Thermal management method and device of air-cooled fuel cell, electronic equipment and storage medium

Abstract

The invention relates to a thermal management method and device of an air-cooled fuel cell, electronic equipment and a storage medium, and the method comprises the steps: when the air-cooled fuel cell works actually, according to the current environment temperature and target output power, initial operation parameters are determined by using a linear interpolation method, and the operation parameters comprise a fan rotating speed and a hydrogen exhaust valve opening period; the initial operation parameters are used as starting points, and an improved chaotic particle swarm algorithm is used for optimizing the operation parameters; in the operation parameter optimization processing process, each time a group of new operation parameters are generated, the new operation parameters are input into a pre-trained neural network performance prediction model so as to predict the performance parameters of the air-cooled fuel cell corresponding to the current operation parameters; and according to the multi-target optimization function and the performance parameters corresponding to each group of operation parameters, determining target operation parameters so as to obtain the optimal operation condition of the air-cooled fuel cell.

Description

Technical Field

[0001] The present invention relates to the technical field of air-cooled fuel cells, and in particular to a thermal management method, device, electronic equipment and storage medium for air-cooled fuel cells. Background Art

[0002] As an efficient and clean energy, hydrogen energy has received more and more attention in recent years. Air-cooled fuel cells are widely used in low and medium power demand scenarios due to their advantages such as low operating temperature, zero emission, high power density, light weight, simple system and low cost, especially in the fields of drones, express delivery vehicles, sightseeing vehicles, etc.

[0003] The key components of the air-cooled fuel cell system include the fuel cell, hydrogen cylinder, pressure reducing valve, fan and other valves and sensors. The fan provides cooling air and reaction gas (oxygen) and discharges the water produced on the cathode side; the hydrogen cylinder is always open and the water produced on the anode side is discharged by periodically opening the exhaust valve. The fan speed and the opening cycle of the hydrogen exhaust valve are key operating parameters.

[0004] However, thermal management of air-cooled fuel cells is a technical difficulty. Compared with water-cooled fuel cells, thermal management is more complicated. The fan must provide both cooling air and reactant gas, resulting in complete coupling of temperature control and reactant gas flow control. Increasing or decreasing the fan speed will affect both the heat dissipation rate and battery performance. In addition, the opening cycle of the hydrogen exhaust valve also needs to be carefully adjusted: a cycle that is too short will lead to hydrogen loss, and a cycle that is too long may cause liquid water accumulation, causing water blockage in the fuel cell, making it difficult to balance hydrogen utilization and drainage efficiency. Summary of the invention

[0005] The present invention aims to solve one of the technical problems in the related art at least to a certain extent.

[0006] To this end, the first purpose of the present invention is to propose a thermal management method for an air-cooled fuel cell. The method uses an improved chaotic particle swarm algorithm to perform global optimization on the fan speed and the opening period of the hydrogen exhaust valve of the air-cooled fuel cell, and predicts the performance of the fuel cell based on a pre-trained neural network model, thereby obtaining the optimal operating conditions.

[0007] A second objective of the present invention is to provide a thermal management device for an air-cooled fuel cell.

[0008] A third objective of the present invention is to provide an electronic device.

[0009] A fourth object of the present invention is to provide a computer-readable storage medium.

[0010] A fifth object of the present invention is to provide a computer program product.

[0011] To achieve the above-mentioned purpose, the first embodiment of the present invention proposes a thermal management method for an air-cooled fuel cell, including: when the air-cooled fuel cell is actually working, according to the current ambient temperature and the target output power, the initial operating parameters are determined by linear interpolation method; wherein the operating parameters include fan speed and hydrogen exhaust valve opening period; taking the initial operating parameters as the starting point, the operating parameters are optimized using an improved chaotic particle swarm algorithm; in the process of optimizing the operating parameters, each time a new set of operating parameters is generated, the new operating parameters are input into a pre-trained neural network performance prediction model to predict the performance parameters of the air-cooled fuel cell corresponding to the current operating parameters; and the target operating parameters are determined according to the multi-objective optimization function and the performance parameters corresponding to each set of operating parameters.

