Fuel cell stacking method and device, electronic equipment and storage medium

Abstract

The invention provides a fuel cell stacking method and device, electronic equipment and a storage medium, the method is applied to a fuel cell, and the method comprises the following steps: acquiring material thickness data of the fuel cell; according to the material thickness data, a permutation and combination result meeting the material stacking requirement is predicted, and a target material taking sequence from the multiple material bins is determined based on the permutation and combination result; and the materials are taken from the corresponding material bins according to the target material taking sequence and stacked repeatedly. According to the method, the material thickness data of the fuel cell can be analyzed and processed, the permutation and combination result meeting the material stacking requirement is predicted according to the thickness data, and intelligent material taking and stacking of the materials are achieved, so that stacking at a fixed height is achieved, the stacking height of a reactor core is consistent, and the product quality is improved.

Description

Technical Field

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

[0002] At present, in the production and manufacturing process of hydrogen fuel cells, when the bipolar plates and membrane electrodes are being prepared, the materials may come from different production batches, and the processing errors will cause a certain dimensional tolerance in the thickness of the materials. Since the number of materials may be as high as 1,000 or more when the core is repeatedly stacked, there will be a large cumulative error in the size of the entire stack, and the consistency of the core height will be difficult to ensure. For the subsequent press-fitting process, if the core height is uneven, the internal stress of each stack after press-fitting cannot be guaranteed to be consistent, and the product quality cannot be guaranteed.

[0003] The problem of inconsistent core stacking height is generally solved by adjusting the pressing force to meet the technical requirements of the battery stack packaging process. However, this method will lead to the following problems: due to different stacking heights, the pressing force selection is different, resulting in different stresses in the core; the electrochemical reaction zones inside different battery stack cores are subjected to different forces, and the consistency of product quality cannot be guaranteed; the force in the reaction zone will affect the contact area of ​​the reaction zone, thereby affecting the reaction effect and power performance of each single battery. Summary of the invention

[0004] The present application provides a fuel cell stacking method, device, electronic device and storage medium. The method can analyze and process the material thickness data of the fuel cell, predict the permutation and combination results that meet the material stacking requirements based on the thickness data, realize intelligent material extraction and stacking, thereby achieving fixed height stacking, making the core stacking height consistent, and improving product quality.

[0005] In a first aspect, a fuel cell stacking method is provided, the method comprising: acquiring material thickness data of the fuel cell; predicting permutation and combination results that meet material stacking requirements based on the material thickness data, and determining a target material collection sequence from multiple material bins based on the permutation and combination results; and collecting materials from corresponding material bins according to the target material collection sequence for repeated stacking.

[0006] Through the above technical scheme, the material thickness data of the fuel cell can be analyzed and processed, and the arrangement and combination results that meet the material stacking requirements can be predicted based on the thickness data, so as to realize intelligent material retrieval and stacking, thereby achieving fixed height stacking, making the core stacking height consistent, and improving product quality.

[0007] In combination with the first aspect, in some possible implementations, the material stacking requirements include a target stacking height and a target quantity, and predicting the permutation and combination results that meet the material stacking requirements based on the material thickness data includes: inputting the target height, the target quantity and the material thickness data into a pre-trained intelligent sorting model, and the intelligent sorting model outputs the permutation and combination results of the material stacking.

[0008] Through the above technical solution, the arrangement and combination results of material stacking that meet the target height, target quantity and material thickness can be output through the intelligent sorting model.

[0009] In combination with the first aspect and the above-mentioned implementation methods, in some possible implementation methods, the sorting process of the intelligent sorting model includes: identifying all material thicknesses in the material thickness data; defining a state list and a state transition probability list of the material, wherein the state list includes a first state indicating that the material is in a material bin and a second state indicating that the material is in a stacking device, and the state transition probability list includes the probability of the material transferring between the first state and the second state; selecting membrane electrodes and bipolar plates as materials from the material bin or the stacking device according to the target height, the target quantity, the state list and the state transition probability list, and obtaining the permutation and combination results of the material stacking.

[0010] Through the above technical scheme, the state list and state transition probability list of the material can be defined, and the membrane electrode and bipolar plate can be selected from the material bin or stacking equipment according to the target height, target quantity, state list and state transition probability list to obtain the permutation and combination results of the material stacking, and realize automatic calculation and analysis of the permutation and combination results of the material stacking that meet the design requirements.

[0011] In combination with the first aspect and the above-mentioned implementation methods, in some possible implementation methods, the membrane electrode and bipolar plate are selected from the material bin or the stacking device according to the target height, the target quantity, the state list and the state transition probability list, including: if the current state is the first state, the membrane electrode and / or bipolar plate are selected from the material bin based on the state transition probability list, and the material corresponding to the state list is updated from the first state to the second state; if the current state is the second state, the membrane electrode and / or bipolar plate are selected from the stacking device based on the state transition probability list, and the state of the material corresponding to the state list is updated according to the change in the stacking height; after the state list is updated, the state transition probability of the material in the state transition probability list is updated according to the selected material thickness, the target height and the target quantity, until the stacking height of the material in the second state reaches the target height and the quantity reaches the target quantity.

[0012] Through the above technical scheme, it is possible to determine how to select membrane electrodes and bipolar plates based on the current state and the state transition probability list so that the material stacking height and quantity meet the design requirements. After the material is selected, the current state of the material and the state transition probability are updated, thereby realizing the update of the material state for the subsequent selection of stacked materials.

[0013] In combination with the first aspect and the above-mentioned implementation methods, in some possible implementation methods, after defining the state list and state transition probability list of the material, it also includes: initializing the state list and the state transition probability list, wherein the values ​​of all elements in the initialized state list are the same, and the state transition probabilities of all materials in the initialized state transition probability list are the same.

[0014] Through the above technical solution, the state list and the state transition probability list can be initialized to facilitate the subsequent intelligent sorting of materials.

[0015] In combination with the first aspect and the above-mentioned implementation methods, in some possible implementation methods, before inputting the target height, the target quantity and the material thickness data into the pre-trained intelligent sorting model, it also includes: obtaining the stacking data of the membrane electrode and the bipolar plate in the fuel cell; generating a training data set and a verification set based on the stacking data; selecting a machine learning or deep learning model as the intelligent sorting model, training the intelligent sorting model according to the training data set, and verifying the trained intelligent sorting model according to the verification set, and completing the training when the performance and accuracy of the intelligent sorting model meet the preset requirements.

[0016] Through the above technical solution, a large amount of data can be used to train and verify the intelligent sorting model until the performance and accuracy of the intelligent sorting model meet the preset requirements. The training is completed, thereby improving the accuracy of the intelligent sorting model and the accuracy of the prediction, and better assisting production.

[0017] In combination with the first aspect and the above-mentioned implementation methods, in some possible implementation methods, the material includes membrane electrodes and bipolar plates, and the material is taken from the corresponding material bin according to the target material taking order and repeatedly stacked, including: taking materials from the corresponding material bin according to the target material taking order; repeatedly and alternately stacking the taken out membrane electrodes and bipolar plates until the stacking quantity of the membrane electrodes and the bipolar plates reaches the target quantity.

[0018] Through the above technical solution, materials can be taken from the corresponding material bin in the target material taking order, and the membrane electrode and bipolar plates can be repeatedly stacked alternately until the target quantity is met and the stacking is stopped, thereby realizing the automatic stacking of fuel cell cores.

[0019] In a second aspect, a fuel cell stacking device is provided, which includes: an acquisition module for acquiring material thickness data of the fuel cell; a determination module for predicting permutation and combination results that meet material stacking requirements based on the material thickness data, and determining a target material collection sequence from multiple material bins based on the permutation and combination results; and a stacking module for collecting materials from corresponding material bins for repeated stacking according to the target material collection sequence.

[0020] In combination with the second aspect, in some possible implementations, the material stacking requirements include a target stacking height and a target quantity, and the determination module is further used to: input the target height, the target quantity and the material thickness data into a pre-trained intelligent sorting model, and the intelligent sorting model outputs the permutation and combination results of the material stacking.

[0021] In combination with the second aspect and the above-mentioned implementation methods, in some possible implementation methods, the sorting process of the intelligent sorting model includes: identifying all material thicknesses in the material thickness data; defining a state list and a state transition probability list of the material, wherein the state list includes a first state indicating that the material is in a material bin and a second state indicating that the material is in a stacking device, and the state transition probability list includes the probability of the material transferring between the first state and the second state; selecting membrane electrodes and bipolar plates as materials from the material bin or the stacking device according to the target height, the target quantity, the state list and the state transition probability list, and obtaining the permutation and combination results of the material stacking.

[0022] In combination with the second aspect and the above-mentioned implementation methods, in some possible implementation methods, the membrane electrode and bipolar plate are selected from the material bin or the stacking device according to the target height, the target quantity, the state list and the state transition probability list, including: if the current state is the first state, the membrane electrode and / or bipolar plate are selected from the material bin based on the state transition probability list, and the corresponding material in the state list is updated from the first state to the second state; if the current state is the second state, the membrane electrode and / or bipolar plate are selected from the stacking device based on the state transition probability list, and the state of the material corresponding to the state list is updated according to the change in the stacking height; after the state list is updated, the state transition probability of the material in the state transition probability list is updated according to the selected material thickness, the target height and the target quantity, until the stacking height of the material in the second state reaches the target height and the quantity reaches the target quantity.

[0023] In combination with the second aspect and the above-mentioned implementation methods, in some possible implementation methods, after defining the state list and state transition probability list of the material, it also includes: initializing the state list and the state transition probability list, wherein the values ​​of all elements in the state list after initialization are the same, and the state transition probabilities of all materials in the state transition probability list after initialization are the same.

[0024] In combination with the second aspect and the above-mentioned implementation methods, in some possible implementation methods, before inputting the target height, the target quantity and the material thickness data into a pre-trained intelligent sorting model, it also includes: obtaining the stacking data of the membrane electrode and the bipolar plate in the fuel cell; generating a training data set and a verification set based on the stacking data; selecting a machine learning or deep learning model as the intelligent sorting model, training the intelligent sorting model according to the training data set, and verifying the trained intelligent sorting model according to the verification set, and completing the training when the performance and accuracy of the intelligent sorting model meet the preset requirements.

[0025] In combination with the second aspect and the above-mentioned implementation methods, in some possible implementation methods, the material includes membrane electrodes and bipolar plates, and the stacking module is further used to: take materials from the corresponding material bin according to the target material taking order; repeatedly and alternately stack the taken out membrane electrodes and bipolar plates until the stacking quantity of the membrane electrodes and the bipolar plates reaches the target quantity.

[0026] In a third aspect, an electronic device is provided, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement a fuel cell stacking method as described in the first aspect or any possible implementation of the first aspect.

[0027] In a fourth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed, a fuel cell stacking method as described in the first aspect or any possible implementation of the first aspect is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 is a structural diagram of a thickness measuring device provided in an embodiment of the present application;

[0029] Figure 2 is a flow chart of a fuel cell stacking method provided in an embodiment of the present application;

[0030] Figure 3 is a structural diagram of a stacking device provided in an embodiment of the present application;

[0031] Figure 4is an architecture diagram of a fuel cell stack system provided in an embodiment of the present application;

[0032] Figure 5 is a schematic structural diagram of a fuel cell stacking device provided in an embodiment of the present application;

[0033] Figure 6 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0034] The technical solution in the present application will be described clearly and in detail below in conjunction with the accompanying drawings. In the description of the embodiments of the present application, unless otherwise specified, " / " means or, for example, A / B can mean A or B: "and / or" in the text is only a description of the association relationship of associated objects, indicating that there can be three relationships, for example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, in the description of the embodiments of the present application, "multiple" means two or more than two.

[0035] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features.

[0036] Hydrogen fuel cells use perfluorosulfonic acid solid polymers as electrolytes, Pt / C as catalysts, hydrogen or reformed hydrogen as fuel, and air or oxygen as oxidants to convert chemical energy into electrical energy. They are a new type of energy that is efficient and clean. The production and manufacturing process of hydrogen fuel cell stacks is generally that industrial robots use vacuum suction cups configured on the gripper mechanism to grab the bipolar plates and membrane electrodes from the silo, and then use the robot to transport them to the stacking station where the blind end plates are placed. The two materials are stacked alternately according to the process requirements. After the stacking operation of the core is completed, it enters the press-fitting process for packaging, and finally completes the packaging process of the stack.

[0037] Taking the 150kw hydrogen fuel cell model commonly seen in the market as an example, the design BOM list of its core includes the number of membrane electrode and bipolar plates, which totals about 1000. The processing error range of the bipolar plate is ±0.1mm, and the processing error range of the membrane electrode is ±0.15mm. The height difference of the whole stack after stacking can reach more than ±100mm, which greatly affects the consistency of the product status.

[0038] In this state, if the constant pressure working mode is selected for the press-fitting method, then when press-fitting the battery cores of different heights, the height of the pressed cores will deviate from or exceed the design requirements. For example, if the core height is lower than the design requirements, it is necessary to add fake batteries to compensate for the size, resulting in cost waste; if the core height exceeds the design requirements, it will affect the packaging process of the battery stack, and the assembly cannot be completed, which will increase the difficulty of product quality control and reduce the reasonable rate.

[0039] If the constant size working mode is selected for the press-fitting method, the stress conditions of the single cells in the core will be different. For example, in order to ensure that the height after press-fitting meets the design requirements, if the core height is lower than the design requirements, the press-fitting equipment needs to output a press-fitting force lower than the standard to compress the core to the specified height. At this time, the force on the single cell is also lower than the standard value, which may lead to problems such as insufficient contact area of ​​the proton exchange membrane reaction zone and affect the electrochemical performance; if the core height exceeds the design requirements, the press-fitting equipment needs to output a press-fitting force higher than the standard to compress the core to the specified height. At this time, the force on the single cell will exceed the standard value, which may lead to problems such as reduced gas flow channels or reduced proton exchange membrane permeability and affect the electrochemical performance.

[0040] Therefore, the present application provides a fuel cell stacking method, which uses a thickness measuring device to accurately measure the thickness of the bipolar plate and membrane electrode and record the data before the core is stacked. Then, based on the measurement results, these materials are screened and grouped according to the size of the thickness tolerance and stored in different silos, and the arrangement and combination strategy of material stacking is further predicted. According to the strategy results, the specified materials are selected for repeated stacking, so as to achieve the consistency of the core height to meet the design requirements.

[0041] Before describing the stacking method of the fuel cell of the present application, the thickness measuring device used to measure the thickness of the fuel cell material and the partitioned storage material is first described in detail. The structural principle of the thickness measuring device is as follows: Figure 1 As shown, the thickness measuring equipment includes: MEA (membrane electrode) feeding and placement port 1, BPP (bipolar plate) feeding and placement port 2, a fixed thickness monitoring sensor unit 3, MEA material conveying module 4, BPP material conveying module 5, MEA grabbing and translation mechanism 6, BPP grabbing and translation mechanism 7, MEA storage bins arranged according to tolerance size 8, BPP storage bins arranged according to tolerance size 9, equipment storage area 10, and translation mechanism bracket 11.

[0042] Figure 2 It is a schematic flow chart of a fuel cell stacking method provided in an embodiment of the present application.

[0043] For example, Figure 2 As shown, the method includes:

[0044] In step S101, the material thickness data of the fuel cell is obtained.

[0045] Among them, the materials of the fuel cell include bipolar plates and membrane electrodes. The embodiment of the present application can measure the thickness of the materials of the fuel cell by the above-mentioned thickness measuring equipment.

[0046] Since the materials used to manufacture fuel cells may come from different production batches, processing errors may cause certain dimensional errors in material thickness. Therefore, the order of stacking materials needs to be determined based on the thickness of the materials.

[0047] In step S102, the permutation and combination results that meet the material stacking requirements are predicted according to the material thickness data, and the target material extraction sequence from multiple material bins is determined based on the permutation and combination results.

[0048] Among them, material stacking requirements include target stacking height and target quantity.

[0049] It is understandable that the embodiments of the present application can predict the permutation and combination results that meet the material stacking requirements based on the material thickness data, and determine the target material retrieval order from the material bin based on the multiple permutation and combination results.

[0050] In an embodiment of the present application, the permutation and combination results that meet the material stacking requirements are predicted based on the material thickness data, including: inputting the target height, target quantity and material thickness data into a pre-trained intelligent sorting model, and the intelligent sorting model outputs the permutation and combination results of the material stacking.

[0051] It can be understood that the embodiments of the present application can output the target height required for the fuel cell, the target quantity of the material, and the thickness data of the material to a trained intelligent sorting model, and the intelligent sorting model outputs the arrangement and combination results of the material stacking. The specific sorting process of the intelligent sorting model is explained in the following embodiments and will not be repeated here.

[0052] It should be noted that since the thickness distribution of bipolar plates and membrane electrodes is random, and the results after permutations and combinations may be of very large orders of magnitude, it is necessary to use the powerful computing power and scientific algorithms of industrial AI to collect and calculate the thickness measurement data to obtain convergent calculation results, that is, to output the permutation and combination results of material stacking according to the intelligent sorting model. The intelligent sorting model is pre-trained. After a large amount of data feeding and training, the trained intelligent sorting model will have better generalization ability and interpretability, and the prediction accuracy will also be higher. Among them, the generalization ability refers to the performance of the model on unseen data, and the interpretability refers to whether the structure and parameters of the model can be understood and explained by people.

[0053] In an embodiment of the present application, the sorting process of the intelligent sorting model includes: identifying all material thicknesses in the material thickness data; defining a state list and a state transition probability list of the material, wherein the state list includes a first state indicating that the material is in a material bin and a second state indicating that the material is in a stacking device, and the state transition probability list includes the probability of the material transferring between the first state and the second state; selecting membrane electrodes and bipolar plates as materials from the material bin or stacking device according to the target height, target quantity, state list and state transition probability list, and obtaining the permutation and combination results of the material stacking.

[0054] Among them, the state list includes a first state and a second state, the first state is the material thickness of the material in the material bin, and the second state is the material at the stacking height of the stacking device; the state transition probability list is the probability of the material transferring between the thickness of the material bin and the stacking height of the stacking device.

[0055] It can be understood that the embodiments of the present application can select membrane electrodes and bipolar plates as materials from a material bin or stacking device based on the target height, target quantity, a customized state list, and a customized state transition probability list to obtain the permutation and combination results of the material stacking, so as to subsequently determine the material retrieval order based on the permutation and combination results of the material stacking.

[0056] In an embodiment of the present application, membrane electrodes and bipolar plates are selected from a material bin or a stacking device based on a target height, a target quantity, a state list and a state transition probability list, including: if the current state is the first state, membrane electrodes and / or bipolar plates are selected from the material bin based on the state transition probability list, and the corresponding material in the state list is updated from the first state to the second state; if the current state is the second state, membrane electrodes and / or bipolar plates are selected from the stacking device based on the state transition probability list, and the state of the corresponding material in the state list is updated according to the change in the stacking height; after the state list is updated, the state transition probability of the material in the state transition probability list is updated according to the selected material thickness, target height and target quantity, until the stacking height of the material in the second state reaches the target height and the quantity reaches the target quantity.

[0057] It is understandable that the embodiments of the present application can determine how to select membrane electrode and bipolar plate from the material bin or stacking device based on the current state of the material, specifically:

[0058] If the current state is the first state, select materials, membrane electrode, and bipolar plate from the material bin according to the state transition probability list, and at the same time update the corresponding materials in the state list from the first state to the second state;

[0059] If the current state is the second state, materials, membrane electrodes, and bipolar plates are selected from the stacking device according to the state transition probability list, and the states of the corresponding materials in the state list are updated according to the changes in the stacking height.

[0060] After the state list is updated, the state transition probability of the material in the state transition probability list is updated according to the thickness, target height and target quantity of the selected material until the stacking height of the material in the second state (i.e., the stacking device) reaches the target height and the quantity reaches the target quantity, then stacking is stopped.

[0061] It can be understood that the embodiment of the present application defines a state update rule, which is used to randomly take out a certain number of materials from the current state and add them to another state, and then randomly take out a certain number of materials from the current state, and update the state and probability distribution of the materials according to the size of the random number. After the state list is updated, the state transition probability of the materials in the state transition probability list is updated according to the thickness, target height and target quantity of the selected materials, until the material stacking height reaches the target height and the quantity reaches the target quantity, and then the stacking stops.

[0062] For example, if the current state is the first state, randomly select some materials from the material bin, and add 1 to their material thickness and stacking height, then add the new state to the second state, and update the second state and transition probability of the stacking device; if the current state is the second state, randomly select some materials from the stacking device, and perform some random operations on their stacking height. For example, there is a 50% probability of reducing the stacking height by 1, and a 50% probability of adding 1 to the stacking height. If the stacking height is 0, update the second state to 1, and then iterate the update state, using the second state and transition probability as input to obtain the updated state and transition probability.

[0063] In an embodiment of the present application, after defining the state list and state transition probability list of the material, it also includes: initializing the state list and the state transition probability list, wherein the values ​​of all elements in the initialized state list are the same, and the state transition probabilities of all materials in the initialized state transition probability list are the same.

[0064] It can be understood that the embodiment of the present application can initialize the state list and the state transition probability list, the values ​​of all elements in the initialized state list are the same, and the state transition probabilities of all materials in the initialized state transition probability list are the same. For example, all elements in the initialized state list are 0, and the initialized transition probabilities are all 1.

[0065] In an embodiment of the present application, before inputting the target height, target quantity and material thickness data into a pre-trained intelligent sorting model, it also includes: obtaining the stacking data of the membrane electrode and bipolar plate in the fuel cell; generating a training data set and a verification set based on the stacking data; selecting a machine learning or deep learning model as the intelligent sorting model, training the intelligent sorting model according to the training data set, and verifying the trained intelligent sorting model according to the verification set, and completing the training when the performance and accuracy of the intelligent sorting model meet the preset requirements.

[0066] It can be understood that in the embodiment of the present application, a machine learning or deep learning model can be selected as an intelligent sorting model. The specific model can be selected according to the characteristics of the data and the application scenario. A training data set and a verification set are generated based on the stacking data of the membrane electrode and the bipolar plate in the fuel cell. The training data set is used for training, and the verification set is used to verify the trained intelligent sorting model. The training is completed when the performance and accuracy of the intelligent sorting model meet the preset requirements, and the final intelligent sorting model is deployed and applied to practical applications. The intelligent sorting model can be continuously iterated and optimized through continuous data feeding training, and highly matched with the current battery stacking application scenario, thereby achieving more efficient and accurate intelligent sorting model strategy generation and better auxiliary production.

[0067] The training process of the intelligent sorting model includes:

[0068] Step 1: Data collection and cleaning: First, you need to collect a large amount of data and clean the data to remove noise and outliers.

[0069] Step 2: Feature Engineering: Perform feature engineering on the data to extract important features from the data in order to better characterize the data.

[0070] Step 3: Model selection: Select a suitable machine learning or deep learning model based on the characteristics of the data and the application scenario.

[0071] Step 4: Model training: Use the data set to train the model and iteratively optimize the model parameters so that the model can accurately predict or analyze the data.

[0072] Step 5: Model evaluation: Use the validation set to evaluate the trained model to determine the performance and accuracy of the model.

[0073] Step 6: Model optimization: Based on the evaluation results, optimize the model to improve its performance and accuracy.

[0074] Step 7: Deployment: Deploy the final optimized model to the production environment for practical application.

[0075] In step S103, materials are retrieved from corresponding material bins according to the target material retrieval sequence and repeatedly stacked.

[0076] It can be understood that, based on the target material fetching sequence obtained in the above steps, materials are fetched from the corresponding material bins according to the target material fetching sequence and repeatedly stacked, thereby achieving a fixed stacking height and achieving consistency in core height to meet design requirements.

[0077] In an embodiment of the present application, materials are taken from corresponding material bins according to a target material taking sequence and repeatedly stacked, including: taking materials from corresponding material bins according to a target material taking sequence; repeatedly and alternately stacking the taken-out membrane electrode and bipolar plates until the stacking numbers of the membrane electrode and bipolar plates reach the target number.

[0078] It can be understood that the embodiments of the present application can take out the membrane electrode and bipolar plate from the corresponding material bin according to the target material collection order, and repeatedly stack the membrane electrode and bipolar plate alternately until the stacking number reaches the target number. On the basis of reaching the target number, it means that the core height reaches the required core height design requirement, thereby achieving consistency in core height and avoiding the problem of inconsistent core stacking height, which makes it impossible to ensure the consistency of internal stress of each battery stack after pressing, and further fails to ensure product quality.

[0079] Specifically, the method for stacking the above-mentioned fuel cells is mainly divided into two parts, thickness measurement and stacking. The specific process is as follows:

[0080] Step 1: Use thickness measuring equipment to accurately measure the thickness of bipolar plates and membrane electrodes and record the data.

[0081] Step 2: Based on the measurement results, the materials are screened and grouped according to the thickness tolerance and stored in different material bins.

[0082] Step 3: After the material has completed the thickness measurement process, it enters the stacking station for use.

[0083] Step 4: Before grabbing the material, analyze and calculate according to the material thickness grouping data and the set stacking height, predict the arrangement and combination strategy of material stacking, and select the specified material according to the strategy results.

[0084] Step 5: The strategy of step 4 will be output as a command signal to the control system of the six-axis robot, and the action will be executed to pick up materials and repeatedly stack them in sequence from the material bin specified by the strategy.

[0085] Among them, the structural principle diagram of the stacking device is as follows Figure 3 As shown, the stacking equipment includes: a bipolar plate handling robot 12, a membrane electrode handling robot 13, a bipolar plate storage bin 14, a membrane electrode storage bin 15, and a core stacking station tooling 16. The working principle of the equipment is: the robot absorbs materials from the material bin, transports them to the stacking station for stacking, and repeats this process to complete the stacking operation.

[0086] In addition, regarding thickness measurement data collection, in order to improve the immediacy and volume of data transmission, the edge computing + intelligent sorting model solution can be used to achieve this, that is, by deploying industrial edge computing equipment and control systems, and connecting the thickness measurement equipment and the PLC control system of the stacking equipment. At the same time, the intelligent sorting model can be deployed in the edge computing control system, with the thickness data collected by the edge computing system as input, and the strategy results calculated by the intelligent sorting model can be output to the stacking equipment through the edge computing system to achieve guidance and intervention in the stacking operation process. The system architecture principle is as follows: Figure 4 shown.

[0087] The analysis and calculation in step 4 above (intelligent sorting model) can use the Markov chain algorithm to predict the state of the material and the stacking height at a certain moment in the future, and calculate the probability distribution between them to subsequently achieve a fixed stacking height.

[0088] It should be noted that the Markov chain algorithm is a probabilistic model used to process sequence data. The core idea is to regard the sequence data as a state space, in which each state has a probability distribution. The Markov chain algorithm calculates the transition probability between states by constructing a state transition matrix, thereby realizing the modeling of sequence data. The Markov chain algorithm has the ability to generalize and can process data of different lengths, different features, and different patterns, and can automatically adapt to changes in data. That is, it has the ability to play its role in a wider range of data. This is because the Markov chain algorithm has the properties of a state transition matrix, which can automatically calculate the transition probability between states based on the input data, thereby ensuring the generalization ability of the algorithm. The Markov chain algorithm can be deployed on various structures, including artificial neural networks, stochastic gradient descent, decision trees, etc., and general structures can also be used.

[0089] The Markov chain algorithm is an algorithm for processing random processes and state transitions, and can be used to randomly select some elements from a given data set and stack them together to achieve a fixed stacking height. In the stacking scenario of this application, the Markov chain algorithm can be used to represent the process of randomly taking out a certain number of bipolar plates and membrane electrodes from a thickness measuring device, and applying their state transition to the stacking device to achieve a fixed stacking height.

[0090] The implementation principles and methods are as follows:

[0091] 1. Define the state transfer equation: According to the structure and state of the stacking device, the state transfer equation can be defined to describe the process of randomly taking out a certain number of bipolar plates and membrane electrodes from the thickness measuring device. For example, the state transfer equation can be defined to describe the process of randomly taking out a certain number of bipolar plates from a certain thickness and placing them in different positions in the stacking device.

[0092] 2. Initialize the Markov chain: Before starting the stack, the Markov chain needs to be initialized. A random number generator can be used to generate some random bipolar plate and membrane electrode states and add them to the state chain.

[0093] 3. State transfer: When some bipolar plates and membrane electrodes need to be selected from the state chain, the random selection algorithm in the Markov chain algorithm can be used to select some states. For example, a random number generator can be used to select states, or the Markov state transfer equation in the Markov chain algorithm can be used to select states.

[0094] 4. Update the state chain: After a certain number of bipolar plates and membrane electrodes are selected, they need to be placed in a stacking device to achieve a fixed stacking height. The state chain can be updated using the update algorithm in the Markov chain algorithm to reflect the process of removing bipolar plates and membrane electrodes from the stacking device. For example, a new state can be added to the state chain, or an old state can be updated to a new state.

[0095] 5. Output: The output of the Markov chain algorithm can be used to predict future stack heights. For example, the output of the Markov chain algorithm can be used to train a model to predict which bipolar plates and membrane electrodes will be placed into a stack to achieve a fixed stack height.

[0096] Specifically, the Markov chain algorithm is used in the stacking method of the fuel cell in the embodiment of the present application, and two states can be defined: material thickness and stacking height. Each state has a probability distribution, which represents the probability of transferring from this state to other states. First, two initial states and initial probability distributions are defined. Then a state update rule is defined to randomly take out a certain number of materials from the current state and add them to another state. Next, the algorithm randomly takes out a certain number of materials from the current state, and updates the state and probability distribution of these materials according to the size of the random number. Finally, the algorithm returns the updated state and probability distribution. Thereby realizing the state used to predict the material and stacking height at a certain moment in the future, and calculating the probability distribution between them. In addition, the algorithm can also be used to design some automated systems, such as automatic batching systems or automatic stacking systems.

[0097] According to the fuel cell stacking method proposed in the embodiment of the present application, the material thickness data of the fuel cell can be analyzed and processed, and the arrangement and combination results that meet the material stacking requirements can be predicted based on the thickness data, so as to realize intelligent material collection and stacking, thereby achieving fixed height stacking, making the core stacking height consistent, and improving product quality.

[0098] Figure 5It is a schematic structural diagram of a fuel cell stacking device provided in an embodiment of the present application.

[0099] For example, Figure 5 As shown, the device 20 may include:

[0100] Acquisition module 100: used to acquire material thickness data of the fuel cell;

[0101] Determination module 200: used to predict the permutation and combination results that meet the material stacking requirements according to the material thickness data, and determine the target material extraction sequence from multiple material bins based on the permutation and combination results;

[0102] The stacking module 300 is used to retrieve materials from the corresponding material bins and stack them repeatedly according to the target retrieval sequence.

[0103] In an embodiment of the present application, the material stacking requirements include a target stacking height and a target quantity, and the determination module 200 is further used to: input the target height, target quantity and material thickness data into a pre-trained intelligent sorting model, and the intelligent sorting model outputs the permutation and combination results of the material stacking.

[0104] In an embodiment of the present application, the sorting process of the intelligent sorting model includes: identifying all material thicknesses in the material thickness data; defining a state list and a state transition probability list of the material, wherein the state list includes a first state indicating that the material is in a material bin and a second state indicating that the material is in a stacking device, and the state transition probability list includes the probability of the material transferring between the first state and the second state; selecting membrane electrodes and bipolar plates as materials from the material bin or stacking device according to the target height, target quantity, state list and state transition probability list, and obtaining the permutation and combination results of the material stacking.

[0105] In an embodiment of the present application, membrane electrodes and bipolar plates are selected from a material bin or a stacking device based on a target height, a target quantity, a state list and a state transition probability list, including: if the current state is the first state, membrane electrodes and / or bipolar plates are selected from the material bin based on the state transition probability list, and the corresponding material in the state list is updated from the first state to the second state; if the current state is the second state, membrane electrodes and / or bipolar plates are selected from the stacking device based on the state transition probability list, and the state of the corresponding material in the state list is updated according to the change in the stacking height; after the state list is updated, the state transition probability of the material in the state transition probability list is updated according to the selected material thickness, target height and target quantity, until the stacking height of the material in the second state reaches the target height and the quantity reaches the target quantity.

[0106] In an embodiment of the present application, after defining the state list and state transition probability list of the material, it also includes: initializing the state list and the state transition probability list, wherein the values ​​of all elements in the initialized state list are the same, and the state transition probabilities of all materials in the initialized state transition probability list are the same.

[0107] In an embodiment of the present application, before inputting the target height, target quantity and material thickness data into a pre-trained intelligent sorting model, it also includes: obtaining the stacking data of the membrane electrode and bipolar plate in the fuel cell; generating a training data set and a verification set based on the stacking data; selecting a machine learning or deep learning model as the intelligent sorting model, training the intelligent sorting model according to the training data set, and verifying the trained intelligent sorting model according to the verification set, and completing the training when the performance and accuracy of the intelligent sorting model meet the preset requirements.

[0108] In an embodiment of the present application, the material includes membrane electrodes and bipolar plates, and the stacking module 300 is further used to: take materials from the corresponding material bin according to the target material taking order; repeatedly and alternately stack the taken-out membrane electrodes and bipolar plates until the stacking numbers of the membrane electrodes and bipolar plates reach the target number.

[0109] According to the fuel cell stacking device proposed in the embodiment of the present application, the material thickness data of the fuel cell can be analyzed and processed, and the arrangement and combination results that meet the material stacking requirements can be predicted based on the thickness data, so as to realize intelligent material extraction and stacking, thereby achieving fixed height stacking, making the core stacking height consistent, and improving product quality.

[0110] Figure 6 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device may include:

[0111] A memory 601 , a processor 602 , and a computer program stored in the memory 601 and executable on the processor 602 .

[0112] When the processor 602 executes the program, the fuel cell stacking method provided in the above embodiment is implemented.

[0113] Furthermore, the electronic device further comprises:

[0114] The communication interface 606 is used for communication between the memory 601 and the processor 602 .

[0115] The memory 601 is used to store computer programs that can be executed on the processor 602 .

[0116] The memory 601 may include a high-speed RAM (Random Access Memory) memory, and may also include a non-volatile memory, such as at least one disk memory.

[0117] If the memory 601, the processor 602 and the communication interface 606 are implemented independently, the communication interface 606, the memory 601 and the processor 602 can be connected to each other through a bus and communicate with each other. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0118] Optionally, in a specific implementation, if the memory 601, the processor 602 and the communication interface 606 are integrated on a chip, the memory 601, the processor 602 and the communication interface 606 can communicate with each other through an internal interface.

[0119] The processor 602 may be a CPU (Central Processing Unit), or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present application.

[0120] The embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the above-mentioned fuel cell stacking method is implemented.

[0121] 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 this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0122] 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 application. In this specification, the schematic representations of the above terms are not necessarily directed 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.

[0123] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A fuel cell stacking method, characterized in that: The method comprises: Obtain material thickness data of fuel cells; Predicting the permutation and combination results that meet the material stacking requirements according to the material thickness data, and determining the target material extraction sequence from multiple material bins based on the permutation and combination results; According to the target material taking sequence, materials are taken from the corresponding material bins for repeated stacking.

2. The method according to claim 1, characterized in that The material stacking requirements include a target stacking height and a target quantity, and the prediction of the permutation and combination results that meet the material stacking requirements based on the material thickness data includes: The target height, the target quantity and the material thickness data are input into a pre-trained intelligent sorting model, and the intelligent sorting model outputs the arrangement and combination results of the material stacking.

3. The method according to claim 2, characterized in that The sorting process of the intelligent sorting model includes: Identifying all material thicknesses in the material thickness data; Define a state list and a state transition probability list of a material, wherein the state list includes a first state indicating that the material is in a material bin and a second state indicating that the material is in a stacking device, and the state transition probability list includes the probability of the material transitioning between the first state and the second state; Membrane electrodes and bipolar plates are selected as materials from the material bin or the stacking device according to the target height, the target quantity, the state list and the state transition probability list to obtain a permutation and combination result of the material stacking.

4. The method according to claim 3, characterized in that The selecting a membrane electrode and a bipolar plate from the material bin or the stacking device according to the target height, the target quantity, the state list, and the state transition probability list includes: If the current state is the first state, selecting a membrane electrode and / or a bipolar plate from the material bin based on the state transition probability list, and updating the material corresponding to the state list from the first state to the second state; If the current state is the second state, selecting a membrane electrode and / or a bipolar plate from the stacking device based on the state transition probability list, and updating the state of the material corresponding to the state list according to the change of the stacking height; After the state list is updated, the state transition probability of the material in the state transition probability list is updated according to the selected material thickness, the target height and the target quantity until the stacking height of the material in the second state reaches the target height and the quantity reaches the target quantity.

5. The method according to claim 3, characterized in that: After defining the material's state list and state transition probability list, it also includes: Initialize the state list and the state transition probability list, wherein the values ​​of all elements in the initialized state list are the same, and the state transition probabilities of all materials in the initialized state transition probability list are the same.

6. The method according to any one of claims 2 to 5, characterized in that: Before inputting the target height, the target quantity and the material thickness data into the pre-trained intelligent sorting model, the method further includes: Obtain stacking data of membrane electrode and bipolar plates in fuel cells; Generate a training data set and a validation set based on the stacked data; A machine learning or deep learning model is selected as the intelligent sorting model, the intelligent sorting model is trained according to the training data set, and the trained intelligent sorting model is verified according to the verification set, and the training is completed when the performance and accuracy of the intelligent sorting model meet the preset requirements.

7. The method according to claim 1, characterized in that The materials include membrane electrodes and bipolar plates, and the process of taking materials from corresponding material bins and repeatedly stacking them according to the target material taking sequence includes: Pick up materials from the corresponding material bin according to the target picking order; The taken-out membrane electrodes and bipolar plates are repeatedly stacked alternately until the stacking numbers of the membrane electrodes and the bipolar plates reach the target number.

8. A fuel cell stacking device, characterized in that: The device comprises: An acquisition module, used to acquire material thickness data of a fuel cell; A determination module, used to predict the permutation and combination results that meet the material stacking requirements according to the material thickness data, and determine the target material extraction sequence from multiple material bins based on the permutation and combination results; The stacking module is used to fetch materials from the corresponding material bins and stack them repeatedly according to the target fetching sequence.

9. An electronic device, characterized in that: The electronic device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor executes the program to implement the fuel cell stacking method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed, the fuel cell stacking method according to any one of claims 1 to 7 is implemented.

Images