CN116306288A - Agent model-based lithium battery optimal design method, system, device and medium - Google Patents
Agent model-based lithium battery optimal design method, system, device and medium Download PDFInfo
- Publication number
- CN116306288A CN116306288A CN202310274472.5A CN202310274472A CN116306288A CN 116306288 A CN116306288 A CN 116306288A CN 202310274472 A CN202310274472 A CN 202310274472A CN 116306288 A CN116306288 A CN 116306288A
- Authority
- CN
- China
- Prior art keywords
- lithium battery
- design
- model
- proxy model
- variables
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 title claims abstract description 179
- 229910052744 lithium Inorganic materials 0.000 title claims abstract description 179
- 238000013461 design Methods 0.000 title claims abstract description 169
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000012549 training Methods 0.000 claims abstract description 41
- 238000013528 artificial neural network Methods 0.000 claims description 47
- 238000005457 optimization Methods 0.000 claims description 34
- 238000004590 computer program Methods 0.000 claims description 12
- 238000003860 storage Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 abstract description 7
- 239000003795 chemical substances by application Substances 0.000 description 12
- 238000010586 diagram Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 7
- 238000007796 conventional method Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 description 5
- 229910001416 lithium ion Inorganic materials 0.000 description 5
- 239000007791 liquid phase Substances 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 4
- 239000007790 solid phase Substances 0.000 description 4
- 238000012795 verification Methods 0.000 description 4
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000012938 design process Methods 0.000 description 3
- 238000013401 experimental design Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 239000002245 particle Substances 0.000 description 3
- 238000004088 simulation Methods 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000007774 positive electrode material Substances 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 210000004027 cell Anatomy 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 239000011800 void material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
- Y02E60/10—Energy storage using batteries
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computer Hardware Design (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Biomedical Technology (AREA)
- Geometry (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Charge And Discharge Circuits For Batteries Or The Like (AREA)
Abstract
The application provides a lithium battery optimal design method, a system, a device and a medium based on a proxy model, wherein the lithium battery optimal design method based on the proxy model comprises the following steps: constructing and training a surrogate model, the surrogate model input comprising a plurality of lithium battery design variables, the surrogate model output comprising at least one lithium battery performance parameter; and inputting design combination values of the plurality of lithium battery design variables into the proxy model to predict and obtain the lithium battery performance parameters. According to the method, the agent model is used for processing the design variables of the lithium battery so as to predict the performance parameters of the lithium battery, the result can be output rapidly, the time of battery design is shortened, and the efficiency of optimal design of the lithium battery is improved.
Description
Technical Field
The application belongs to the technical field of lithium batteries, and relates to a lithium battery optimal design method, a system, a device and a medium based on a proxy model.
Background
Lithium ion batteries have been widely used in various fields of application ranging from consumer electronics to electric automobiles. In order to meet the increasing demands for higher energy and power capability, durability and safety of batteries, the design of lithium ion batteries has become critical to avoid any unexpected performance loss. Experimental-based cell designs are time consuming and expensive. In contrast, simulation-based designs are not only more efficient, but also allow for a deeper understanding of the mechanisms controlling battery performance.
Current lithium battery designs can be broadly divided into two categories: a high power type battery and a high energy type battery. The energy type battery is characterized by high energy density and is mainly used for high energy output; the power type battery is characterized by high power density and is mainly used for instantaneous high-power input and output batteries. How to quantitatively measure the settings of different parameters in the design process of the lithium battery is an important subject for different types of batteries with different factors to be considered in the design process.
In the electrochemical model of the lithium battery, the P2D model can simulate the whole battery structure from the theory of a porous electrode, and is generally used for optimizing important electrode parameters such as electrode thickness, void ratio, particle size and the like. However, starting from the P2D model, the coupled partial differential equation needs to be solved for the design of the lithium battery, which is time-consuming, consumes a lot of resources, and has high calculation cost in the direct battery design.
Disclosure of Invention
The purpose of the application is to provide a lithium battery optimal design method, a system, a device and a medium based on a proxy model, which are used for solving the problems existing in the prior art.