[0012] According to the thermal management method of the air-cooled fuel cell of the embodiment of the present invention, when the air-cooled fuel cell is actually working, the initial operating parameters are determined by linear interpolation method according to the current ambient temperature and the target output power, and the operating parameters include the fan speed and the opening period of the hydrogen exhaust valve. Then, the operating parameters are optimized by using the improved chaotic particle swarm algorithm with the initial operating parameters as the starting point. In the process of optimizing the operating parameters, each time a new set of operating parameters is generated, the new operating parameters are input into the pre-trained neural network performance prediction model to predict the performance parameters of the air-cooled fuel cell corresponding to the current operating parameters. Finally, the target operating parameters are determined according to the multi-objective optimization function and the performance parameters corresponding to each set of operating parameters. Therefore, the method uses the improved chaotic particle swarm algorithm to globally optimize the fan speed and the opening period of the hydrogen exhaust valve of the air-cooled fuel cell, and predicts the performance of the fuel cell according to the pre-trained neural network model, so as to obtain the optimal operating conditions.

[0013] In addition, the thermal management method for an air-cooled fuel cell provided in the first aspect of the present invention may also have the following additional technical features:

[0014] In one embodiment of the present invention, the performance parameter includes the total voltage V sum , Maximum single chip voltage V max , minimum single chip voltage V min and hydrogen utilization Determining the target operating parameters according to the multi-objective optimization function and the performance parameters corresponding to each set of operating parameters includes:

[0015] Substituting the performance parameters corresponding to each set of operating parameters into the multi-objective optimization function to obtain the quality value corresponding to each performance parameter;

[0016] Determining the target operating parameters according to a plurality of merit values;

[0017] Wherein, the expression of the multi-objective optimization function is:

[0018]

[0019] Among them, a, b, and c are the first coefficient, the second coefficient, and the third coefficient respectively.

[0020] In one embodiment of the present invention, the training process of the neural network performance prediction model includes:

[0021] Under different ambient temperatures and different output powers, a plurality of different operating parameters are given, and corresponding performance parameters are calibrated to form a training data set for the neural network performance prediction model;

[0022] Setting neural network parameters to obtain an initial neural network performance prediction model;

[0023] The initial neural network performance prediction model is trained using the training data set to obtain a trained neural network performance prediction model.

[0024] In one embodiment of the present invention, the method further comprises:

[0025] After the air-cooled fuel cell operates according to the target operating parameters, recording the performance parameters of the air-cooled fuel cell;

[0026] The training data set of the neural network performance prediction model is modified according to the recorded performance parameters and corresponding operating conditions.

[0027] To achieve the above-mentioned purpose, the second embodiment of the present invention proposes a thermal management device for an air-cooled fuel cell, including: a first determination module, which is used to determine the initial operating parameters by linear interpolation method according to the current ambient temperature and the target output power when the air-cooled fuel cell is actually working; wherein the operating parameters include the fan speed and the opening period of the hydrogen exhaust valve; an optimization module, which is used to optimize the operating parameters using an improved chaotic particle swarm algorithm based on the initial operating parameters; a prediction module, which is used to input the new operating parameters into a pre-trained neural network performance prediction model each time a new set of operating parameters is generated during the optimization process of the operating parameters, so as to predict the performance parameters of the air-cooled fuel cell corresponding to the current operating parameters; a second determination module, which is used to determine the target operating parameters based on the multi-objective optimization function and the performance parameters corresponding to each set of operating parameters.

[0028] According to the thermal management device of the air-cooled fuel cell of the embodiment of the present invention, the first determination module determines the initial operating parameters according to the current ambient temperature and the target output power by using the linear interpolation method when the air-cooled fuel cell is actually working, and the operating parameters include the fan speed and the opening period of the hydrogen exhaust valve. The optimization module uses the initial operating parameters as the starting point and uses the improved chaotic particle swarm algorithm to optimize the operating parameters. In the process of optimizing the operating parameters, the prediction module generates a new set of operating parameters each time, and inputs the new operating parameters into the pre-trained neural network performance prediction model to predict the performance parameters of the air-cooled fuel cell corresponding to the current operating parameters. The second determination module determines the target operating parameters according to the multi-objective optimization function and the performance parameters corresponding to each set of operating parameters. Thus, the device uses the improved chaotic particle swarm algorithm to perform global optimization on the fan speed and the opening period of the hydrogen exhaust valve of the air-cooled fuel cell, and predicts the fuel cell performance according to the pre-trained neural network model, so as to obtain the optimal operating conditions.

