary code corresponding to the operating status parameter of the fuel cell into a decimal number, and G n ' is the value obtained by converting the 5-digit n-ary code of the second generation random population code into a decimal number, and x min is the minimum value of the operating status parameter of the fuel cell, and x max is the maximum value of the operating condition parameter of the fuel cell, x is the value of the decoded operating condition parameter, The loop judgment module judges whether the iteration number meets a stopping condition, and if so, ends the simulation and records the third generation random population code; if not, selects, crosses and mutates the third generation random population code again; According to the decoding result of the second generation random population code, a combination of the fuel cell operating status parameters of power and operating efficiency is selected, and the corresponding combination of the fuel cell operating status parameters is coded according to the digit number of sensitivity to obtain the third generation random population, and all the individuals of the third generation random population are sorted in a non-dominant manner; generating a fourth generation random population according to a genetic algorithm operation; merging the individuals of the third generation random population and the fourth generation random population as a parent population; re-sorting the parent population into non-dominated groups, calculating crowding distances, and matching the population to the initial state value through survival of the fittest; re-running the genetic algorithm operation to generate a new generation of offspring population; determining whether the number of iterations is less than a maximum algebra, and if so, returning to the second step, otherwise recording the data and terminating the simulation; Specifically including: A fuel cell simulation optimization system based on sensitivity and genetic analysis.

7. 7. A fuel cell simulation optimization system based on sensitivity and genetic analysis according to claim 6, comprising: The multi-objective optimization module includes: Specifically including determining a value of a multi-objective optimization function based on the power value and the operational efficiency value. A fuel cell simulation optimization system based on sensitivity and genetic analysis.