In a first aspect, the present application provides a method for optimizing a lithium battery design based on a proxy model, the method comprising: constructing and training a surrogate model, the surrogate model input comprising a plurality of lithium battery design variables, the surrogate model output comprising at least one lithium battery performance parameter; and inputting design combination values of the plurality of lithium battery design variables into the proxy model to predict and obtain the lithium battery performance parameters.
In one implementation of the first aspect, the proxy model is a neural network-based proxy model.
In one implementation manner of the first aspect, the proxy model includes a classification neural network for classifying the design combination values of the plurality of lithium battery design variables to output whether the design combination values of the plurality of lithium battery design variables are valid or not, and a fitting neural network for fitting the valid design combination values of the plurality of lithium battery design variables to output the lithium battery performance parameter.
In one implementation of the first aspect, the at least one lithium battery performance parameter includes at least one of lithium battery energy and lithium battery power.
In one implementation manner of the first aspect, the method further includes: and comparing the performance parameters of the lithium battery with the design requirements based on the prediction, adjusting the design combination values of the design variables of the lithium battery, predicting again by adopting the agent model, and executing circularly until the design requirements are met.
In one implementation of the first aspect, training the proxy model includes training the classification neural network and the fitting neural network separately.
In a second aspect, the present application provides a lithium battery optimization design system based on a proxy model, the system comprising: a model building training module configured to build and train a proxy model, an input of the proxy model comprising a plurality of lithium battery design variables, an output of the proxy model comprising at least one lithium battery performance parameter; a prediction module configured to input design combination values of the plurality of lithium battery design variables to the proxy model to predictively obtain the lithium battery performance parameters.
In one implementation manner of the second aspect, the system further includes a loop optimization module configured to obtain a comparison of the lithium battery performance parameter and the design requirement based on the prediction, adjust the design combination value of the plurality of lithium battery design variables, and predict again using the proxy model, and perform the loop until the design requirement is satisfied.
In a third aspect, the present application provides a lithium battery optimization design device based on a proxy model, where the device includes: a memory configured to store a computer program; and a processor configured to invoke the computer program to perform the proxy model-based lithium battery optimization design method according to the first aspect of the present application.
In a fourth aspect, the present application provides a computer-readable storage medium having stored thereon a computer program that is executed to implement the proxy model-based lithium battery optimization design method according to the first aspect of the present application.
As described above, the agent model-based lithium battery optimization design method, system, device and medium have the following beneficial effects:
(1) According to the method, the agent model is used for processing the design variables of the lithium battery so as to predict the performance parameters of the lithium battery, the result can be output rapidly, the time of battery design is shortened, and the efficiency of optimal design of the lithium battery is improved.
(2) According to the method, an agent model is built based on a classified neural network and a fitting neural network, firstly, the design combination of a plurality of lithium battery design variables can be used for classifying, whether the design variables meet the practical physical significance is determined, secondly, when the design variables are effective, the lithium battery performance parameters of the battery corresponding to the design variables can be rapidly output, the prediction accuracy and speed of the agent model are guaranteed, compared with a conventional method for solving a P2D electrochemical model, the time of the agent model is mainly spent on training of the neural network, once training is finished, the result can be rapidly output, and the conventional method is limited by finite elements, the time for solving a group of variables is usually several minutes, and in battery manufacturing and design, the parameter combination usually needs multiple tests, and the conventional method can consume a large amount of time.
Drawings
Fig. 1 is a flowchart of a method for optimizing a design of a lithium battery based on a proxy model according to an embodiment of the present application.
Fig. 2A is a schematic diagram of a structure of a classification neural network according to an embodiment of the present application.
Fig. 2B is a schematic diagram of a fitting neural network in an embodiment of the present application.
Fig. 3 is a flowchart of a method for optimizing a design of a lithium battery based on a proxy model according to another embodiment of the present application.
