A method, apparatus, electronic device, and storage medium for quantitative analysis of influence degree

Abstract

The present application discloses a method, apparatus, electronic device, and storage medium for quantitative analysis of influence degree. The method and apparatus are applied to an electronic device, specifically for obtaining M-dimensional physical quantity data of a fuel cell system, where the M-dimensional physical quantity data includes M working conditions; constructing a neural network model; based on a step-by-step neural network model feature selection method, traversing the M-dimensional physical quantity data to train the neural network model; each time after the neural network model training is completed, recording the corresponding model training statistical error value; and calculating the influence degree of the corresponding working condition according to the model training statistical error value. Compared with the existing simulation methods, the solution of the present application can quantitatively analyze the influence degree of each working condition on the performance of the fuel cell system in the working state, thereby supporting the adaptation of working conditions in the fuel cell durability process and improving the system durability and efficiency.

Description

Technical Field

[0001] The present application relates to the technical field of fuel cells, and more specifically, to a quantitative analysis method, device, electronic device, and storage medium. Background Art

[0002] The durability of a fuel cell system is one of the key factors for the commercialization of fuel cells, and the adaptation of operating conditions has a significant impact on the durability of the system. Among various operating conditions, the influence degree of each operating condition on the fuel cell system is different. In order to determine the operating conditions that need to be prioritized for attention and adjustment and better adapt the operating conditions, it is necessary to quantitatively analyze the influence degree of various operating conditions on the operating performance of the fuel cell during its operation, so that designers can design and optimize the fuel cell system according to the influence degree.

[0003] Currently, mainstream simulation methods cannot quantitatively analyze the influence degree of the operating conditions of a fuel cell system during durable operation, and it is difficult to guide the adaptation of the operating conditions of a fuel cell that changes in real time. Therefore, for a fuel cell system in operation, there is an urgent need for a method for quantitatively analyzing the influence degree of operating conditions to support the adaptation of operating conditions during the durability process of the fuel cell and improve the durability and efficiency of the system. Summary of the Invention

[0004] In view of this, the present application provides a quantitative analysis method, device, electronic device, and storage medium for the influence degree, which is used to quantitatively analyze the influence degree of each operating condition on the performance of a fuel cell system in a working state, so as to support the adaptation of operating conditions during the durability process of the fuel cell and improve the durability and efficiency of the system.

[0005] In order to achieve the above object, the following solutions are proposed:

[0006] A quantitative analysis method for the influence degree, which is applied to an electronic device and is used to quantitatively analyze the influence degree of each operating condition on the operating performance of a fuel cell system. The quantitative analysis method includes the steps of:

[0007] Obtain M-dimensional physical quantity data of the fuel cell system, where the M-dimensional physical quantity data includes M operating conditions, and M is a positive integer;

[0008] Construct a neural network model, where the input data dimension of the neural network model is N, the number of neurons in the input layer is N, and the number of neurons in the output layer is 1, and N is a positive integer less than M;

[0009] Based on the step-by-step neural network model feature selection method, traverse the M-dimensional physical quantity data and train the neural network model;

[0010] After each completion of the neural network model training, record the corresponding model training statistical error value;

[0011] Calculate the influence degree of the corresponding working condition according to the model training statistical error value.

[0012] Optionally, the multiple working conditions include some or all of the running time, air flow rate, air inlet pressure, hydrogen inlet pressure, air inlet temperature, hydrogen inlet temperature, coolant outlet temperature, coolant inlet temperature, and their derived variables.

[0013] Optionally, the step-by-step neural network model feature selection method traverses the M-dimensional physical quantity data and trains the neural network model, including the following methods:

[0014] Based on the backward method, select (N - 1) working conditions from the N working conditions as the model input, and use the performance index of the fuel cell system as the model output to train the neural network model;

[0015] Alternatively, based on the forward method, select 1 working condition from the M - N - 2 co-group conditions, use N + 1 working conditions as the model input, and use the performance index as the model output to train the neural network model;

[0016] Alternatively, based on the two-way method, select N + n working conditions as the model input, and use the performance index as the model output to train the neural network model, where n satisfies N > 1 - n and N < M - n - 2.

