Method and system for predicting health state of lithium ion battery
By establishing an electrochemical P2D model for lithium-ion batteries and combining aging parameters with machine learning models, the problems of slow speed and low accuracy in predicting the health status of lithium-ion batteries in existing technologies have been solved, achieving more efficient and accurate prediction of the health status of lithium-ion batteries.
Patent Information
- Application Number
- CN202511365024.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-23
- Publication Date
- 2025-12-26
AI Technical Summary
Existing lithium-ion battery aging models have low accuracy in predicting health status. Models based on electrochemical mechanisms are less accurate, while data-driven models are slow and susceptible to noise, making it difficult to accurately reflect the battery's health status.
An electrochemical P2D model for lithium-ion batteries was established, aging parameters and aging reactions were added, and the health status relationship of lithium-ion batteries was fitted by combining external and internal parameters through a machine learning model to predict the health status of lithium-ion batteries.
This method improves the speed and accuracy of predicting the health status of lithium-ion batteries. By coupling a high-precision electrochemical mechanism model of SEI growth and lithium plating side reactions, a battery life degradation dataset is obtained. Parameters during the battery cycle aging process are extracted as training data, achieving more efficient and accurate prediction.
Smart Images
Figure CN121208641A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of machine learning technology, specifically to a method and system for predicting the health status of lithium-ion batteries within the field of machine learning technology. Background Technology
[0002] With the rapid development of new energy and electronic technologies, lithium-ion batteries have been widely used in various fields, including new energy vehicles. However, during use, the capacity of lithium-ion batteries gradually decreases, and their lifespan is limited by many factors, the key ones being the growth of the solid electrolyte interface (SEI) film and lithium plating. When multiple aging mechanisms coexist, clarifying the degradation rate caused by different aging mechanisms is of paramount importance for battery management and use.
[0003] In related technologies, there are many models for lithium-ion battery aging, each with its own advantages and disadvantages. Models based on electrochemical mechanisms can accurately reflect changes in battery performance, but due to the complexity of the lithium-ion aging process itself, the reaction rate is too slow. Data-driven models can quickly reflect the battery's State of Health (SOH), but their accuracy is biased and they are easily affected by noisy data, resulting in significant errors in the prediction results. Summary of the Invention
[0004] The purpose of this invention is to provide a method and system for predicting the health status of lithium-ion batteries. The specific technical solution adopted is as follows: In a first aspect, embodiments of the present invention provide a method for predicting the health status of a lithium-ion battery, the method comprising: Establish an electrochemical P2D model for lithium-ion batteries; By adding aging parameters and aging reactions to the lithium-ion battery electrochemical P2D model, an updated P2D model is obtained. Based on the initial and stopping conditions of the updated P2D model, the cycle aging parameters and battery health status sample results are determined. Based on the cycle aging parameters and battery health status sample results, external and internal parameters are determined; wherein, the internal parameters include: negative electrode particle radius, negative electrode thickness, and separator porosity; the external parameters include: depth of discharge, initial cycle temperature, and discharge rate. A machine learning model is trained based on the external parameters and the internal parameters to fit the correspondence between the external parameters and the internal parameters and the health status of the lithium-ion battery, so as to predict the health status of the lithium-ion battery through the target machine learning model.
[0005] Secondly, embodiments of the present invention provide a system for predicting the health status of a lithium-ion battery, the system comprising: A module is created to build an electrochemical P2D model for lithium-ion batteries; An add module is used to add aging parameters and aging reactions to the lithium-ion battery electrochemical P2D model to obtain an updated P2D model. The first determining module is used to determine the cycle aging parameters and battery health status sample results based on the initial and stopping conditions of the updated P2D model. The second determining module is used to determine external and internal parameters based on the cycle aging parameters and battery health status sample results; wherein, the internal parameters include: negative electrode particle radius, negative electrode thickness, and separator porosity; the external parameters include: depth of discharge, initial cycle temperature, and discharge rate; The prediction module is used to train a machine learning model based on the external parameters and the internal parameters, fit the correspondence between the external parameters and the internal parameters and the health status of the lithium-ion battery, so as to predict the health status of the lithium-ion battery through the target machine learning model.
[0006] Thirdly, a computer program product is provided, comprising: computer program code, which, when run on a computer, causes the computer to perform the method described in the first aspect.
[0007] Fourthly, a computer-readable storage medium is provided that stores computer program code, which, when executed on a computer, causes the computer to perform the method described in the first aspect.