Fig. 4 is a schematic structural diagram of a lithium battery optimization design system based on a proxy model according to an embodiment of the present application.
Fig. 5 is a schematic structural diagram of a lithium battery optimization design system based on a proxy model according to another embodiment of the present application.
Fig. 6 is a schematic structural diagram of a lithium battery optimization design device based on a proxy model according to an embodiment of the present application.
Description of element reference numerals
4. Lithium battery optimization design system based on proxy model
41. Model building training module
42. Prediction module
43. Circulation optimization module
6. Lithium battery optimal design device based on agent model
61. Memory device
62. Processor and method for controlling the same
S1 to S3 steps
Detailed Description
Other advantages and effects of the present application will become apparent to those skilled in the art from the present disclosure, when the following description of the embodiments is taken in conjunction with the accompanying drawings. The present application may be embodied or carried out in other specific embodiments, and the details of the present application may be modified or changed from various points of view and applications without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments may be combined with each other without conflict.
It should be noted that, the illustrations provided in the following embodiments merely illustrate the basic concepts of the application by way of illustration, and only the components related to the application are shown in the drawings and are not drawn according to the number, shape and size of the components in actual implementation, and the form, number and proportion of the components in actual implementation may be arbitrarily changed, and the layout of the components may be more complex.
The application provides a lithium battery optimal design method, system, device and medium based on a proxy model. According to the method and the device, the proxy model is used for processing the plurality of lithium battery design variables to predict the performance parameters of the lithium battery, and the result can be output rapidly, so that the prediction of the performance parameters of the lithium battery under different combinations of the lithium battery design variables can be realized, and the time of battery design is shortened.
The following describes the technical solutions in the embodiments of the present application in detail with reference to the drawings in the embodiments of the present application.
As shown in fig. 1, the present embodiment provides a method for optimizing a lithium battery design based on a proxy model, which includes the following steps S1 and S2.
In step S1, a surrogate model is constructed and trained, the inputs of which include a plurality of lithium battery design variables, and the outputs of which include at least one lithium battery performance parameter.
This application focuses on the variables that are controllable during battery manufacturing. Specifically, in this embodiment, only the positive electrode is concerned in the lithium battery design process, and the lithium battery design variables concerned are usually the inherent properties of the electrode, and 6 variables including the positive electrode thickness, the positive electrode solid phase volume fraction, the positive electrode Bruggeman coefficient, the positive electrode active material particle radius, the charge-discharge rate C and the initial liquid phase lithium ion concentration are selected as the lithium battery design variables in this embodiment.
The design of lithium batteries is divided into energy batteries and power batteries, and in order to measure whether the designed battery meets the power or energy requirements, the lithium battery performance parameters include at least one of lithium battery energy and lithium battery power.
In a preferred embodiment, the proxy model is a neural network-based proxy model, including a classification neural network for classifying the design combination values of the plurality of lithium battery design variables to output whether the design combination values of the plurality of lithium battery design variables are valid or not, and a fitting neural network for fitting the valid design combination values of the plurality of lithium battery design variables to output the lithium battery performance parameter.
Fig. 2A is a schematic diagram of a classification neural network, which includes an input layer, a hidden layer, and an output layer.
In a preferred embodiment, the input layer is the above-determined 6 lithium battery design variables, the hidden layer comprises 10 neurons, the hidden layer uses a sigmoid activation function, the output layer uses a softmax function for classification, an output form of 0 indicates that the parameter combination is not available against the physical meaning, and an output of 1 indicates that the parameter combination is available within the physical meaning.
Fig. 2B is a schematic diagram of a fitting neural network, which has a similar structure to that of a classification neural network, and includes an input layer, a hidden layer, and an output layer.
In a preferred embodiment, the input of the fitting neural network is a combination of parameters corresponding to an output of 1 in the classification neural network, the hidden layer is 10 neural units, the activation function is a sigmoid function, and the output layer uses a linear function for fitting calculation to obtain lithium battery energy and/or power.