[0029] In addition, the thermal management device for an air-cooled fuel cell provided in the second aspect of the present invention may also have the following additional technical features:

[0030] In one embodiment of the present invention, the performance parameter includes the total voltage V sum , Maximum single chip voltage V max , minimum single chip voltage V min and hydrogen utilization The second determination module is used to determine the target operating parameters according to the multi-objective optimization function and the performance parameters corresponding to each group of operating parameters, including:

[0031] Substituting the performance parameters corresponding to each set of operating parameters into the multi-objective optimization function to obtain the quality value corresponding to each performance parameter;

[0032] Determining the target operating parameters according to a plurality of merit values;

[0033] Wherein, the expression of the multi-objective optimization function is:

[0034]

[0035] Among them, a, b, and c are the first coefficient, the second coefficient, and the third coefficient respectively.

[0036] In one embodiment of the present invention, the training process of the neural network performance prediction model in the prediction module includes:

[0037] Under different ambient temperatures and different output powers, a plurality of different operating parameters are given, and corresponding performance parameters are calibrated to form a training data set for the neural network performance prediction model;

[0038] Setting neural network parameters to obtain an initial neural network performance prediction model;

[0039] The initial neural network performance prediction model is trained using the training data set to obtain a trained neural network performance prediction model.

[0040] In one embodiment of the present invention, the device further comprises:

[0041] A correction module is used to record the performance parameters of the air-cooled fuel cell after the air-cooled fuel cell operates according to the target operating parameters, and to correct the training data set of the neural network performance prediction model according to the recorded performance parameters and corresponding operating conditions.

[0042] To achieve the above object, the third aspect of the present invention further provides an electronic device, including:

[0043] at least one processor; and,

[0044] a memory communicatively connected to the at least one processor; wherein,

[0045] The memory stores instructions that can be executed by the at least one processor. The instructions are executed by the at least one processor so that the at least one processor can perform the above method.

[0046] The electronic device of the embodiment of the present invention executes the above method and uses the improved chaotic particle swarm algorithm to globally optimize the fan speed of the air-cooled fuel cell and the opening period of the hydrogen exhaust valve, and predicts the performance of the fuel cell according to the pre-trained neural network model, thereby obtaining the optimal operating conditions.

[0047] To achieve the above objectives, the fourth aspect of the present invention further proposes a computer-readable storage medium, wherein the computer instructions are used to enable the computer to execute the above method.

[0048] The computer-readable storage medium of an embodiment of the present invention executes the above method and uses an improved chaotic particle swarm algorithm to perform global optimization on the fan speed of the air-cooled fuel cell and the opening period of the hydrogen exhaust valve, and predicts the performance of the fuel cell according to a pre-trained neural network model, thereby obtaining the optimal operating conditions.

[0049] To achieve the above-mentioned purpose, a fifth aspect of the present invention further provides a computer program product, which executes the above-mentioned method when an instruction processor in the computer program product executes it.

[0050] The computer program product of the embodiment of the present invention executes the above method and uses an improved chaotic particle swarm algorithm to globally optimize the fan speed of the air-cooled fuel cell and the opening period of the hydrogen exhaust valve, and predicts the performance of the fuel cell based on a pre-trained neural network model, thereby obtaining the optimal operating conditions.

[0051] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0053] Figure 1 is a flow chart of a thermal management method for an air-cooled fuel cell according to an embodiment of the present invention;

[0054] Figure 2 is a schematic diagram of an air-cooled fuel cell and a system thereof according to an embodiment of the present invention;

[0055] Figure 3 4 is a block diagram of a thermal management device for an air-cooled fuel cell according to an embodiment of the present invention. DETAILED DESCRIPTION

[0056] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0057] The following describes a thermal management method, device, electronic device, and storage medium for an air-cooled fuel cell according to an embodiment of the present invention with reference to the accompanying drawings.