[0017] Optionally, the calculating the influence degree of the corresponding working condition according to the model training statistical error value includes the steps:

[0018] When n < 0, define the sensitivity as the model training statistical error value;

[0019] When n > 0, define the sensitivity as the reciprocal of the model training statistical error value.

[0020] An influence degree quantitative analysis device is applied to an electronic device and is used to quantitatively analyze the influence degree of each working condition on the working performance of the fuel cell system. The quantitative analysis device includes:

[0021] A data acquisition module configured to acquire the M-dimensional physical quantity data of the fuel cell system. The M-dimensional physical quantity data includes M working conditions, and M is a positive integer;

[0022] A model construction module, configured to construct a neural network model, where the input data dimension of the neural network model is N, the number of neurons in the input layer is N, the number of neurons in the output layer is 1, and N is a positive integer less than M;

[0023] A model training module, configured to traverse the M-dimensional physical quantity data based on a step-by-step neural network model feature selection method, and train the neural network model;

[0024] An error recording module, configured to record the corresponding model training statistical error value each time the neural network model training is completed;

[0025] A quantization calculation module, configured to calculate the influence degree of the corresponding working condition according to the model training statistical error value.

[0026] Optionally, the multiple working conditions include some or all of the running time, air flow rate, air inlet pressure, hydrogen inlet pressure, air inlet temperature, hydrogen inlet temperature, coolant outlet temperature, coolant inlet temperature, and their derived variables.

[0027] Optionally, the model training module includes:

[0028] A first training unit, configured to select (N - 1) working conditions from the N working conditions as the model input based on the backward method, and use the performance index of the fuel cell system as the model output to train the neural network model;

[0029] Or, a second training unit, configured to select 1 working condition from the M - N - 2 working conditions based on the forward method, use N + 1 working conditions as the model input, and use the performance index as the model output to train the neural network model;

[0030] Or a third training unit, configured to select N + n working conditions as the model input based on the bidirectional method, and use the performance index as the model output to train the neural network model, where n satisfies N > 1 - n and N < M - n - 2.

[0031] Optionally, the quantization calculation module includes:

[0032] A first quantization unit, configured to define the sensitivity as the model training statistical error value when n < 0;

[0033] A second quantization unit, configured to define the sensitivity as the reciprocal of the model training statistical error value when n > 0.

[0034] An electronic device, including at least one processor and a memory connected to the processor, where:

[0035] The memory is used to store computer programs or instructions;

[0036] The processor is used to execute the computer programs or instructions, so that the electronic device implements the quantization analysis method of the influence degree as described above.

[0037] A storage medium is applied to an electronic device. The storage medium carries one or more computer programs. When the electronic device executes the one or more computer programs, the quantization analysis method of the influence degree as described above can be realized.

[0038] As can be seen from the above technical solutions, this application discloses a quantization analysis method, device, electronic device and storage medium of an influence degree. The method and device are applied to an electronic device. Specifically, M-dimensional physical quantity data of a fuel cell system is obtained. The M-dimensional physical quantity data includes M working conditions; a neural network model is constructed; based on the step-by-step neural network model feature selection method, the M-dimensional physical quantity data is traversed to train the neural network model; after each neural network model training is completed, the corresponding model training statistical error value is recorded; the influence degree of the corresponding working condition is calculated according to the model training statistical error value. Compared with the existing mainstream simulation methods, the solution of this application can quantitatively analyze the influence degree of each working condition on the performance of the fuel cell system in the working state, so as to support the working condition adaptation of the fuel cell durability process and improve the system durability and efficiency. Brief Description of the Drawings

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required to be used in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0040] Figure 1 It is a flowchart of a quantization analysis method of an influence degree according to an embodiment of the present application;