In the embodiment shown in fig. 2B, the output layer is 2 layers, and lithium battery energy and lithium battery power are simultaneously output for 2 lithium battery performance parameters. It should be noted that, in other embodiments, the output layer in the network may be adjusted to be 1 layer, so as to output 1 lithium battery performance parameter, which may be any one of lithium battery energy or lithium battery power, according to design requirements.
Training the surrogate model in step S1 includes training the classification neural network and the fitting neural network separately.
The specific training process comprises the following steps:
(1) Acquisition of training data
First, after determining the design variables of the lithium battery, the relevant literature is referred to, giving an approximate range of the corresponding variables, depending on the battery actually used. And generating 900 groups of experimental design variables based on Latin hypercube sampling, wherein the 900 groups of experimental design variables comprise effective variable combinations, labels are 1 and invalid variable combinations, and the labels of 0,0 and 1 are added after manual judgment, and the generated data are used for training a classified neural network.
Further, in order to train the fitting neural network, the lithium battery performance parameters under the combination of effective lithium battery design variables need to be obtained through simulation, which specifically includes:
numerical modeling is carried out on the lithium battery based on a physical-chemical theory basis of a pseudo two-dimensional electrochemical model, and specifically, the pseudo two-dimensional model comprises four partial differential equations and an algebraic equation; the four partial differential equations are solid phase potential, solid phase mass transfer, liquid phase potential and liquid phase mass transfer, respectively, and the algebraic equation is the Butler-Volmer equation. Through the five equations, the physical and chemical state quantity in the internal time space of the lithium battery can be accurately described. In addition to the above equation, it is also necessary to limit the boundary and give corresponding boundary conditions such as lithium ion concentration at the initial time, physical quantity continuity at the solid-liquid boundary, and the like.
And (3) selecting a proper numerical method to carry out numerical solution on the partial differential equation in the step one. Common numerical methods include finite difference, finite element, finite volume, and the like. Wherein the finite difference is the easiest to implement, easy to program, but is condition stable; the finite element method not only has high solving precision, but also can adapt to any complex shape. To ensure accuracy of the training data, we choose here a finite element method to solve partial differential equations in the P2D model. After the finite element method is solved, a large amount of data can be obtained. The input parameter values (effective lithium battery design variable combinations) and output voltage values in each simulation experiment are recorded.
Calculating lithium battery energy and/or lithium battery power by:
wherein P is lithium battery power, E is lithium battery energy, I is input current, V is output voltage, m is lithium battery mass, wherein the lithium battery mass is the mass combination of a current collector, a separator and an electrode, and t is d Is the time required for the battery to discharge to the lowest voltage.
Thus, the calculated lithium battery energy and/or lithium battery power is used as a label value for training the fitting neural network.
(2) Training a classified neural network
Training of the classification neural network requires the use of the 900 set of experimental variables described above. The 900 sets of experimental variables are labeled accordingly, with either 0 or 1,0 indicating that the 6 parameter combinations are not available against the actual physical meaning, and 1 indicating that the 6 parameter combinations are available within the physical meaning. During training, 900 groups of experimental variables are divided according to a training set and a verification set which are 8:2, and the training set and the verification set are mutually independent and do not cross. The purpose of the training set is to evaluate the classification ability of the neural network one, for example, to classify whether the input variable is within a physical range. The 720 sets of data are used for training the classified neural network, and the other 180 sets of experimental variables are used for verifying whether the training of the classified neural network is accurate or not, and when the accuracy reaches more than 95%, the training of the classified neural network is finished.
(3) Training fitting neural networks
After the 900 groups of experimental design variables are classified by the classified neural network, the parameter combination with the output result corresponding to 1 accords with the actual physical meaning. And taking the parameter combination with the output result corresponding to 1 as the input of the step fitting neural network. The parameter combination with the output result of 1 is firstly simulated and solved by using a pseudo two-dimensional model, and simultaneously the input current I and the output voltage V of each group of parameter sets are recorded. The obtained numerical solution is still divided according to the ratio of 8:2 of a training set and a verification set, the training set is used for fitting data samples to obtain lithium battery energy and/or lithium battery power, and the verification set is used for evaluating the ability of fitting the neural network to fit real data.