[0058] Figure 1 is a flow chart of a thermal management method for an air-cooled fuel cell according to an embodiment of the present invention.

[0059] like Figure 1 As shown, the thermal management method of an air-cooled fuel cell according to an embodiment of the present invention includes:

[0060] S1, when the air-cooled fuel cell is actually working, the initial operating parameters are determined by linear interpolation method according to the current ambient temperature and the target output power; wherein the operating parameters include the fan speed and the opening period of the hydrogen exhaust valve.

[0061] It should be noted that when determining the initial fan speed and the initial hydrogen exhaust valve opening period, it is also necessary to determine the upper and lower limits of the fan speed and the hydrogen exhaust valve opening period.

[0062] Preferably, in the air-cooled fuel cell system, the output pressure of the hydrogen pressure reducing valve is 30 kPa.

[0063] Preferably, the air-cooled fuel cell has a rated output power of 2 kW. The target output power needs to be less than the rated output power.

[0064] Preferably, for the air-cooled fuel cell, the lower limit of the fan speed is the fan speed corresponding to the air stoichiometric ratio of 1.9, and the upper limit is the fan speed corresponding to the air stoichiometric ratio of 5.3. The upper limit of the opening cycle of the hydrogen exhaust valve is once every 20 seconds, and the lower limit of the opening cycle of the hydrogen exhaust valve is once every 7 seconds.

[0065] For example, at 0.5A / cm 2 The duty cycle that can be operated at each current working point is artificially given, and the duty cycle compensation amount is given according to the room temperature. When the current is constant, two adjacent current points are selected, and the duty cycle is obtained by linear interpolation. It is added to the duty cycle compensation amount selected according to the room temperature to obtain the initial value of the duty cycle at each room temperature and current. The opening cycle of the hydrogen exhaust valve is uniformly set to 14s.

[0066] S2, taking the initial operating parameters as the starting point, uses the improved chaotic particle swarm algorithm to optimize the operating parameters.

[0067] S3, during the process of optimizing the operating parameters, each time a new set of operating parameters is generated, the new operating parameters are input into the pre-trained neural network performance prediction model to predict the performance parameters of the air-cooled fuel cell corresponding to the current operating parameters. The performance parameters include the total voltage V sum , Maximum single chip voltage V max , minimum single chip voltage V min and hydrogen utilization

[0068] In the embodiment of the present invention, an air-cooled fuel cell with 50 sections and its system are used, including fuel cells, hydrogen cylinders and valves, fans, sensors and other components, such as Figure 2 As shown in Figure 1, the training process of the neural network performance prediction model is as follows:

[0069] Under different ambient temperatures and different output powers, multiple groups of different operating parameters are given, and the corresponding performance parameters are calibrated to form a training data set for the neural network performance prediction model; the neural network parameters are set to obtain an initial neural network performance prediction model; the initial neural network performance prediction model is trained using the training data set to obtain a trained neural network performance prediction model.

[0070] Among them, the training data set can be calibrated under different working conditions according to the following parameter table (Table 1):

[0071] Table 1

[0072]

[0073]

[0074]

[0075] For each working condition, the current is loaded from small to large, and the current is loaded at a certain interval. Each working condition has at least 12 data points, and a total of 63 working conditions. However, for some working conditions with large and small duty cycles, such as the working condition of 15% duty cycle, the phenomenon of insufficient gas occurs when the load is pulled to a certain extent. This working condition has less than 12 data points. After testing all the above working conditions, about 500 data points can be obtained, which can form a neural network training set.

[0076] Based on the above data as the training data set, the neural network parameters are set, including the maximum number of iterations, the number of network layers, the number of neurons in each layer and the activation function, initial value, learning rate, error function and other parameters, to train the neural network performance prediction model.