[0041] Figure 2 It is a schematic diagram of a multi-layer perceptron neural network according to an embodiment of the present application;

[0042] Figure 3 It is a bar chart of the influence degree of neural network training completed according to the forward selection principle in an embodiment of the present application;

[0043] Figure 4 It is a change diagram of the influence degree of neural network training completed according to the forward selection principle in an embodiment of the present application;

[0044] Figure 5Block diagram of a quantization analysis device for the influence degree according to an embodiment of the present application;

[0045] Figure 6 Block diagram of an electronic device according to an embodiment of the present application. Detailed implementation manners

[0046] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0047] The neural network model can extract the relationship between the input data and the output data. If the influence degree of a certain fuel cell working condition is high, more information about the change of performance indicators is contained in its feedback signal; conversely, less information about the change of performance indicators is contained in this working condition. According to information theory, the information entropy that measures the uncertainty of event X is defined as:

[0048]

[0049] The probability p(x) of event X = x ∈ [0, 1]. Therefore, the information entropy H(X) is non - negative. Assuming that after obtaining the information, the probabilities of all events can be set to 1, then the amount of information is equal to the information entropy. At the same time, this also means that in this case, the uncertainty brought by the information is eliminated. If the input data contains more information about the change of performance indicators, then, on the premise of making full use of the information, the uncertainty of judging the output data is smaller, and the model training error is smaller; conversely, the model training error is larger. Based on the above principles, the principles of using the backward method and the forward method to judge the influence degree of fuel cell working conditions can be analyzed as follows:

[0050] For the backward method, the larger the statistical error, the greater the influence of the removed variable (working condition) on the model training, that is, this variable contains the information determining the performance indicators that other variables do not have, and its influence degree is higher; for the forward method, the smaller the mean square error, the greater the influence of the selected variable on the model training, that is, this variable contains the most important information determining the performance indicators, and its influence degree is higher. The logics of the two methods are not completely the same, and are not limited to the increase or decrease of single variables. The effectiveness of the results of the two methods' logics should depend on the relative size of the statistical errors.

[0051] The above is the inventive principle of the present application. Based on the above description, the following specific embodiments are particularly proposed.

[0052] Embodiment 1

[0053] Figure 1Flow chart of a method for quantitatively analyzing influence degree according to an embodiment of the present application.

[0054] As Figure 1 shown, the quantitative analysis method provided in this embodiment is applied to an electronic device, which can be understood as a computer or a server with information processing capabilities and data calculation capabilities, and is used to quantitatively analyze and calculate the influence degree of each working condition on the working performance of the fuel cell system, so as to support the adaptation of working conditions during the fuel cell durability process and improve the system durability and efficiency. The quantitative analysis method includes the following steps:

[0055] S1. Obtain M-dimensional physical quantity data.

[0056] Generally, the above M-dimensional physical quantity data can be obtained through a test bench of the fuel cell system. Here, M is an integer greater than zero, and the M-dimensional physical quantity data actually includes M - 2 working conditions of the fuel cell system. The working conditions in this embodiment include operating time, working voltage, working current, air flow rate, air inlet pressure, hydrogen inlet pressure, air inlet temperature, hydrogen inlet temperature, coolant outlet temperature, and coolant inlet temperature and their derived variables. Here, M is 10. The derived variables include, but are not limited to, temperature difference, flow resistance, stoichiometric ratio, etc.

[0057] After obtaining the above working conditions, store the working conditions in the form of an array in chronological order and in text format in the memory of the electronic device for subsequent further processing.

[0058] S2. Build a neural network model.

[0059] In this application, the input data dimension N of the built neural network model satisfies the condition that N < M - 2. The number of neurons in the input layer is N, and the number of neurons in the output layer is 1. Preferably, the neural network model is a multi-layer perceptron neural network, a recurrent neural network, an autoencoder, or a convolutional neural network.

[0060] Specifically in this embodiment, the number of neurons in the input layer of the multi-layer perceptron neural network is 1, the number of hidden layers is 3, the number of neurons in each layer is 10, and the number of neurons in the output layer is 1, as Figure 2 shown.

[0061] S3. Train the neural network model based on a step-by-step neural network model adjustment selection method.

[0062] Based on the step-by-step neural network model feature selection method, traverse the aforementioned M-dimensional physical quantity data and train the neural network model. The specific solution is as follows:

[0063] Based on the backward method, after removing M-N variables such as the operating time and performance indicators of the fuel cell system, N-1 working conditions are selected from the remaining N working conditions as the input of the neural network model, and the performance indicator is used as the model output to train the neural network. The performance indicators here include but are not limited to the working voltage, the power of the fuel cell system, the variance of the fuel cell inspection voltage, etc.

[0064] Based on the forward method, after removing M-N working conditions such as the operating time and performance indicators of the fuel cell, one is selected from the removed M-N-2 working conditions, and N+1 working conditions are used as the input of the neural network model, and the performance indicator is used as the model output to train the model.

[0065] Based on the two-way method, N+n-dimensional working conditions are selected as the input of the neural network model (satisfying N>1-n and N<M-n-2), and the performance indicator is used as the model output to train the neural network. Here, n satisfies N>1-n and N<M-n-2. It can be seen from the formula that when n = 1, the two-way method degenerates into the forward method; when n = -1, the two-way method degenerates into the backward method.

[0066] S4. Record the model training statistical error value for each round of model training.

[0067] Each time the neural network training is completed, the corresponding model training statistical error is recorded. Generally, the model training statistical error is the mean square error, and the mean square error is defined as follows:

[0068]

[0069] where, {x i} is the true value sequence, the model training predicted value sequence. The model training statistical error can also select the Euclidean distance or Mahalanobis distance between the model training predicted value and the true value.

[0070] S5. Calculate the influence degree of the corresponding working condition according to the model training statistical error value.

[0071] Specifically, examine the model training statistical error values of the neural network training under different situations, and determine the influence degree of each working condition on the working performance of the fuel cell system in the operating state according to the model training statistical error value, and sort the influence degrees.

[0072] Both the aforementioned backward method and forward method are special cases of the two-way method. Based on this, when n<0, the influence degree of the working condition is defined as the aforementioned model training statistical error value; when n>0, the influence degree of the working condition is defined as the reciprocal of the model training statistical error value of the neural network model. According to the above definitions, the influence degree values of the corresponding working conditions can be calculated and sorted, so as to provide a basis for the working condition adaptation in the whole process of fuel cell operation.

[0073] As can be seen from the above technical solution, this embodiment provides a method for quantitative analysis of influence degree, which is applied to an electronic device. Specifically, it obtains M-dimensional physical quantity data of a fuel cell system, and the M-dimensional physical quantity data includes M working conditions; constructs a neural network model; based on the step-by-step neural network model feature selection method, traverses the M-dimensional physical quantity data to train the neural network model; after each training of the neural network model is completed, records the corresponding model training statistical error value; calculates the influence degree of the corresponding working condition according to the model training statistical error value. Compared with the existing simulation methods, the solution of this application can quantitatively analyze the influence degree of each working condition on the performance of the fuel cell system in the working state, so as to support the adaptation of working conditions in the fuel cell durability process and improve the system durability and efficiency.

[0074] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0075] Although the operations are depicted in a particular order, this should not be construed as requiring that the operations be performed in the particular order shown or in sequential order. In certain circumstances, multitasking and parallel processing may be advantageous.

[0076] It should be understood that the various steps recited in the method embodiments of the present disclosure may be executed in a different order and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this regard.

[0077] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages or combinations thereof. The above-mentioned programming languages include, but are not limited to, object-oriented programming languages - such as Python, Julia, Java, Smalltalk, C++, and also include conventional procedural programming languages - such as the C language or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network - including a local area network (LAN) or a wide area network (WAN) - or may be connected to an external computer.

[0078] Embodiment 2

[0079] Figure 5 It is a block diagram of a device for quantitatively analyzing the influence degree of an embodiment of the present application.

[0080] As Figure 5 As shown, the quantitative analysis device provided in this embodiment is applied to an electronic device. The electronic device can be understood as a computer or server with information processing capabilities and data calculation capabilities, and is used to quantitatively analyze and calculate the influence degree of each working condition on the working performance of the fuel cell system, so as to support the adaptation of the working conditions in the fuel cell durability process and improve the system durability and efficiency. The quantitative analysis device includes a data acquisition module 10, a model construction module 20, a model training module 30, an error recording module 40, and a quantitative calculation module 50.

[0081] The data acquisition module is used to acquire M-dimensional physical quantity data.

[0082] Generally, the above-mentioned M-dimensional physical quantity data can be obtained through a test bench of the fuel cell system. Here, M is a positive integer, and the M-dimensional physical quantity data actually includes M - 2 working conditions of the fuel cell system. The working conditions in this embodiment include operating time, working voltage, working current, air flow rate, air inlet pressure, hydrogen inlet pressure, air inlet temperature, hydrogen inlet temperature, coolant outlet temperature, and coolant inlet temperature and their derived variables. Here, M is 10. The derived variables include, but are not limited to, temperature difference, flow resistance, stoichiometric ratio, etc.

[0083] After obtaining the above-mentioned working conditions, the working conditions are stored in the memory of the electronic device in the form of an array in chronological order and in text format for subsequent further processing.

[0084] The model construction module is used to construct a neural network model.

[0085] The dimensionality N of the input data of the neural network model constructed in this application, where N satisfies the condition N < M - 2, the number of neurons in the input layer is N, and the number of neurons in the output layer is 1. Preferably, the neural network model is a multi-layer perceptron neural network, a recurrent neural network, an autoencoder, or a convolutional neural network.

[0086] Specifically in this embodiment, the number of neurons in the input layer of this multi-layer perceptron neural network is 1, the number of hidden layers is 3, the number of neurons in each layer is 10, and the number of neurons in the output layer is 1, as Figure 2 shown.

[0087] The model training module is used to train the neural network model based on the step-by-step neural network model adjustment selection method.

[0088] Based on the step-by-step neural network model feature selection method, traverse the aforementioned M-dimensional physical quantity data to train the neural network model. This module includes a first training unit, a second training unit, and a third training unit.

[0089] The first training unit is used to select N - 1 working conditions from the remaining N working conditions as the input of the neural network model and the performance index as the model output for model training of the neural network based on the backward method after removing M - N variables such as the operating time and performance index of the fuel cell system. The performance index here includes but is not limited to the working voltage, the power of the fuel cell system, the variance of the fuel cell inspection voltage, etc.

[0090] The second training unit is used to select 1 from the removed M - N - 2 working conditions based on the forward method after removing M - N working conditions such as the operating time and performance index of the fuel cell, and use N + 1 working conditions as the input of the neural network model and the performance index as the model output for model training.

[0091] The third training unit is used to select N + n - dimensional working conditions as the input of the neural network model (satisfying N > 1 - n and N < M - n - 2) and use the performance index as the model output to train the neural network. Here, n satisfies N > 1 - n and N < M - n - 2. It can be seen from the formula that when n = 1, the two-way method degenerates into the forward method; when n = -1, the two-way method degenerates into the backward method.

[0092] The error recording module is used to record the model training statistical error value of each round of model training.

[0093] Each time the neural network training is completed, the corresponding model training statistical error is recorded. Generally, the model training statistical error is the mean square error, and the mean square error is as follows:

[0094]

[0095] Among them, {x i} is the true value sequence, and the model training predicted value sequence. The model training statistical error can also select the Euclidean distance or the Mahalanobis distance between the model training predicted value and the true value.

[0096] The quantization calculation module is used to calculate the influence degree of the corresponding working condition according to the model training statistical error value.

[0097] Specifically, the model training statistical error values of neural network training in different situations are investigated, and according to the model training statistical error values, the influence degrees of each working condition on the working performance of the fuel cell system in the operating state are determined, and the influence degrees are sorted.

[0098] The aforementioned backward method and forward method are both special cases of the two-way method. This module includes a first quantization unit and a second quantization unit. The first quantization unit is used to define the influence degree of the working condition as the aforementioned model training statistical error value when n < 0; the second quantization unit is used to define the influence degree of the working condition as the reciprocal of the model training statistical error value of the neural network model when n > 0. According to the above definitions, the influence degree values of the corresponding working conditions can be calculated and sorted, so as to provide a basis for the working condition adaptation in the whole process of fuel cell operation.

[0099] The units involved in the embodiments described in this disclosure can be implemented in software or in hardware. Among them, the name of the unit does not constitute a limitation to the unit itself in some cases. For example, the first acquisition unit can also be described as "the unit for acquiring at least two Internet protocol addresses".

[0100] The functions described above in this article can be performed at least in part by one or more hardware logic components. For example, without limitation, the exemplary types of hardware logic components that can be used include: Field Programmable Gate Array (FPGA), Application Specific Integrated Circuit (ASIC), Application Specific Standard Product (ASSP), System on Chip (SOC), Complex Programmable Logic Device (CPLD), and so on.

[0101] Embodiment III

[0102] Figure 6 It is a block diagram of an electronic device according to an embodiment of the present application.

[0103] Refer to Figure 6As shown, it shows a schematic structural diagram of an electronic device suitable for implementing the electronic device in the embodiments of the present disclosure. The terminal device in the embodiments of the present disclosure may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Tablet Computers), PMPs (Portable Multimedia Players), vehicle terminals (such as vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 6 The electronic device shown is merely an example and should not impose any limitation on the functions and usage scope of the embodiments of the present disclosure.

[0104] As Figure 6 shown, the electronic device may include a processing device (such as a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 602 or the program loaded from the storage device 606 into the random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device are also stored. The processing device 601, the ROM 602, and the RAM 603 are connected to each other through a bus 604. The input / output (I / O) interface 605 is also connected to the bus 604.

[0105] Generally, the following devices may be connected to the I / O interface 605: an input device 606 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 606 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 609. The communication device 609 can allow the electronic device to communicate with other devices wirelessly or wiredly to exchange data. Although Figure 6 the electronic device with various devices is shown, it should be understood that it is not required to implement or have all the shown devices. More or fewer devices may be alternatively implemented or had.

[0106] Embodiment Four

[0107] This embodiment provides a computer-readable storage medium, which is applied to an electronic device. The above computer-readable medium carries one or more programs. When the above one or more programs are executed by the electronic device, the electronic device can implement the quantitative analysis method of the influence degree in Embodiment One.

[0108] It should be noted that the computer-readable medium described above in the present disclosure may be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium that contains or stores a program, which can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0109] Each embodiment in this specification is described in a progressive manner. The key point of each embodiment is to illustrate the differences from other embodiments. For the same or similar parts among the embodiments, reference may be made to each other.

[0110] Although the preferred embodiments of the embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications once they know the basic creative concept. Therefore, the appended claims are intended to be construed as including the preferred embodiments and all changes and modifications falling within the scope of the embodiments of the present invention.

[0111] Finally, it should also be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or terminal device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or terminal device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or terminal device comprising the element.

[0112] The technical solutions provided by the present invention have been introduced in detail above. Specific examples are used in this text to elaborate on the principles and implementation manners of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A method for quantitative analysis of influence degree, which is applied to an electronic device and used for quantitatively analyzing the influence degree of each working condition on the working performance of a fuel cell system, is characterized in that The quantization analysis method includes: Obtaining M-dimensional physical quantity data of the fuel cell system, where the M-dimensional physical quantity data includes M-2 working conditions, and M is a positive integer; Constructing a neural network model, where the input data dimension of the neural network model is N, the number of neurons in the input layer is N, and the number of neurons in the output layer is 1, and N is a positive integer less than M-2; Based on the step-by-step neural network model feature selection method, traverse the M-dimensional physical quantity data and train the neural network model, specifically including: Based on the backward method, select N-1 working conditions from the N working conditions as the model input, and use the performance index of the fuel cell system as the model output to train the neural network model; Alternatively, based on the forward method, select 1 working condition from the M-N-2 working conditions, use N+1 working conditions as the model input, and use the performance index as the model output to train the neural network model; Alternatively, based on the bidirectional method, select N+n working conditions as the model input, and use the performance index as the model output to train the neural network model, where n satisfies N>1-n and N<M-n-2; Among them, when n = 1, the bidirectional method degenerates into the forward method; when n = -1, the bidirectional method degenerates into the backward method; After each neural network model training is completed, record the corresponding model training statistical error value; Calculate the influence degree of the corresponding working condition on the working performance of the fuel cell system in the working state according to the model training statistical error value, specifically including: when n<0, define the influence degree as the model training statistical error value; when n>0, define the influence degree as the reciprocal of the model training statistical error value.

2. The quantitative analysis method according to claim 1, characterized in that The working conditions include some or all of the running time, air flow rate, air inlet pressure, hydrogen inlet pressure, air inlet temperature, hydrogen inlet temperature, coolant outlet temperature, coolant inlet temperature and their derived variables.

3. A quantization analysis device for influence degree, which is applied to an electronic device and is used to quantitatively analyze the influence degree of each working condition on the working performance of a fuel cell system, is characterized in that The quantization analysis device includes: A data acquisition module configured to acquire M-dimensional physical quantity data of the fuel cell system, where the M-dimensional physical quantity data includes M-2 working conditions, and M is a positive integer; A model construction module configured to construct a neural network model, where the input data dimension of the neural network model is N, the number of neurons in the input layer is N, and the number of neurons in the output layer is 1, and N is a positive integer less than M-2; A model training module configured to traverse the M-dimensional physical quantity data and train the neural network model based on the step-by-step neural network model feature selection method; An error recording module configured to record the corresponding model training statistical error value after each neural network model training is completed; A quantization calculation module configured to calculate the influence degree of the corresponding working condition on the working performance of the fuel cell system in the working state according to the model training statistical error value; The model training module includes: The first training unit is configured to select N - 1 working conditions from the N working conditions as model inputs based on the backward method, and use the performance index of the fuel cell system as the model output to train the neural network model; Alternatively, a second training unit is configured to select 1 working condition from the M - N - 2 working conditions based on the forward method, use N + 1 working conditions as model inputs, and use the performance index as the model output to train the neural network model; Alternatively, a third training unit is configured to select N + n working conditions as model inputs based on the bidirectional method, and use the performance index as the model output to train the neural network model, where n satisfies N > 1 - n and N < M - n - 2; Wherein, when n = 1, the bidirectional method degenerates into the forward method; when n = -1, the bidirectional method degenerates into the backward method; The quantization calculation module includes: A first quantization unit configured to define the influence degree as the model training statistical error value when n < 0; A second quantization unit configured to define the influence degree as the reciprocal of the model training statistical error value when n > 0.

4. The quantization analysis device according to claim 3, wherein The working conditions include some or all of the running time, air flow rate, air inlet pressure, hydrogen inlet pressure, air inlet temperature, hydrogen inlet temperature, coolant outlet temperature, coolant inlet temperature, and their derivative variables.

5. An electronic device, characterized in that, A memory including at least one processor and connected to the processor, wherein: The memory is used to store computer programs or instructions; The processor is used to execute the computer programs or instructions so that the electronic device implements the quantization analysis method of the influence degree as described in any one of claims 1 to 2.

6. A storage medium, applied to an electronic device, characterized in that, The storage medium carries one or more computer programs, and when the electronic device executes the one or more computer programs, it can implement the quantization analysis method of the influence degree as described in any one of claims 1 to 2.

Images