Thus, training of the proxy model is completed to enable the proxy model to process a plurality of lithium battery design variables to predict performance parameters of the lithium battery.
In step S2, design combination values of the plurality of lithium battery design variables are input to the proxy model to predict and obtain the lithium battery performance parameters.
In the step S2, the proxy model trained in the step S1 is adopted to predict the performance parameters of the lithium battery, namely, the lithium battery energy and/or lithium battery power data can be obtained only by inputting the design combination value of the lithium battery design variables into the proxy model, and therefore whether the combination of the lithium battery design variables meets the actual requirements is judged. In a specific embodiment, the design variables of the lithium battery include 6 variables including positive electrode thickness, positive electrode solid phase volume fraction, positive electrode Bruggeman coefficient, positive electrode active material particle radius, charge-discharge multiplying power C and initial liquid phase lithium ion concentration, and the design values of the 6 variables are input into a proxy model to be predicted, so as to obtain lithium battery energy and/or battery power data, and whether the design values of the 6 variables meet actual requirements is judged based on the predicted lithium battery energy and/or battery power data. It should be noted that, when the design values of the 6 variables are not available against the physical meaning, the classification neural network in the proxy model will directly output 0, and no data fitting is performed to obtain the lithium battery performance parameters.
As shown in fig. 3, in another preferred embodiment, the present embodiment provides a method for optimizing a lithium battery based on a proxy model, which includes the following steps S1 to S3.
In step S1, a surrogate model is constructed and trained, the inputs of which include a plurality of lithium battery design variables, and the outputs of which include at least one lithium battery performance parameter.
In step S2, design combination values of the plurality of lithium battery design variables are input to the proxy model to predict and obtain the lithium battery performance parameters.
The specific solutions of the above steps S1 and S2 are consistent with the foregoing preferred embodiments, and will not be repeated here.
In step S3, comparing the performance parameter of the lithium battery with the design requirement based on the prediction, adjusting the design combination values of the design variables of the lithium battery, predicting again by using the proxy model, and executing circularly until the design requirement is met.
In step S3, the predicted performance parameters of the lithium battery are compared with the design requirements, and the sizes of the design variables are adjusted to repeat the prediction, so that the corresponding parameter ranges can be very conveniently adjusted, the battery design requirements are realized, and the time for determining the variable ranges in the battery design is greatly reduced.
According to the method and the device, the proxy model is used for processing the design variables of the plurality of lithium batteries so as to predict the performance parameters of the lithium batteries, the result can be output rapidly, and the time of battery design is shortened. Further, the agent model is built based on the classification neural network and the fitting neural network, firstly, the design combination of a plurality of lithium battery design variables can be used for classifying, whether the design variables are in accordance with actual physical significance is determined, secondly, when the design variables are effective, the lithium battery performance parameters of the battery corresponding to the design variables can be rapidly output, the prediction accuracy and speed of the agent model are guaranteed, compared with a conventional method for solving the P2D electrochemical model, the time of the agent model is mainly consumed in training of the neural network, once training is completed, the result can be rapidly output, and the conventional method is limited by finite elements, the time for solving a group of variables is usually several minutes, in battery manufacturing and design, the parameter combination usually needs multiple tests, and the conventional method can consume a large amount of time.
The protection scope of the lithium battery optimization design method based on the agent model according to the embodiment of the application is not limited to the execution sequence of the steps listed in the embodiment, and all the schemes realized by increasing or decreasing the steps and replacing the steps according to the prior art made according to the principles of the application are included in the protection scope of the application.
The embodiment of the application also provides a lithium battery optimization design system based on the proxy model, which can realize the lithium battery optimization design method based on the proxy model, but the implementation device of the lithium battery optimization design method based on the proxy model, which is disclosed by the application, comprises but is not limited to the structure of the lithium battery optimization design system based on the proxy model, and all the structural deformation and replacement of the prior art according to the principles of the application are included in the protection scope of the application.