[0077] It should be noted that the particle swarm algorithm is a heuristic swarm intelligence algorithm. It is an optimization algorithm born from the study and simulation of the group behavior of birds and fish. It regards the individuals in the population as particles flying in a D-dimensional space. After the particles are randomly placed in the solution space, all particles will fly in the solution space at a certain speed and present a continuous trend towards their own historical optimal solution p. best and the global historical optimal solution g best The trend of getting closer, thus achieving the evolution of the population. As described in formulas (1) to (6):

[0078] In the D-dimensional solution space, there are N particles forming a population, where the i-th particle is represented by a D-dimensional vector, which indicates the position of the particle in the solution space, expressed as the following formula (1):

[0079] X i =(x i1 ,x i2 ,...,xiD ),i=1,2,...N (1)

[0080] The flight speed Vi of the i-th particle is also a D-dimensional vector, expressed as the following formula (2):

[0081] V i =(v i1 ,v i2 ,...,v iD ),i=1,2,...N(2)

[0082] The historical optimal position solution of the i-th particle is its own historical optimal solution p best , expressed as the following formula (3):

[0083] p best =(p i1 ,p i2 ,...,p iD ),i=1,2,...N (3)

[0084] The global optimal position solution of the entire particle swarm population is the global historical optimal solution g best , expressed as the following formula (4):

[0085] g best =(g1,g2,...,g D ) (4)

[0086] The particle updates its velocity and position according to the following formulas, which are expressed as the following formulas (5) and (6) respectively:

[0087] v ij (t+1)=v ij (t)+c1r1(t)[p ij (t)-x ij (t)]+c2r2(t)[p gj (t)-x ij (t)](5)

[0088] x ij (t+1)=x ij (t)+v ij (t+1)(6)

[0089] In the above formulas (5) and (6), v ij(t) represents the j-th dimension component of the parameters X, V, and P corresponding to the particle i in the k-th iteration. c1 and c2 are called learning factors, which reflect the influence of the individual experience and population experience of the particle on the particle. r1 and r2 are random numbers between [0,1], which increase the randomness of the particle flight. It can be seen that the speed of the particle is composed of three parts. One part is the previous speed, which represents inertia, in order to ensure the global convergence ability of the algorithm. The last two parts represent the particle moving towards its own historical optimal solution p best and the global historical optimal solution g best The trend of approaching.

[0090] Subsequently, on the basis of the above formula (5), the inertia weight is introduced to solve the problem of global or local non-convergence. The improved velocity update formulas are shown in formulas (7) and (8).

[0091] v ij (t+1)=w·v ij (t)+c1r1(t)[p ij (t)-x ij (t)]+c2r2(t)[p gj (t)-x ij (t)](7)

[0092]

[0093] Where, T max Indicates the maximum number of iterations; w min represents the minimum inertia weight; w max represents the maximum inertia weight; t represents the current iteration number.

[0094] Heuristic swarm intelligence algorithms generally have the disadvantage of being easily trapped in local optimal solutions. In order to improve the global search capability of the algorithm, the speed change formula (7) of the standard particle swarm algorithm is modified to make the learning factor c2 change according to the sinusoidal chaos sequence as the number of iterations increases, and the following formula (9) is obtained:

[0095]

[0096] The dynamic properties of the chaotic map are important to ensure that the solutions generated by the algorithm are diverse enough to reach every state in the search space. Therefore, compared with standard random search, chaotic search is more likely to escape from the local optimal solution. The random parameters in the particle swarm algorithm can be replaced by a one-dimensional chaotic map, which is called chaotic optimization. Due to the random ergodic nature of chaos, the chaotic optimization algorithm converges faster than the standard random search.

[0097] S4, determining target operating parameters according to the multi-objective optimization function and the performance parameters corresponding to each set of operating parameters.

[0098] The process of step S4 is as follows: the performance parameters corresponding to each group of operating parameters are respectively substituted into the multi-objective optimization function to obtain the merit value corresponding to each performance parameter, and the target operating parameter is determined according to the multiple merit values. For example, the operating parameter corresponding to the maximum value among the multiple merit values ​​is used as the target operating parameter.

[0099] The expression of the multi-objective optimization function is as follows:

[0100]

[0101] Among them, a, b, and c are the first coefficient, the second coefficient, and the third coefficient respectively.

[0102] After obtaining the target operating parameters, the air-cooled fuel cell is operated with the target operating parameters. Therefore, the method uses the improved chaotic particle swarm algorithm to globally optimize the fan speed and the opening period of the hydrogen exhaust valve of the air-cooled fuel cell, and predicts the fuel cell performance based on the pre-trained neural network model, thereby obtaining the optimal operating conditions.