[0008] This invention offers the following advantages: By establishing a quasi-two-dimensional (P2D) electrochemical model of lithium-ion batteries, aging parameters and aging reactions are added to the P2D model to obtain an updated P2D model. Then, using the initial and stopping conditions of the updated P2D model, cycle aging parameters and battery health status sample results are determined. Based on these cycle aging parameters and battery health status sample results, external and internal parameters are determined. The internal parameters include: negative electrode particle radius, negative electrode thickness, and separator porosity. The external parameters include: depth of discharge, initial cycle temperature, and discharge rate. Finally, a machine learning model is trained using the external and internal parameters to fit the correspondence between the external and internal parameters and the health status of the lithium-ion battery, thereby predicting the health status of the lithium-ion battery through the target machine learning model. Thus, by using a high-precision electrochemical mechanism model based on the P2D model and coupling it with SEI growth and lithium plating side reactions, a battery life degradation dataset is obtained, and parameters during battery cycle aging are extracted as training data. Furthermore, machine learning training is performed using external and internal parameters to obtain a target machine learning model. This target machine learning model can then be used to predict the health status of lithium-ion batteries, thereby improving the speed and accuracy of prediction. Attached Figure Description
[0009] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0010] Figure 1 This is a schematic diagram illustrating the implementation process of a method for predicting the health status of a lithium-ion battery according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the SOH curve after 2000 cycles at 25 degrees Celsius (°C) with different rp_neg values, provided in an embodiment of the present invention. Figure 3 This is a schematic diagram of the SOH curves after 2000 cycles at 25 degrees Celsius (°C) and different L_neg values, provided in an embodiment of the present invention. Figure 4 This is a schematic diagram of the SOH curve after 2000 cycles of epsl_sep at 25 degrees Celsius (°C) provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of the L_neg simulation and prediction relationship curves after 2000 cycles at 25 degrees Celsius (°C) provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the SOH curve after 2000 cycles at 25 degrees Celsius (°C) and different C-rates, provided in an embodiment of the present invention. Figure 7 The different T provided in the embodiments of the present invention init A schematic diagram of the SOH curve after 2000 cycles; Figure 8 This is a schematic diagram of the SOH curve after 2000 cycles at 25 degrees Celsius (°C) with different DOD values, provided in an embodiment of the present invention. Figure 9 This is a schematic diagram of the relationship between DOD simulation and prediction curves after 2000 cycles at 25 degrees Celsius (°C) provided in an embodiment of the present invention; Figure 10 This is a schematic diagram of battery experimental data and predicted data curves at 25 degrees Celsius (°C) provided in an embodiment of the present invention; Figure 11 This is a schematic diagram of the composition structure of a lithium-ion battery health status prediction system provided in an embodiment of the present invention; Figure 12 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. Detailed Implementation
[0011] To further illustrate the technical methods adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and applications of a method for predicting the health status of a lithium-ion battery according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments may be combined from any suitable form.
[0012] In the description of the embodiments of the present invention, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of the present invention, "multiple" means two or more.
[0013] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting 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 that feature.
[0014] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.
[0015] The specific details of the method for predicting the health status of a lithium-ion battery provided by the present invention are described below with reference to the accompanying drawings. Please refer to... Figure 1 The diagram illustrates a flowchart of a method for predicting the health status of a lithium-ion battery according to an embodiment of the present invention. The method includes: 101. Establish an electrochemical P2D model for lithium-ion batteries.
[0016] Here, the battery is equivalent to a negative electrode, a separator, and a positive electrode. A one-dimensional geometric structure of the lithium battery is drawn, and a P2D model of the lithium-ion battery is built based on the lithium battery modeling control equations listed in Table 1. The lithium-ion battery electrochemical P2D model is a coupled model. The modeling theory of the lithium-ion battery electrochemical P2D model is based on the P2D model proposed by Newman, using lithium iron phosphate (LFP / graphite battery) as the object. The control equations are shown in Table 1, and the symbols and explanations are shown in Table 2. The geometric structure is a one-dimensional structure, divided into a negative electrode, a separator, and a positive electrode.
[0017] Table 1 Electrochemical governing equations
[0018] Table 2. Explanation of symbols for electrochemical and thermal model equations.
[0019] 102. Add aging parameters and aging reactions to the lithium-ion battery electrochemical P2D model to obtain an updated P2D model.
[0020] Here, in the lithium-ion battery electrochemical P2D model, aging mechanism, solid electrolyte interface growth aging side reaction and lithium plating aging side reaction are added to obtain an intermediate P2D model; and aging parameters are added to the intermediate P2D model to obtain the updated P2D model.
[0021] In the lithium-ion battery electrochemical P2D model, an aging mechanism is added, including: determining the negative electrode computational domain in the lithium-ion battery electrochemical P2D model; and adding an aging mechanism for solid electrolyte interface growth and a lithium plating aging mechanism to the negative electrode computational domain. The aging mechanism includes the solid electrolyte interface growth aging mechanism and the lithium plating aging mechanism. Solid electrolyte interface growth and lithium plating are behaviors constrained by diffusion kinetics; that is, solid electrolyte interface growth and lithium plating are behaviors simultaneously constrained by both kinetics and diffusion kinetics.
[0022] In some possible implementations, an aging mechanism for SEI growth and a lithium plating aging mechanism are added to the negative electrode computational domain in the electrochemical model. Furthermore, SEI growth is a behavior simultaneously limited by kinetics and diffusion, with an equilibrium potential of 0.4 V for the SEI growth reaction and an initial SEI film thickness of 5 nm. The SEI growth is analyzed by formulating equation (6) in Table 3, establishing a function of local current density and parameters related to SEI side reactions.
[0023] Furthermore, the aging caused by lithium plating is that the local negative electrode potential becomes negative relative to Li / Li+, and a side reaction of lithium metal deposition will occur on the graphite negative electrode. The calculation equation for the lithium plating reaction rate is shown in equation (12) in Table 3. The control equations for SEI growth and lithium plating aging model are shown in Table 3, and the symbols and explanations are shown in Table 4.
[0024] Table 3 Equations for SEI growth and aging models and lithium plating aging models
[0025] Table 4 Symbols and Explanations
[0026] 103. Based on the initial and stopping conditions of the updated P2D model, determine the cycle aging parameters and battery health status sample results.
[0027] The initial conditions include: an initial state of charge of 1, an initial voltage of 3.2 V to 3.65 V, and an electrolyte concentration of 800 mol / m3 to 1000 mol / m3. The stopping conditions include: a cutoff voltage of 2.3V to 2.5V and a cycle count of 500 to 2000.
[0028] In some possible implementations, while setting initial and stopping conditions, the operating conditions of the updated P2D model are also set, and cyclic aging is performed to obtain cyclic aging parameters. The operating conditions of the updated P2D model include charge and discharge conditions, which are cycled with charging, resting, discharging, and then resting (e.g., charging first, then resting, then charging again, and then resting again), and the battery charging and discharging time is equal to the resting time.
[0029] 104. Based on the cycle aging parameters and battery health status sample results, determine the external and internal parameters.
[0030] The internal parameters include: negative electrode particle radius, negative electrode thickness, and membrane porosity; the external parameters include: depth of discharge (DOD); and initial cycle temperature (T). init ); discharge rate (hereinafter referred to as C-rate).
[0031] Here, the aging process of the lithium-ion battery is set when different internal and external parameters are set; and the machine learning model is trained based on the external parameters, the internal parameters, and the aging process of the lithium-ion battery.
[0032] In some possible implementations, the internal parameters can be determined in the following ways: The radius of the negative electrode particles is measured using a scanning electron microscope or a transmission electron microscope. That is, the radius of the negative electrode particles (rp_neg) can be directly observed and measured using a scanning electron microscope (SEM) or a transmission electron microscope (TEM). The thickness of the negative electrode is obtained by X-ray computed tomography (CT), that is, the precise value of the thickness of the negative electrode (L_neg) can be obtained by X-ray computed tomography (CT). Microscopic images are acquired using an electron microscope or X-ray computed tomography (CT), and the porosity of the membrane is determined based on the microscopic images. That is, microscopic images can be acquired using an electron microscope or X-ray computed tomography (CT), and the membrane porosity (epsl_sep) can be analyzed using software.
[0033] 105. A machine learning model is trained based on the external parameters and the internal parameters to fit the correspondence between the external parameters and the internal parameters and the health status of the lithium-ion battery, so as to predict the health status of the lithium-ion battery through the target machine learning model.
[0034] Here, firstly, unsupervised learning methods in machine learning are used to exclude faulty data from the external and internal parameters, resulting in updated external and internal parameters. Secondly, the updated external and internal parameters are divided into training and testing datasets, with a portion of experimental data used as a validation dataset. Thirdly, the machine learning model is trained using the training dataset, tested using the testing dataset, and validated using the validation dataset. The machine learning module can be an LSTM-XGBoost hybrid model, a CNN-LSTM model, a PINN model, or an XLSTM-informer model. Finally, another portion of experimental data is used to refine the machine learning model, resulting in a target machine learning model for predicting the health status of the lithium-ion battery. For example, the training dataset is used to train the model for deep learning; after successful training, a test dataset is used for initial testing; and then experimental data is used to validate the model's accuracy and fitting speed. Using a large amount of experimental data to train the machine learning model allows for continuous model refinement, and by learning from a large amount of historical data, the model becomes more aligned with actual needs.
[0035] In some possible implementations, a coupled model that can predict SOH during the aging process of lithium-ion batteries is constructed, including the following steps: First, an electrochemical model of a lithium-ion battery is established. Based on the P2D modeling theory, a one-dimensional geometric structure is determined, including the electrode dimension and the particle dimension.
[0036] Secondly, a lumped thermal model is established, which simplifies the actual structure of the battery cell to a one-dimensional structure and couples it with the electrochemical model. The coupling mechanism between the electrochemical model and the thermal model is that the heat generated by the electrochemical model is used as a heat source to provide an interface to the thermal model, and the temperature is used as a variable input to the electrochemical model. This allows the heat-dependent parameters in the electrochemical model to be dynamically updated according to the temperature input, thus constructing a one-dimensional electrochemical-thermal coupled model and inputting the battery model parameters.
[0037] Furthermore, an aging mechanism was added to the electrochemical-thermal coupling model, introducing SEI growth aging side reactions and lithium plating aging side reactions, and adding aging parameters to the model. SEI growth is a behavior simultaneously limited by kinetics and diffusion; the equilibrium potential of the SEI growth reaction is 0.4V, and the initial SEI film thickness is 5nm. The SEI growth was analyzed using a formula, and the results were substituted into the analytical expression for further analysis. The aging induction caused by lithium plating is that the local negative electrode potential becomes negative relative to Li / Li+, leading to lithium metal deposition side reactions at the graphite negative electrode.
[0038] Finally, the initial and stopping conditions of the coupled model are set, the operating conditions are set and calculations are performed, the results are obtained and post-processed for analysis.
[0039] In the above process, after establishing the P2D electrochemical model of lithium-ion battery, the thermal model adopts a lumped thermal model, which is coupled with the P2D model to construct a one-dimensional electrochemical-thermal coupled model and assign parameters to the battery model.
[0040] An aging mechanism was added to the electrochemical-thermal coupling model, introducing SEI growth aging side reactions and lithium plating aging side reactions. Aging parameters were added to the model. Initial and stopping conditions were set for the electrochemical-thermal coupling model coupling SEI growth and active material loss aging mechanisms. Operating conditions were set and calculations were performed to obtain L_neg, rp_neg, and epsl_sep results, which were then post-processed and analyzed. The simulation data from the coupling model were organized and summarized, and faulty data were excluded using unsupervised learning methods in machine learning. The data was classified into training and testing datasets, and a portion of the experimental data was used as a validation dataset to verify the model's fitting speed and accuracy.
[0041] By introducing a machine learning model into PyCharm, the training dataset is used to train the model for deep learning. After successful training, a test dataset is used for initial testing, and then experimental data is used to verify the model's accuracy and fitting speed.
[0042] A large amount of experimental data is used to train the machine learning model, which is then continuously improved. By learning from a large amount of historical data, the model becomes more in line with actual needs.
[0043] The relationship between rp_neg and SOH in the simulation model is as follows: Figure 2 As shown, the variation law of SOH with rp_neg was established. The relationship between L_neg and SOH in the simulation model is as follows. Figure 3 As shown, the variation law of SOH with L_neg was established. The relationship between epsl_sep and SOH in the simulation model is as follows. Figure 4 As shown, the variation law of SOH with epsl_sep was established. The relationship between the three internal parameters and SOH in the simulation model is as follows. Figure 5 As shown, curve 501 represents the SOH simulation curve, and curve 502 represents the SOH prediction curve, which verifies the close relationship between the internal parameters and SOH, and also verifies the accuracy of the machine learning model in predicting the internal parameters.
[0044] In this embodiment of the invention, an electrochemical P2D model for lithium-ion batteries is established. Aging parameters and aging reactions are added to this P2D model to obtain an updated model. Then, using the initial and stopping conditions of the updated P2D model, cycle aging parameters and battery health status sample results are determined. Based on these cycle aging parameters and battery health status sample results, external and internal parameters are determined. The internal parameters include: negative electrode particle radius, negative electrode thickness, and separator porosity. The external parameters include: depth of discharge, initial cycle temperature, and discharge rate. Finally, a machine learning model is trained using the external and internal parameters to fit the correspondence between the external and internal parameters and the health status of the lithium-ion battery, thereby predicting the health status of the lithium-ion battery through the target machine learning model. Thus, by using a high-precision electrochemical mechanism model based on the P2D model and coupled with SEI growth and lithium plating side reactions, a battery life degradation dataset is obtained, and parameters during battery cycle aging are extracted as training data. Furthermore, machine learning training is performed using external and internal parameters to obtain a target machine learning model. This target machine learning model can then be used to predict the health status of lithium-ion batteries, thereby improving the speed and accuracy of prediction.
[0045] In some embodiments, the method for predicting the health status of a lithium-ion battery provided in this invention can be implemented through the following process: First, an electrochemical model of a lithium-ion battery is established.
[0046] The battery is equivalent to a negative electrode, a separator, and a positive electrode. The one-dimensional geometric structure of the lithium battery is drawn. Based on the lithium battery modeling control equations listed in Table 1, a P2D model of the lithium-ion battery is built.
[0047] Secondly, the thermal model adopts a lumped thermal model, coupled with the P2D model, to construct a one-dimensional electrochemical-thermal coupled model, and assigns parameters to the battery model.
[0048] Furthermore, an aging mechanism was added to the electrochemical-thermal coupling model, introducing SEI growth aging side reactions and lithium plating aging side reactions. Aging parameters were also added to the model.
[0049] Next, the initial and stopping conditions of the electrochemical-thermal coupling model that couples the SEI growth and active material loss aging mechanism were set, the operating conditions were set and calculations were performed to obtain DOD and T. init The C-rate results were analyzed and post-processed.
[0050] Finally, the data simulated by the coupled model were organized and summarized, and faulty data were eliminated using unsupervised learning methods in machine learning. The data was classified into training datasets and test datasets, and a portion of the experimental data was used as a validation dataset to verify the model's fitting speed and accuracy.
[0051] By introducing a machine learning model into PyCharm, the training dataset is used to train the model for deep learning. After successful training, a test dataset is used for initial testing, and then experimental data is used to verify the model's accuracy and fitting speed.
[0052] A large amount of experimental data is used to train the machine learning model, which is then continuously improved. By learning from a large amount of historical data, the model becomes more in line with actual needs.
[0053] The relationship between C-rate and SOH in the simulation model is as follows: Figure 6 As shown, the variation law of SOH with C-rate was established. In the simulation model, T... init The relationship between SOH and SOH is as follows Figure 7 As shown, the relationship between SOH and T was established. init The changing pattern. The relationship between DOD and SOH in the simulation model is shown in the figure below. Figure 8 Therefore, the variation law of SOH with DOD was established. For example... Figure 9As shown, the relationship between the three external parameters and SOH in the simulation model is illustrated. Curve 901 represents the simulation curve of the relationship between the three external parameters and SOH, and curve 902 represents the predicted curve of the relationship between the three external parameters and SOH. This verifies the close relationship between the external parameters and SOH, and also verifies the accuracy of the machine learning model in predicting the external parameters.
[0054] In some embodiments, the method for predicting the health status of a lithium-ion battery provided in this invention can be implemented through the following process: First, establish a lithium-ion battery aging cycle dataset. The battery aging data obtained from the experiment were divided into a training set, a validation set, and a test set, with proportions of 60%, 20%, and 20% respectively. The training set was used for adaptive training of the model, the validation set was used to verify whether the model conformed to the experimental rules, and the test set was used to test the existing experimental data.
[0055] Secondly, after the test is completed, the model will be used to predict new data from actual experiments in order to determine the SOH corresponding to each stage of the lithium-ion battery aging process. Furthermore, the operating conditions used in the experiment were: a temperature of 25℃, and multiples of 0.5C, 1C, and 2C.
[0056] Finally, experimental data under different working conditions were used to verify the accuracy of the established machine learning model.
[0057] like Figure 10 As shown, the data represent experimental and predicted data. Curve 1001 represents the experimental data for SOH, and curve 1002 represents the predicted data for SOH. Figure 10 It can be seen that the experimental data and the predicted data fit well.
[0058] Table 5 Battery Design Parameters
[0059] Table 6 Aging parameters of SEI and LAM models
[0060] In this embodiment of the invention, combining a machine learning model with an electrochemical model to predict the state of harmonics (SOH) of lithium-ion batteries fully leverages the advantages of both. The electrochemical model, based on the physicochemical principles of the battery, can describe in detail the internal reaction processes and mechanisms, providing a theoretical foundation for prediction. The machine learning model, trained on a large amount of historical data, can automatically learn complex patterns and regularities in the data, improving prediction efficiency. This combined approach can more comprehensively consider the battery's operating state and aging mechanisms, improving the accuracy, reliability, and efficiency of prediction. It also better adapts to different operating conditions and data conditions, exhibiting stronger robustness.
[0061] This invention provides a system for predicting the health status of lithium-ion batteries. Please refer to [link / reference]. Figure 11 This illustration shows a schematic diagram of the composition structure of a lithium-ion battery health state prediction system according to an embodiment of the present invention. The system 1100 includes: a building module 1101 for building an electrochemical P2D model of a lithium-ion battery; an adding module 1102 for adding aging parameters and aging reactions to the lithium-ion battery electrochemical P2D model to obtain an updated P2D model; a first determining module 1103 for determining cycle aging parameters and battery health state sample results based on the initial and stopping conditions of the updated P2D model; a second determining module 1104 for determining external and internal parameters based on the cycle aging parameters and battery health state sample results; wherein the internal parameters include: negative electrode particle radius, negative electrode thickness, and separator porosity; the external parameters include: depth of discharge, initial cycle temperature, and discharge rate; and a prediction module 1105 for training a machine learning model based on the external and internal parameters, fitting the correspondence between the external and internal parameters and the health state of the lithium-ion battery, so as to predict the health state of the lithium-ion battery through a target machine learning model.
[0062] In some possible implementations, the adding module 1102 is also used to add aging mechanism, solid electrolyte interface growth aging side reaction and lithium plating aging side reaction to the lithium-ion battery electrochemical P2D model to obtain an intermediate P2D model; and to add aging parameters to the intermediate P2D model to obtain the updated P2D model.
[0063] In some possible implementations, the adding module 1102 is further configured to determine the negative electrode calculation domain in the lithium-ion battery electrochemical P2D model; and to add an aging mechanism for solid electrolyte interface growth and a lithium plating aging mechanism to the negative electrode calculation domain; wherein the aging mechanism includes the aging mechanism for solid electrolyte interface growth and the lithium plating aging mechanism.
[0064] In some possible implementations, solid electrolyte interface growth and lithium plating are behaviors constrained by diffusion kinetics.
[0065] In some possible implementations, the second determining module 1104 is further configured to set different aging processes of the lithium-ion battery corresponding to the external parameters and internal parameters; and to train the machine learning model based on the aging processes of the external parameters, the internal parameters and the lithium-ion battery.
[0066] In some possible implementations, the second determining module 1104 is further configured to measure the radius of the negative electrode particles using a scanning electron microscope or a transmission electron microscope; obtain the thickness of the negative electrode using an X-ray computed tomography (CT) scan; acquire microscopic images using an electron microscope or an X-ray computed tomography (CT) scan; and determine the membrane porosity based on the microscopic images.
[0067] In some possible implementations, the initial conditions include: an initial state of charge of 1, an initial voltage of 3.2V-3.65V, and an electrolyte concentration of 800-1000 mol / m3; the stopping conditions include: a cutoff voltage of 2.3V-2.5V, and a cycle count of 500-2000.
[0068] In some possible implementations, the first determining module 1103 is further configured to set the operating conditions of the updated P2D model and perform cyclic aging to obtain the cyclic aging parameters; wherein the operating conditions of the updated P2D model include charging and discharging conditions, the charging and discharging conditions are cycled with charging, resting, discharging and then resting again, and the battery charging and discharging time is equal to the resting time.
[0069] In some possible implementations, the prediction module 1105 is further configured to employ unsupervised learning methods in machine learning to exclude fault data from the external parameters and internal parameters, thereby obtaining updated external parameters and updated internal parameters; divide the updated external parameters and updated internal parameters into training datasets and test datasets, and use a portion of the experimental data as a validation dataset; train the machine learning model using the training dataset, test the machine learning model using the test dataset after training, and validate the machine learning model using the validation dataset; and correct the machine learning model using another portion of the experimental data to obtain a target machine learning model for predicting the health status of the lithium-ion battery.
[0070] Optionally, the transmission medium can be a wired link (e.g., but not limited to, coaxial cable, optical fiber, and Digital Subscriber Line (DSL)) or a wireless link (e.g., but not limited to, Wireless Fidelity (WIFI), Bluetooth, and mobile device networks). It should be noted that the apparatus provided in the above embodiments is only illustrative of the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device can be divided into different functional modules to complete all or part of the functions described above. Furthermore, the method embodiments provided in the above embodiments belong to the same concept, and their specific implementation processes are detailed in the method embodiments, and will not be repeated here.
[0071] Figure 12 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. For example, as shown... Figure 12 As shown, the computer device 1200 includes: a memory 1201, a processor 1202, and a computer program 1203 stored in the memory 1201 and running on the processor 1202, wherein when the processor 1202 executes the computer program 1203, the computer device can execute any of the aforementioned methods for predicting the health status of lithium-ion batteries.
[0072] Furthermore, this embodiment of the invention also protects a system that may include a memory and a processor. The memory stores executable program code, and the processor is used to call and execute the executable program code to perform a method for predicting the health status of a lithium-ion battery provided in this embodiment of the invention. This embodiment can divide the system into functional modules based on the above method example. For example, each module can correspond to a specific function, or two or more functions can be integrated into a single processing module. The integrated module can be implemented in hardware. It should be noted that the module division in this embodiment is illustrative and only represents a logical functional division; other division methods may be used in actual implementation. It should also be noted that all relevant content of each step involved in the above method embodiment can be referenced to the functional description of the corresponding functional module, and will not be repeated here.
[0073] It should be understood that the apparatus provided in this embodiment is used to execute the above-described method for predicting the health status of a lithium-ion battery, and therefore can achieve the same effect as the above-described implementation method. When using integrated units, the system may include a processing module and a storage module. When the system is applied to a device, the processing module can be used to control and manage the device's operations. The storage module can be used to support the device in executing mutual program code, etc. The processing module may be a processor or a controller, which can implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this invention. The processor may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of Digital Signal Processing (DSP) and a microprocessor, etc., and the storage module may be a memory.
[0074] Furthermore, the apparatus provided in the embodiments of the present invention may specifically be a chip, component, or module. The chip may include a connected processor and a memory; wherein, the memory is used to store instructions, and when the processor calls and executes the instructions, the chip can execute the method for predicting the health status of a lithium-ion battery provided in the above embodiments. This embodiment also provides a computer-readable storage medium storing computer program code. When the computer program code is run on a computer, the computer executes the aforementioned method steps to implement the method for predicting the health status of a lithium-ion battery provided in the above embodiments.
[0075] This embodiment also provides a computer program product. When the computer program product is run on a computer, it causes the computer to execute the aforementioned related steps to realize the method for predicting the health status of a lithium-ion battery provided in the above embodiment. The device, computer-readable storage medium, computer program product, or chip provided in this embodiment are all used to execute the corresponding methods provided above. Therefore, the beneficial effects they achieve can be referred to the beneficial effects in the corresponding methods provided above, and will not be repeated here. Through the description of the above embodiments, those skilled in the art can understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In the embodiments provided by this invention, it should be understood that the disclosed device and method can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of modules or units is merely a logical functional division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other forms.
[0076] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multiple task processing and parallel processing are possible or may be advantageous. The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. The above content is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be covered within the protection scope of the present invention.
Claims
1. A method for predicting the health status of a lithium-ion battery, characterized in that, The method for predicting the health status of the lithium-ion battery includes: Establish an electrochemical P2D model for lithium-ion batteries; By adding aging parameters and aging reactions to the lithium-ion battery electrochemical P2D model, an updated P2D model is obtained. Based on the initial and stopping conditions of the updated P2D model, the cycle aging parameters and battery health status sample results are determined. Based on the cycle aging parameters and battery health status sample results, external and internal parameters are determined; wherein, the internal parameters include: negative electrode particle radius, negative electrode thickness, and separator porosity; the external parameters include: depth of discharge, initial cycle temperature, and discharge rate. A machine learning model is trained based on the external parameters and the internal parameters to fit the correspondence between the external parameters and the internal parameters and the health status of the lithium-ion battery, so as to predict the health status of the lithium-ion battery through the target machine learning model.
2. The method for predicting the health status of a lithium-ion battery according to claim 1, characterized in that, The process of adding aging parameters and aging reactions to the lithium-ion battery electrochemical P2D model to obtain an updated P2D model includes: In the aforementioned lithium-ion battery electrochemical P2D model, aging mechanism, solid electrolyte interface growth aging side reaction and lithium plating aging side reaction are added to obtain an intermediate P2D model; An aging parameter is added to the intermediate P2D model to obtain the updated P2D model.
3. The method for predicting the health status of a lithium-ion battery according to claim 1, characterized in that, In the aforementioned lithium-ion battery electrochemical P2D model, an aging mechanism is added, including: Determine the negative electrode computational domain in the aforementioned lithium-ion battery electrochemical P2D model; An aging mechanism for solid electrolyte interface growth and a lithium plating aging mechanism are added to the negative electrode computational domain; wherein the aging mechanism includes: the aging mechanism for solid electrolyte interface growth and the lithium plating aging mechanism.
4. The method for predicting the health status of a lithium-ion battery according to claim 3, characterized in that, Solid electrolyte interface growth and lithium deposition are behaviors constrained by diffusion kinetics.
5. The method for predicting the health status of a lithium-ion battery according to claim 1, characterized in that, The method further includes: The aging process of the lithium-ion battery is set according to different external and internal parameters; The machine learning model is trained based on the external parameters, the internal parameters, and the aging process of the lithium-ion battery.
6. The method for predicting the health status of a lithium-ion battery according to claim 1, characterized in that, The determination of internal parameters based on the cycle aging parameters and battery health status sample results includes: The radius of the negative electrode particles was measured using a scanning electron microscope or a transmission electron microscope. The thickness of the negative electrode was obtained using X-ray computed tomography. Microscopic images are acquired using an electron microscope or X-ray computed tomography (CT), and the porosity of the membrane is determined based on the microscopic images.
7. The method for predicting the health status of a lithium-ion battery according to claim 1, characterized in that, The initial conditions include: an initial state of charge of 1, an initial voltage of 3.2 V-3.65 V, and an electrolyte concentration of 800-1000 mol / m³. 3 ; The stopping conditions include: a cutoff voltage of 2.3V-2.5V and a cycle count of 500-2000.
8. The method for predicting the health status of a lithium-ion battery according to claim 1, characterized in that, The method further includes: The updated P2D model is set to the operating conditions and subjected to cyclic aging to obtain the cyclic aging parameters; wherein, the updated P2D model's operating conditions include charge and discharge conditions, the charge and discharge conditions are cycled with charging, resting, discharging, and resting again, and the battery charging and discharging time is equal to the resting time.
9. The method for predicting the health status of a lithium-ion battery according to claim 1, characterized in that, The step of training a machine learning model based on the external parameters and the internal parameters, fitting the correspondence between the external parameters and the internal parameters and the health status of the lithium-ion battery, and predicting the health status of the lithium-ion battery through the machine learning model, includes: By employing unsupervised learning methods in machine learning, faulty data in the external parameters and internal parameters are excluded to obtain updated external parameters and updated internal parameters; The updated external parameters and updated internal parameters are divided into training datasets and test datasets, and a portion of the experimental data is used as a validation dataset. The machine learning model is trained using the training dataset, tested using the test dataset, and validated using the validation dataset after training. The machine learning model was modified using another set of experimental data to obtain a target machine learning model for predicting the health status of the lithium-ion battery.
10. A system for predicting the health status of a lithium-ion battery, characterized in that, The health status prediction system for the lithium-ion battery includes: A module is created to build an electrochemical P2D model for lithium-ion batteries; An add module is used to add aging parameters and aging reactions to the lithium-ion battery electrochemical P2D model to obtain an updated P2D model. The first determining module is used to determine the cycle aging parameters and battery health status sample results based on the initial and stopping conditions of the updated P2D model. The second determining module is used to determine external and internal parameters based on the cycle aging parameters and battery health status sample results; wherein, the internal parameters include: negative electrode particle radius, negative electrode thickness, and separator porosity; the external parameters include: depth of discharge, initial cycle temperature, and discharge rate; The prediction module is used to train a machine learning model based on the external parameters and the internal parameters, fit the correspondence between the external parameters and the internal parameters and the health status of the lithium-ion battery, so as to predict the health status of the lithium-ion battery through the target machine learning model.