As shown in fig. 4, the present embodiment provides a lithium battery optimization design system based on a proxy model, and the lithium battery optimization design system 4 based on the proxy model includes a model building training module 41 and a prediction module 42. Model building training module 41 is configured to build and train a proxy model whose inputs include a plurality of lithium battery design variables and whose outputs include at least one lithium battery performance parameter; the prediction module 42 is configured to input design combination values of the plurality of lithium battery design variables to the proxy model to predict the obtained lithium battery performance parameters.
As shown in fig. 5, the present embodiment provides a lithium battery optimization design system based on a proxy model, where the lithium battery optimization design system 4 based on a proxy model includes a model building training module 41 and a prediction module 42, and further includes a cycle optimization module 43. Model building training module 41 is configured to build and train a proxy model whose inputs include a plurality of lithium battery design variables and whose outputs include at least one lithium battery performance parameter; the prediction module 42 is configured to input design combination values of the plurality of lithium battery design variables to the proxy model to predict the obtained lithium battery performance parameters;
the loop optimization module 43 is configured to compare the lithium battery performance parameter with the design requirement based on the prediction, adjust the design combination value of the plurality of lithium battery design variables, predict again using the proxy model, and loop until the design requirement is satisfied.
In the embodiments of fig. 4 and 5 described above, the proxy model constructed by the model construction training module 41 is a neural network-based proxy model, and includes a classification neural network for classifying the design combination values of the plurality of lithium battery design variables to output whether the design combination values of the plurality of lithium battery design variables are valid or not, and a fitting neural network for fitting the valid design combination values of the plurality of lithium battery design variables to output the lithium battery performance parameters. The training method of the classification neural network and the fitting neural network are respectively trained, and the training method is described in detail in some preferred embodiments, and is not described herein.
As shown in fig. 6, the present embodiment provides a lithium battery optimization design device based on a proxy model, and the lithium battery optimization design device 6 based on the proxy model includes: a memory 61 configured to store a computer program; and a processor 62 configured to invoke the computer program to perform the above-described proxy model-based lithium battery optimization design method.
Preferably, the memory 61 includes: various media capable of storing program codes, such as ROM, RAM, magnetic disk, U-disk, memory card, or optical disk.
Preferably, the processor 62 may be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), etc.; but also digital signal processors (Digital Signal Processor, DSP for short), application specific integrated circuits (Application Specific Integrated Circuit, ASIC for short), field programmable gate arrays (Field Programmable Gate Array, FPGA for short) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.
In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, or methods may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of modules/units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple modules or units may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or modules or units, which may be in electrical, mechanical or other forms.
The modules/units illustrated as separate components may or may not be physically separate, and components shown as modules/units may or may not be physical modules, i.e., may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules/units may be selected according to actual needs to achieve the purposes of the embodiments of the present application. For example, functional modules/units in various embodiments of the present application may be integrated into one processing module, or each module/unit may exist alone physically, or two or more modules/units may be integrated into one module/unit.
Those of ordinary skill would further appreciate that the elements and algorithm steps of the examples described in connection with the embodiments disclosed herein may be embodied in electronic hardware, in computer software, or in a combination of the two, and that the elements and steps of the examples have been generally described in terms of function in the foregoing description to clearly illustrate the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
Embodiments of the present application also provide a computer-readable storage medium. Those of ordinary skill in the art will appreciate that all or part of the steps in the method implementing the above embodiments may be implemented by a program to instruct a processor, where the program may be stored in a computer readable storage medium, where the storage medium is a non-transitory (non-transitory) medium, such as a random access memory, a read only memory, a flash memory, a hard disk, a solid state disk, a magnetic tape (magnetic tape), a floppy disk (floppy disk), an optical disk (optical disk), and any combination thereof. The storage media may be any available media that can be accessed by a computer or a data storage device such as a server, data center, or the like that contains an integration of one or more available media. The usable medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a digital video disc (digital video disc, DVD)), or a semiconductor medium (e.g., a Solid State Disk (SSD)), or the like.