[0103] Furthermore, after the air-cooled fuel cell is operated according to the target operating parameters, the performance parameters of the air-cooled fuel cell are recorded; the training data set of the neural network performance prediction model is corrected according to the recorded performance parameters and the corresponding operating conditions. At the same time, the older data in the training data set can be deleted at regular intervals.

[0104] The present invention aims at the problem that the cooling and the supply of the reactant gas of the air-cooled fuel cell are fully coupled, and it is difficult to balance the hydrogen utilization rate and the drainage efficiency, which leads to the difficulty of thermal management. The present invention provides a multi-objective optimization thermal management method for an air-cooled fuel cell system based on a swarm intelligence algorithm. The method predicts the performance of the fuel cell according to the room temperature, output power, fan speed, hydrogen exhaust valve opening cycle and other conditions based on a pre-trained neural network performance prediction model. Subsequently, an improved particle swarm algorithm is used to perform a global combination optimization on the fan speed and the opening cycle of the hydrogen exhaust valve of the air-cooled fuel cell, so as to obtain the optimal operating conditions. At the same time, the training set of the neural network performance prediction model is corrected at regular intervals. It not only plays an important role in the optimization research of the thermal management method of the air-cooled fuel cell system, but also has practical significance in the actual production and application of air-cooled fuel cells.

[0105] In summary, according to the thermal management method of the air-cooled fuel cell of the embodiment of the present invention, when the air-cooled fuel cell is actually working, the initial operating parameters are determined by linear interpolation method according to the current ambient temperature and the target output power, and the operating parameters include the fan speed and the opening period of the hydrogen exhaust valve. Then, the initial operating parameters are used as the starting point, and the operating parameters are optimized using the improved chaotic particle swarm algorithm. In the process of optimizing the operating parameters, each time a new set of operating parameters is generated, the new operating parameters are input into the pre-trained neural network performance prediction model to predict the performance parameters of the air-cooled fuel cell corresponding to the current operating parameters. Finally, the target operating parameters are determined according to the multi-objective optimization function and the performance parameters corresponding to each set of operating parameters. Therefore, the method uses the improved chaotic particle swarm algorithm to perform global optimization on the fan speed and the opening period of the hydrogen exhaust valve of the air-cooled fuel cell, and predicts the performance of the fuel cell according to the pre-trained neural network model, so as to obtain the optimal operating conditions.

[0106] Figure 3 Schematic diagram of a thermal management device for an air-cooled fuel cell according to an embodiment of the present invention.

[0107] like Figure 3 As shown, the thermal management device 300 of an air-cooled fuel cell according to an embodiment of the present invention includes: a first determination module 310 , an optimization module 320 , a prediction module 330 and a second determination module 340 .

[0108] Among them, the first determination module 310 is used to determine the initial operating parameters using the linear interpolation method according to the current ambient temperature and the target output power when the air-cooled fuel cell is actually working; wherein the hydrogen operating parameters include the fan speed and the hydrogen exhaust valve opening period; the optimization module 320 is used to use the improved chaotic particle swarm algorithm to optimize the hydrogen operating parameters with the initial hydrogen operating parameters as the starting point; the prediction module 330 is used to input the new hydrogen operating parameters into the pre-trained neural network performance prediction model each time a new set of operating parameters is generated during the optimization process of the hydrogen operating parameters, so as to predict the performance parameters of the air-cooled fuel cell corresponding to the current operating parameters; the second determination module 340 is used to determine the target operating parameters according to the multi-objective optimization function and the performance parameters corresponding to each set of operating parameters.

[0109] In one embodiment of the present invention, the hydrogen performance parameter includes the total voltage V sum , Maximum single chip voltage V max , minimum single chip voltage V min and hydrogen utilization The second hydrogen determination module 340 is used to determine the target operating parameters according to the multi-objective optimization function and the performance parameters corresponding to each group of operating parameters, including:

[0110] Substitute the performance parameters corresponding to each set of operating parameters into the hydrogen multi-objective optimization function to obtain the quality value corresponding to each performance parameter;

[0111] Determine hydrogen target operating parameters based on multiple merits and demerits;

[0112] Among them, the expression of hydrogen multi-objective optimization function is:

[0113]

[0114] Among them, a, b, and c are the first coefficient, the second coefficient, and the third coefficient respectively.

[0115] In one embodiment of the present invention, the training process of the neural network performance prediction model in the hydrogen prediction module 330 includes:

[0116] Under different ambient temperatures and different output powers, multiple sets of different operating parameters are given and the corresponding performance parameters are calibrated to form a training data set for the hydrogen neural network performance prediction model;

[0117] Setting neural network parameters to obtain an initial neural network performance prediction model;

[0118] The initial neural network performance prediction model for hydrogen is trained using a hydrogen training data set to obtain a trained neural network performance prediction model.

[0119] In one embodiment of the present invention, the hydrogen device further comprises:

[0120] The correction module is used to record the performance parameters of the hydrogen air-cooled fuel cell after the hydrogen air-cooled fuel cell operates according to the hydrogen target operating parameters, and to correct the training data set of the hydrogen neural network performance prediction model according to the recorded performance parameters and the corresponding operating conditions.

[0121] It should be noted that, for details not disclosed in the importance analysis device based on characteristic parameters of a large break loss of coolant accident in an embodiment of the present invention, please refer to the details disclosed in the importance analysis method based on characteristic parameters of a large break loss of coolant accident in an embodiment of the present invention, and no further details will be given here.

[0122] According to the thermal management device of the air-cooled fuel cell of the embodiment of the present invention, the first determination module determines the initial operating parameters according to the current ambient temperature and the target output power by using the linear interpolation method when the air-cooled fuel cell is actually working, and the operating parameters include the fan speed and the opening period of the hydrogen exhaust valve. The optimization module uses the initial operating parameters as the starting point and uses the improved chaotic particle swarm algorithm to optimize the operating parameters. In the process of optimizing the operating parameters, the prediction module generates a new set of operating parameters each time, and inputs the new operating parameters into the pre-trained neural network performance prediction model to predict the performance parameters of the air-cooled fuel cell corresponding to the current operating parameters. The second determination module determines the target operating parameters according to the multi-objective optimization function and the performance parameters corresponding to each set of operating parameters. Thus, the device uses the improved chaotic particle swarm algorithm to perform global optimization on the fan speed and the opening period of the hydrogen exhaust valve of the air-cooled fuel cell, and predicts the fuel cell performance according to the pre-trained neural network model, so as to obtain the optimal operating conditions.

[0123] Based on the above embodiments, the present invention further provides an electronic device.

[0124] An electronic device according to an embodiment of the present invention comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above method.

[0125] The electronic device of the embodiment of the present invention executes the above method and uses the improved chaotic particle swarm algorithm to globally optimize the fan speed of the air-cooled fuel cell and the opening period of the hydrogen exhaust valve, and predicts the performance of the fuel cell according to the pre-trained neural network model, thereby obtaining the optimal operating conditions.

[0126] The computer instructions in the computer-readable storage medium of the embodiment of the present invention are used to enable a computer to execute the above method.

[0127] The computer-readable storage medium of an embodiment of the present invention executes the above method and uses an improved chaotic particle swarm algorithm to perform global optimization on the fan speed of the air-cooled fuel cell and the opening period of the hydrogen exhaust valve, and predicts the performance of the fuel cell according to a pre-trained neural network model, thereby obtaining the optimal operating conditions.

[0128] Based on the above embodiments, the present invention also proposes a computer program product.

[0129] In an embodiment of the present invention, when an instruction processor in a computer program product executes, the above method is performed.

[0130] The computer program product of the embodiment of the present invention executes the above method and uses an improved chaotic particle swarm algorithm to globally optimize the fan speed of the air-cooled fuel cell and the opening period of the hydrogen exhaust valve, and predicts the performance of the fuel cell based on a pre-trained neural network model, thereby obtaining the optimal operating conditions.

[0131] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0132] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0133] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present invention belong.

[0134] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0135] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0136] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0137] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0138] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present invention. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A thermal management method for an air-cooled fuel cell, characterized in that: include: When the air-cooled fuel cell is actually working, the initial operating parameters are determined by linear interpolation method according to the current ambient temperature and the target output power; wherein the operating parameters include the fan speed and the opening period of the hydrogen exhaust valve; Taking the initial operating parameters as a starting point, an improved chaotic particle swarm algorithm is used to optimize the operating parameters; During the operation parameter optimization process, each time a new set of operation parameters is generated, the new operation parameters are input into a pre-trained neural network performance prediction model to predict the performance parameters of the air-cooled fuel cell corresponding to the current operation parameters; The target operating parameters are determined according to the multi-objective optimization function and the performance parameters corresponding to each set of operating parameters.

2. The method according to claim 1, characterized in that The performance parameters include the total voltage V sum , Maximum single chip voltage V max , minimum single chip voltage V min and hydrogen utilization Determining the target operating parameters according to the multi-objective optimization function and the performance parameters corresponding to each set of operating parameters includes: Substituting the performance parameters corresponding to each set of operating parameters into the multi-objective optimization function to obtain the quality value corresponding to each performance parameter; Determining the target operating parameters according to a plurality of merit values; Wherein, the expression of the multi-objective optimization function is: Among them, a, b, and c are the first coefficient, the second coefficient, and the third coefficient respectively.

3. The method according to claim 1, characterized in that The training process of the neural network performance prediction model includes: Under different ambient temperatures and different output powers, a plurality of different operating parameters are given, and corresponding performance parameters are calibrated to form a training data set for the neural network performance prediction model; Setting neural network parameters to obtain an initial neural network performance prediction model; The initial neural network performance prediction model is trained using the training data set to obtain a trained neural network performance prediction model.

4. The method according to claim 3, characterized in that The method further comprises: After the air-cooled fuel cell operates according to the target operating parameters, recording the performance parameters of the air-cooled fuel cell; The training data set of the neural network performance prediction model is modified according to the recorded performance parameters and corresponding operating conditions.

5. A thermal management device for an air-cooled fuel cell, characterized in that: include: The first determination module is used to determine the initial operating parameters by linear interpolation method according to the current ambient temperature and the target output power when the air-cooled fuel cell is actually working; wherein the operating parameters include the fan speed and the opening period of the hydrogen exhaust valve; An optimization module, used to optimize the operating parameters using an improved chaotic particle swarm algorithm based on the initial operating parameters as a starting point; A prediction module, for inputting each new set of operating parameters generated during the operation parameter optimization process into a pre-trained neural network performance prediction model to predict the performance parameters of the air-cooled fuel cell corresponding to the current operating parameters; The second determination module is used to determine the target operating parameters according to the multi-objective optimization function and the performance parameters corresponding to each group of operating parameters.

6. The device according to claim 5, characterized in that The performance parameters include the total voltage V sum , Maximum single chip voltage V max , minimum single chip voltage V min and hydrogen utilization The second determination module is used to determine the target operating parameters according to the multi-objective optimization function and the performance parameters corresponding to each group of operating parameters, including: Substituting the performance parameters corresponding to each set of operating parameters into the multi-objective optimization function to obtain the quality value corresponding to each performance parameter; Determining the target operating parameters according to a plurality of merit values; Wherein, the expression of the multi-objective optimization function is: Among them, a, b, and c are the first coefficient, the second coefficient, and the third coefficient respectively.

7. The device according to claim 5, characterized in that The training process of the neural network performance prediction model in the prediction module includes: Under different ambient temperatures and different output powers, a plurality of different operating parameters are given, and corresponding performance parameters are calibrated to form a training data set for the neural network performance prediction model; Setting neural network parameters to obtain an initial neural network performance prediction model; The initial neural network performance prediction model is trained using the training data set to obtain a trained neural network performance prediction model.

8. The device according to claim 7, characterized in that The device also includes: A correction module is used to record the performance parameters of the air-cooled fuel cell after the air-cooled fuel cell operates according to the target operating parameters, and to correct the training data set of the neural network performance prediction model according to the recorded performance parameters and corresponding operating conditions.

9. An electronic device, characterized in that: include: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 4.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the program instructions are executed by a processor, the method according to any one of claims 1 to 4 is implemented.

Images