Embodiments of the present application may also provide a computer program product comprising one or more computer instructions. When the computer instructions are loaded and executed on a computing device, the processes or functions described in accordance with the embodiments of the present application are produced in whole or in part. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from one website, computer, or data center to another website, computer, or data center by a wired (e.g., coaxial cable, fiber optic, digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.).
The computer program product is executed by a computer, which performs the method according to the preceding method embodiment. The computer program product may be a software installation package, which may be downloaded and executed on a computer in case the aforementioned method is required.
The descriptions of the processes or structures corresponding to the drawings have emphasis, and the descriptions of other processes or structures may be referred to for the parts of a certain process or structure that are not described in detail.
The foregoing embodiments are merely illustrative of the principles of the present application and their effectiveness, and are not intended to limit the application. Modifications and variations may be made to the above-described embodiments by those of ordinary skill in the art without departing from the spirit and scope of the present application. Accordingly, it is intended that all equivalent modifications and variations which may be accomplished by persons skilled in the art without departing from the spirit and technical spirit of the disclosure be covered by the claims of this application.
Claims (10)
1. The lithium battery optimal design method based on the proxy model is characterized by comprising the following steps of:
constructing and training a surrogate model, the surrogate model input comprising a plurality of lithium battery design variables, the surrogate model output comprising at least one lithium battery performance parameter;
and inputting design combination values of the plurality of lithium battery design variables into the proxy model to predict and obtain the lithium battery performance parameters.
2. The proxy model-based lithium battery optimization design method of claim 1, wherein the proxy model is a neural network-based proxy model.
3. The proxy model-based lithium battery optimal design method according to claim 2, wherein the proxy model comprises a classification neural network for classifying the design combination values of the plurality of lithium battery design variables to output whether the design combination values of the plurality of lithium battery design variables are valid or not, and a fitting neural network for fitting the valid design combination values of the plurality of lithium battery design variables to output the lithium battery performance parameter.
4. The proxy model based lithium battery optimization design method of claim 1, wherein the at least one lithium battery performance parameter comprises at least one of lithium battery energy and lithium battery power.
5. The proxy model-based lithium battery optimization design method of claim 1, further comprising: and comparing the performance parameters of the lithium battery with the design requirements based on the prediction, adjusting the design combination values of the design variables of the lithium battery, predicting again by adopting the agent model, and executing circularly until the design requirements are met.
6. The proxy model-based lithium battery optimization design method of claim 3, wherein training the proxy model comprises training the classification neural network and the fitting neural network separately.
7. A lithium battery optimization design system based on a proxy model, the system comprising:
a model building training module configured to build and train a proxy model, an input of the proxy model comprising a plurality of lithium battery design variables, an output of the proxy model comprising at least one lithium battery performance parameter;
a prediction module configured to input design combination values of the plurality of lithium battery design variables to the proxy model to predictively obtain the lithium battery performance parameters.
8. The proxy model-based lithium battery optimal design system of claim 7, further comprising a loop optimization module configured to compare the lithium battery performance parameter to a design requirement based on the prediction and adjust a design combination value of the plurality of lithium battery design variables and predict again using the proxy model, and to perform a loop until the design requirement is met.
9. An agent model-based lithium battery optimization design device, which is characterized by comprising:
a memory configured to store a computer program; and
a processor configured to invoke the computer program to perform the proxy model-based lithium battery optimization design method according to any one of claims 1 to 6.
10. A computer-readable storage medium having stored thereon a computer program, characterized in that the computer program is executed to implement the proxy model-based lithium battery optimization design method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310274472.5A CN116306288A (en) | 2023-03-20 | 2023-03-20 | Agent model-based lithium battery optimal design method, system, device and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310274472.5A CN116306288A (en) | 2023-03-20 | 2023-03-20 | Agent model-based lithium battery optimal design method, system, device and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116306288A true CN116306288A (en) | 2023-06-23 |
Family
ID=86832058
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310274472.5A Pending CN116306288A (en) | 2023-03-20 | 2023-03-20 | Agent model-based lithium battery optimal design method, system, device and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116306288A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117095775A (en) * | 2023-10-18 | 2023-11-21 | 江西五十铃汽车有限公司 | Solid-state lithium battery material design method, system, storage medium and computer |
| CN118070366A (en) * | 2024-04-25 | 2024-05-24 | 深圳市峰和数智科技有限公司 | Structure generation method, device, medium and equipment of porous electrode of new energy battery |
-
2023
- 2023-03-20 CN CN202310274472.5A patent/CN116306288A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117095775A (en) * | 2023-10-18 | 2023-11-21 | 江西五十铃汽车有限公司 | Solid-state lithium battery material design method, system, storage medium and computer |
| CN118070366A (en) * | 2024-04-25 | 2024-05-24 | 深圳市峰和数智科技有限公司 | Structure generation method, device, medium and equipment of porous electrode of new energy battery |
| CN118070366B (en) * | 2024-04-25 | 2024-07-05 | 深圳市峰和数智科技有限公司 | Structure generation method, device, medium and equipment of porous electrode of new energy battery |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116306288A (en) | Agent model-based lithium battery optimal design method, system, device and medium | |
| Liu et al. | Data‐Driven Hybrid Internal Temperature Estimation Approach for Battery Thermal Management | |
| He et al. | State of health estimation of lithium‐ion batteries: A multiscale G aussian process regression modeling approach | |
| Zhang et al. | Intelligent state of charge estimation of battery pack based on particle swarm optimization algorithm improved radical basis function neural network | |
| Tian et al. | 3-D temperature field reconstruction for a lithium-ion battery pack: A distributed Kalman filtering approach | |
| CN112180258B (en) | Method, device, medium, terminal and system for measuring average coulombic efficiency of battery | |
| CN113177290B (en) | Satellite component temperature field prediction method based on depth agent model normalization | |
| CN112666479A (en) | Battery service life prediction method based on charging cycle fusion | |
| CN117175664B (en) | Energy storage charging equipment output power self-adaptive adjusting system based on use scene | |
| Zahid et al. | Machine learning an alternate technique to estimate the state of charge of energy storage devices | |
| CN115828685B (en) | Lithium battery temperature simulation method, device and medium based on Chebyshev spectrum method | |
| Yao et al. | A multi-model feature fusion model for lithium-ion battery state of health prediction | |
| CN114462294B (en) | Two-stage proxy model auxiliary parameter estimation method and system | |
| Li et al. | A hybrid convolutional neural network-long short term memory for discharge capacity estimation of lithium-ion batteries | |
| US20140136169A1 (en) | Systems and methods for improving battery performance | |
| CN115840144A (en) | Battery simulation calculation method and device, computer equipment and storage medium | |
| Zhu et al. | A novel fast estimation and regroup method of retired lithium‐ion battery cells | |
| CN115656824A (en) | Lithium battery nuclear power state prediction method based on CNN-LSTM model | |
| CN114117737A (en) | Lithium ion battery simulation method and device, electronic equipment and readable storage medium | |
| Li et al. | Synthetic thermal convolutional‐memory network for the lithium‐ion battery behaviour diagnosis against noise interruptions | |
| Wang et al. | A conditional random field based feature learning framework for battery capacity prediction | |
| Yun et al. | An integrated framework for minimization of inter lithium‐ion cell temperature differences and the total volume of the cell of battery pack for electric vehicles | |
| Galvez‐Aranda et al. | Time‐Dependent Deep Learning Manufacturing Process Model for Battery Electrode Microstructure Prediction | |
| CN115951236B (en) | Lithium battery state monitoring method, system, device and storage medium | |
| Shen et al. | Simultaneous model selection and parameter estimation for lithium‐ion batteries: A sequential MINLP solution approach |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |