A power battery simulation method, device and medium based on AI parameter identification

Abstract

The present invention relates to the field of computer simulation technology, and in particular, to a power battery simulation method, device and medium based on AI parameter identification. The method comprises constructing a simulation model of a power battery; determining the primary parameters of the simulation model according to the type of the simulation model; determining the secondary parameters and tertiary parameters for correcting the primary parameters according to the application scenario of the simulation model; inputting the primary parameters into the simulation model to obtain the output value of the simulation model; and correcting the primary parameters according to the secondary parameters and tertiary parameters to minimize the deviation of the output value. The present invention identifies the input parameters of the simulation model to obtain a high-precision simulation model.

Description

Technical Field

[0001] The present invention relates to the field of computer simulation technology, and in particular to a power battery simulation method, device and medium based on AI parameter identification. Background Art

[0002] In the current development and optimization of power batteries, traditional modeling methods often face the problem of complex system structure and difficulty in obtaining accurate parameters. Power batteries are decomposed into independent objects according to disciplines and functions. Each object represents a specific part or functional unit in the power battery system, such as a cell, management system, thermal management system, etc. Each object has unique properties (such as capacity, internal resistance, temperature characteristics) and methods (such as charging and discharging process, heat conduction mechanism).

[0003] Through object-oriented modeling methods, these objects can interact and collaborate with each other to simulate the operating status of power batteries under different working conditions. Inheritance and polymorphism mechanisms allow objects to be flexibly expanded and modified, enhancing the flexibility and maintainability of the model. In the modeling process, the acquisition of power battery parameters is often a difficult point. Traditional methods are mainly based on Amesim, Simulink, Cruise and other platforms. Different development platforms have different support for power batteries. For example, Simulink requires its own modeling, which requires developers to have a certain technical foundation. Amesim only has a single equivalent and electrochemical model, and Cruise only has a simple equivalent circuit model. None of the above models have the ability to identify parameters.

[0004] In view of this, the present invention is proposed.

[0005] In the current power battery development and optimization process, traditional modeling methods mainly rely on specific platforms, such as Amesim, Simulink and Cruise. Although these platforms have their own characteristics, they have the following problems in the modeling process:

[0006] Difficulty in parameter acquisition: Traditional platforms usually require manual input of parameters, such as battery cell characteristics, battery thermal management parameters, etc. The process of acquiring these parameters is cumbersome and can easily lead to errors when experimental conditions are insufficient, ultimately affecting the accuracy of the simulation model.

[0007] Single model function: Amesim: Supports a single equivalent circuit and electrochemical model, lacks flexible multi-field coupling capabilities, and is difficult to meet complex thermal-electric coupling modeling needs. Simulink: Requires developers to have strong modeling capabilities. Since there is no built-in parameter identification function, users need to design their own algorithms to optimize the model, which increases the usage threshold. Cruise: Only supports basic equivalent circuit models, and its simplification level cannot meet the performance evaluation requirements of power batteries under extreme working conditions.

[0008] Lack of parameter identification capability: The power battery simulation of existing platforms is mainly based on static or empirical parameters, and cannot dynamically adjust the model input parameters to adapt to different application scenarios. For example, the input parameters of the equivalent circuit model and electrochemical model cannot be effectively corrected under complex working conditions such as extreme temperature and pressure, thereby reducing the simulation accuracy.

[0009] Insufficient scalability: Traditional platforms have limited performance in the construction of complex coupling models (such as thermal-electric-fluid-mechanical coupling models) and high-order models (such as aging models, thermal runaway models), and cannot flexibly handle the coupled behaviors of multi-field interactions in power battery systems. Summary of the invention

[0010] The purpose of the present invention is to provide a power battery simulation method, device and medium based on AI parameter identification, so as to identify the input parameters of the simulation model and obtain a high-precision simulation model.

[0011] In order to achieve the above object, the present invention adopts the following technical solutions:

[0012] In a first aspect, the present invention provides a power battery simulation method based on AI parameter identification, comprising:

[0013] Build a simulation model of the power battery;

[0014] Determining the primary parameters of the simulation model according to the type of the simulation model;

[0015] Determining secondary parameters and tertiary parameters for correcting the primary parameters according to an application scenario of the simulation model;

[0016] Inputting the primary parameters into the simulation model to obtain an output value of the simulation model;

[0017] The primary parameter is modified according to the secondary parameter and the tertiary parameter to minimize the deviation of the output value.

[0018] In a second aspect, the present invention provides an electronic device, comprising:

[0019] at least one processor, and a memory communicatively coupled to at least one of the processors;

[0020] Among them, the memory stores instructions that can be executed by at least one of the processors, and the instructions are executed by at least one of the processors so that at least one of the processors can execute the above-mentioned power battery simulation method based on AI parameter identification.

[0021] In a third aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, wherein the computer instructions are used to enable the computer to execute the above-mentioned power battery simulation method based on AI parameter identification.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] The present invention constructs a complete power battery electrical, thermal and other models in Modelica language, and identifies the input parameters with high precision to improve the model accuracy. The present invention starts with the construction of the simulation model, determines the primary parameters, and then corrects the parameters, forming a complete set of power battery simulation methods, and the modeling process is convenient.

[0024] The present invention uses artificial intelligence technology to automatically identify the primary parameters of the power battery. Through the training of a large amount of experimental data, the AI ​​model can automatically generate the required input parameters. This not only improves the efficiency of modeling, but also ensures the accuracy of the parameters. Through the above-mentioned modeling and parameter identification, designers can perform detailed performance evaluation and optimization of the power battery in a simulation environment. During the simulation process, the system can generate a large amount of data to help designers understand the pros and cons of different design schemes. These data can be used to discover potential problems and guide design improvements.

[0025] In summary, the object-oriented power battery simulation method based on AI parameter identification significantly improves the efficiency and accuracy of power battery modeling and simulation through the combination of object-oriented modeling ideas and artificial intelligence technology. This method not only plays an important role in the design and optimization process, but also provides strong support for patent applications and technological innovation, and promotes the development of power battery technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0027] Figure 1 It is a flow chart of a power battery simulation method based on AI parameter identification provided by the present invention;

[0028] Figure 2 is a schematic diagram of building a power battery simulation model provided by an embodiment of the present invention;

[0029] Figure 3 is a schematic diagram of the structure of a battery pack provided by an embodiment of the present invention;

[0030] Figure 4 is a schematic diagram of correcting primary parameters provided by an embodiment of the present invention;

[0031] Figure 5 It is a flow chart of a method for correcting a primary parameter provided by an embodiment of the present invention;

[0032] Figure 6 is a flow chart of another method for correcting primary parameters provided by an embodiment of the present invention;

[0033] Figure 7 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0034] The following is a description of exemplary embodiments of the present application in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0035] The present invention is further described in detail below with reference to the embodiments.

[0036] Example 1

[0037] Figure 1 This is a flow chart of a power battery simulation method based on AI parameter identification provided by the present invention. This embodiment is applicable to the case of identifying the input parameters of an object-oriented power battery model. The method can be executed by a power battery simulation device based on AI parameter identification, which can be composed of software and / or hardware and is generally integrated in an electronic device. Figure 1 As shown, this embodiment provides a power battery simulation method based on AI parameter identification, comprising the following steps:

[0038] S110. Construct a simulation model of a power battery.

[0039] First, the user selects a model required for modeling in the object-oriented modeling software. In response to the user's selection operation, the model required for power battery modeling is determined, including but not limited to battery connection type, electrical model, coupling model and high-order model.

[0040] Among them, battery connection types include series type and parallel type. Electrical models include equivalent circuit models and electrochemical models. Coupling models include electric-thermal coupling models and electric-thermal-mechanical coupling models. High-order models include aging models and thermal runaway models. The following table shows the models required for more detailed modeling.

[0041] Table 1 Models required to build a power battery

[0042]

[0043] The models in Table 1 are only examples, and those skilled in the art may use the models in Table 1 and / or models other than Table 1 to construct a simulation model of a power battery.

[0044] In Table 1, the Rint model equates the battery to an ideal voltage source Uoc (open circuit voltage) in series with the battery internal resistance R, and the open circuit voltage Uoc is a function of SOC and temperature. The Thevenin model takes into account the characteristics of the battery characteristics that are similar to those of the capacitor. In the model, the ideal voltage source Uoc describes the open circuit voltage of the battery, the resistor R is the battery internal resistance, and the capacitor CP and the resistor RP are connected in parallel to describe the overpotential of the battery. The PNGV model is the standard battery model in the 2001 "PNGV Battery Test Manual". The RC model consists of 2 capacitors and 3 resistors. The Dual Particle Model (DP model) is an extension of the Single Particle Model (SPM), which is mainly used to more accurately describe the different kinetic behaviors of the positive and negative electrodes in lithium-ion batteries.

[0045] The user can select the battery architecture (including the interface connection relationship between models) in the object-oriented modeling software, and automatically combine it with the above-extracted models. The interface is determined according to the type of the model, including but not limited to electrical interface, fluid interface, thermal interface, force interface and other software interfaces for joint simulation. Furthermore, the model interface is automatically connected in series and parallel to form a simulation model of the power battery.

[0046] For example, Figure 2 The specific implementation method of building a power battery simulation model is shown. First, determine the model required for power battery modeling, and judge whether the model selected by the user is a single battery, a battery module, or a battery pack, which determines the model structure. Then, determine the connection form between the selected model and the outside world, which can be electrical connection, thermal connection, fluid connection, force connection, or connection with other joint simulation software. The connection method between these models and the outside world determines the interface type. Figure 2Model 1 on the left is a battery pack or a single battery with two electrical interfaces (a positive interface and a negative interface); Model 2 is a battery module composed of multiple discrete batteries, in which the electrical interfaces are connected in series and parallel according to the battery connection type selected by the user. Figure 2 Only the electrical interface is set on the left side. Figure 2 Models 3 to 5 on the right have thermal interfaces and fluid inlet and outlet interfaces added according to the model type (e.g., electro-thermal coupling model, electro-thermal-fluid-mechanical coupling model). Figure 2 Model 5 on the far right also adds a computational fluid dynamics (CFD) joint simulation interface. Taking the electric-thermal-fluid coupling discrete model as an example, the simulation models of 6 power batteries (i.e. Figure 3 The six electrothermal coupling simulation models in the simulation are connected to form Figure 3 The module series-parallel electrically-thermally coupled battery pack model shown. This embodiment focuses on identifying the input parameters of any simulation model in the battery pack.

[0047] Exemplarily, in response to the user's selection operation on the simulation model, the input parameters of the simulation model are automatically identified. Of course, the input parameters of each simulation model can also be automatically identified in sequence.

[0048] S120. Determine primary parameters of the simulation model according to the type of the simulation model.

[0049] The types of simulation models include, but are not limited to: equivalent circuit model, electrochemical model, electro-thermal coupling model, high-order model, etc., as shown in Table 1.

[0050] Extract input parameters from the simulation model: primary parameters. Different types of simulation models have different primary parameters. If the simulation model is an equivalent circuit model, the primary parameters include at least open circuit voltage, battery internal resistance, charge transfer resistance and charge transfer capacitance. If the simulation model is an electrochemical model, the primary parameters include at least SEI membrane characteristic parameters, positive and negative electrode diffusion coefficients and electrochemical reaction kinetic characteristic parameters. If the simulation model is a thermoelectrically coupled electrochemical model, the primary parameters include at least SEI membrane characteristic parameters, positive and negative electrode diffusion coefficients, electrochemical reaction kinetic characteristic parameters, heat transfer coefficient, flow channel structure and fluid medium physical properties. If the life model of the power battery is activated, the primary parameters need to add calendar life empirical coefficients and cycle life empirical coefficients on the basis of the original primary parameters.

[0051] The initial values ​​of each primary parameter are mainly derived from user input. When the user input is missing, it will be supplemented according to the experimental data pre-stored in the database. Among them, experimental data of batteries of different types and capacities are pre-stored (such as HPPC experimental data, constant current charge and discharge experimental data, entropy thermal coefficient experimental data, temperature test experimental data, etc.). The missing primary parameters are parsed according to these pre-stored experimental data. For example, the internal resistance of the battery is obtained according to the current curve and voltage curve of the battery experiment of the same type and capacity. In summary, the primary parameters can be input by the user, directly recorded in the experiment, or obtained through experimental data.

[0052] S130. Determine secondary parameters and tertiary parameters for correcting primary parameters according to an application scenario of the simulation model.

[0053] If the application scenario is a normal temperature and pressure scenario, the primary parameters are not affected by extreme external conditions, and only the input parameters of other types of simulation models: secondary parameters are used to correct the primary parameters. Here, other types of simulation models should have the same capacity, battery type and application scenario as the simulation model constructed in this embodiment, and only the type of model (construction logic) is different. For the convenience of description and distinction, other types of simulation models different from the simulation model constructed in this embodiment are called reference models.

[0054] The secondary parameters are used to fit the reference values ​​of the primary parameters. For the convenience of description and distinction, the values ​​of the primary parameters obtained by fitting are called the reference values ​​of the primary parameters. Applicable to normal temperature and pressure. For example, if the simulation model is an equivalent circuit model, the secondary parameters at least include the SEI film characteristic parameters, the positive and negative electrode diffusion coefficients and the electrochemical reaction kinetic characteristic parameters in the electrochemical model; if the simulation model is an electrochemical model, the secondary parameters at least include the open circuit voltage, battery internal resistance, charge transfer resistance and charge transfer capacitance in the equivalent circuit model. If the simulation model is a thermoelectrically coupled electrochemical model, the secondary parameters at least include the equivalent circuit electrical characteristic parameters, the mechanical properties related to the battery structure and the heat generation coefficient of the equivalent resistance. The secondary parameters are subjected to physical and chemical operations to fit the reference values ​​of the primary parameters. The present invention does not limit the specific calculation formula, and the conventional parameter calculation formulas in the art can be referred to.

[0055] If the application scenario is an extreme scenario and / or includes a scenario in the non-electrical field, the primary parameters will be affected by extreme external conditions or non-electrical fields. Not only secondary parameters are required, but also input parameters of other types of simulation models: tertiary parameters are required to correct the primary parameters. Among them, extreme scenarios include high pressure, high temperature, and high cold scenarios, and non-electrical scenarios include scenarios of coupled fluid, force, heat, etc. Both secondary and tertiary parameters are used to fit the reference values ​​of the primary parameters; tertiary parameters are parameters affected by extreme scenarios or scenarios in the non-electrical field. For example, when constructing a series-single-particle electrochemical model, the electrochemical characteristic parameters such as solid phase diffusion, liquid phase diffusion, liquid phase migration, and electrochemical reaction of the power battery based on the single-particle model will be set as primary parameters, and secondary and tertiary parameters will be formed according to the equivalent circuit and other quasi-two-dimensional electrochemical parameters. For example, the secondary parameters include: characteristic parameters of the equivalent circuit (such as diffusion resistance, diffusion capacitance, etc.), and the tertiary parameters include correction coefficients under extreme conditions (such as electrochemical reaction dynamic correction coefficients, SEI film characteristic parameter correction coefficients, etc.), dual-particle model electrochemical characteristic parameters, uniform electrochemical characteristic parameters, etc. These correction factors can be obtained through experimental calibration.

[0056] Exemplarily, if the application scenario is an extreme scenario, and the simulation model is an equivalent circuit model, the three-level parameters include at least: correction coefficients of battery voltage, internal resistance, diffusion resistance and diffusion capacitance under extreme conditions. If the application scenario includes non-electrical scenarios, and the simulation model is an equivalent circuit model, the three-level parameters include at least: physical properties of battery structural materials, fluid physical properties and environmental impact factors under extreme conditions. If the application scenario is an extreme scenario, and the simulation model is an electrochemical model, the three-level parameters include at least: correction coefficients of electrochemical reaction kinetics, diffusion coefficients and SEI film characteristic parameters under extreme conditions. If the application scenario includes non-electrical scenarios, and the simulation model is an electrochemical model, the three-level parameters include at least: physical properties of battery structural materials, fluid physical properties and environmental impact factors under extreme conditions.

[0057] For example, the traditional SEI film internal resistance parameters are calculated based on the electrode material, electrolyte composition, capacity, temperature, and humidity. The present invention will additionally consider the battery simulation application scenario and capture additional parameters for further parameter fitting, such as charging density and voltage range, electrolyte additives, etc., and may further consider cycle life and storage time for fitting. This embodiment does not limit the fitting algorithm, and any fitting method that can obtain the parameters of another model through the parameters of one battery simulation model is within the scope of protection of the present invention.

[0058] In a specific embodiment, in the process of constructing the simulation model of the power battery, the model parameters can be collected in real time, and the secondary parameters and tertiary parameters can be obtained by fitting. Figure 4 As shown in the figure, after selecting the battery connection type (series or parallel), the primary parameters are collected; after selecting the electrical model (equivalent circuit model or electrochemical model), the primary parameters are collected, and the secondary and tertiary parameters are determined at the same time; after selecting the coupling model (electrothermal coupling or electrothermal-mechanical coupling), the primary parameters are collected, and the secondary and tertiary parameters are determined at the same time; after selecting the high-order model (aging model or thermal runaway model), the primary parameters are collected, and the secondary and tertiary parameters are determined at the same time. The secondary and tertiary parameters correct the primary parameters to obtain new primary parameters, and the primary parameters are re-entered into the simulation model for calculation.

[0059] S140, inputting the primary parameters into the simulation model to obtain the output value of the simulation model.

[0060] The primary parameters are input into the simulation model to obtain the output value of the simulation model. For example, if the simulation model is an equivalent circuit, the open circuit voltage and internal resistance value are input into the equivalent circuit to obtain the overall voltage value of the equivalent circuit.

[0061] The database pre-stores experimental data of batteries of the same type and capacity. If the output value deviates from the experimental data, for example, the difference between the output value and the experimental data is greater than a threshold, it is considered that the output value of the simulation model deviates, and the primary parameters need to be corrected, that is, the primary parameters are identified, thereby reducing the deviation and making the constructed model closer to the real battery.

[0062] S150, correcting the primary parameters according to the secondary parameters and the tertiary parameters to minimize the deviation of the output value.

[0063] In this step, the primary parameters can be corrected by a neural network model algorithm, as shown in the following two implementations. The primary parameters to be corrected can be determined by the user.

[0064] In the first implementation, the first-level parameters are modified by reinforcement learning of a neural network model (including an input layer, multiple convolutional layers, and a fully connected layer), see Figure 5As shown: After the simulation starts, the primary parameters are determined. The primary parameters are input into the simulation model to obtain the output value, and the output value is subtracted from the true value to obtain the deviation. If the deviation does not meet the accuracy requirements, the primary parameters (such as internal resistance), the reference value of the internal resistance obtained by fitting the secondary parameters and the tertiary parameters, and the deviation of the output value of the simulation model are combined into a three-dimensional vector and input into the neural network model. The neural network model outputs a weight k, which can be greater than 1 or less than 1. The weight k is combined with the primary parameter (such as internal resistance) to correct the primary parameter to obtain a new primary parameter. Among them, the primary parameter can be directly multiplied by the internal resistance, or other calculation methods can be used. Return to the operation of inputting the new primary parameter into the simulation model to obtain a new output value. Iterate the weight k in the neural network model and correct the primary parameter again until the deviation of the output value is less than the set value. It should be noted that the primary parameter is the corrected value and the reference value is a relatively more accurate value, which is directly related to the correction operation; the greater the deviation of the output value of the simulation model, the greater the weight should be, which is also related to the correction operation. Therefore, samples are selected in advance in the database, including: primary parameters, reference values, deviations, and weight labels (obtained by calculating the true value of the primary parameter obtained through experiments with the primary parameter in the sample, for example, dividing the true value of the primary parameter by the primary parameter in the sample). Multiple groups of samples are used to pre-train the neural network model so that the neural network model has the ability to obtain weights based on the primary parameters, reference values, and deviations.

[0065] The second implementation method uses a transfer learning algorithm of a neural network model (including an input layer, multiple convolutional layers, and a fully connected layer) to modify the primary parameters. Figure 6As shown: after the simulation starts, the primary parameters are determined and input into the simulation model; at the same time, the basic data of the reference model is determined according to the user input data or experimental data. The reference model here is the model from which the secondary parameters and the tertiary parameters are derived. For example, if the simulation model is an equivalent circuit model, the reference model is an electrochemical model. The basic data (including secondary parameters and tertiary parameters) are parameters used to input into the reference model, such as the positive and negative electrode diffusion coefficients and the electrochemical reaction kinetic characteristic parameters. This embodiment uses a neural network module to simulate the reference model, which is equivalent to using a neural network model to simulate the internal processing logic of the reference model. The basic data is used as the input of the neural network model, and the neural network simulates and outputs the reference value of the primary parameter. For example, the reference value of the internal resistance is obtained by calculating the SEI film characteristic parameters, the positive and negative electrode diffusion coefficients, etc. in the electrochemical model. The output value of the simulation model is subtracted from the true value to obtain the deviation. Determine whether the deviation meets the accuracy requirements. If not, update the parameters in the neural network model, such as the diffusion coefficient of the electrolyte. The larger the deviation value, the greater the weight of the reference value, and the greater the correction amplitude. The reference value of the primary parameter is used to correct the primary parameter. For example, a reference value and a primary parameter are weighted averaged to obtain a new primary parameter. The new primary parameter is input into the simulation model. This embodiment uses a transfer learning algorithm based on a neural network model to iterate the parameters in the neural network model, thereby indirectly updating the primary parameter until the deviation of the output value is less than the set value and meets the accuracy requirement.

[0066] In summary, in neural network models, the optimization of model parameters is usually adjusted through an incentive-weight mechanism. Specifically, when the deviation obtained by simulation exceeds the preset loss value or accuracy range, the weight will be reduced, and the neural network model will readjust the weight to start a new round of iterative optimization. If the error of the simulation result is within the preset accuracy range, the weight will increase, and the model will be further iterated and updated along the current weight direction, and finally meet the error between the output value and the real data within a certain proportion. It should be noted that although both reinforcement learning and transfer learning involve weight adjustment, they differ in the proportion and type of weights. In reinforcement learning, weight adjustment focuses more on learning from environmental feedback, while transfer learning focuses on transferring the knowledge of the existing model (i.e., the reference model) to the simulation model, and making adaptive adjustments based on the output deviation of the simulation model.

[0067] The embodiments of the present invention have the following technical effects:

[0068] The present invention constructs a complete power battery electrical, thermal and other models in Modelica language, and identifies the input parameters with high precision to improve the model accuracy. The present invention starts with the construction of the simulation model, determines the primary parameters, and then corrects the parameters, forming a complete set of power battery simulation methods, and the modeling process is convenient.

[0069] The present invention uses artificial intelligence technology to automatically identify the primary parameters of the power battery. Through the training of a large amount of experimental data, the AI ​​model can automatically generate the required input parameters. This not only improves the efficiency of modeling, but also ensures the accuracy of the parameters. Through the above-mentioned modeling and parameter identification, designers can perform detailed performance evaluation and optimization of the power battery in a simulation environment. During the simulation process, the system can generate a large amount of data to help designers understand the pros and cons of different design schemes. These data can be used to discover potential problems and guide design improvements.

[0070] In summary, the object-oriented power battery simulation method based on AI parameter identification significantly improves the efficiency and accuracy of power battery modeling and simulation through the combination of object-oriented modeling ideas and artificial intelligence technology. This method not only plays an important role in the design and optimization process, but also provides strong support for patent applications and technological innovation, and promotes the development of power battery technology.

[0071] On the basis of the above embodiment, the method further includes: if the life model of the power battery is activated, the database will automatically iterate the primary parameters according to the running time of the simulation model to improve the accuracy of the life model. Specifically, when the life model is not activated, the primary parameters such as capacity and internal resistance of the battery simulation model are fixed values; after the life model is activated, the primary parameters such as capacity are attenuated. Based on the experimental data in the database (these experimental data are affected by the battery life), the attenuated primary parameters are determined according to the running time of the simulation model. For example, the attenuated primary parameters are obtained by the empirical formula of the battery calendar life. Then, the coefficients in the empirical formula of the battery calendar life are corrected by the mathematical model. The mathematical model here can be a neural network model or other deep learning model, and the coefficients are iterated by the mathematical model to minimize the deviation of the output value of the simulation model. Taking the neural network model as an example, the model output is the weight M, and the input is the deviation value and the current coefficient. The new coefficient is obtained by combining the weight M and the current coefficient, and the empirical formula is updated, so as to obtain the new output value calculated by the simulation model and the life model, and the new deviation value relative to the true value is calculated. The new deviation value and the current coefficient are input into the model again for a new round of iteration. It should be noted that the neural network model needs to be pre-trained. The training samples include: coefficients, biases and weight labels. Multiple groups of samples are used to pre-train the neural network model so that the neural network model has the ability to obtain weights based on the current coefficients and biases.

[0072] Example 2

[0073] like Figure 7 As shown, this embodiment provides an electronic device, including:

[0074] at least one processor; and

[0075] a memory communicatively connected to at least one of the processors; wherein,

[0076] The memory stores instructions that can be executed by at least one of the processors, and the instructions are executed by at least one of the processors to enable at least one of the processors to execute the above-mentioned power battery simulation method based on AI parameter identification. At least one processor in the electronic device can execute the above-mentioned method, and thus has at least the same advantages as the above-mentioned method.

[0077] Optionally, the electronic device also includes interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are connected to each other using different buses and can be installed on a common mainboard or installed in other ways as needed. The processor can process instructions executed in the electronic device, including instructions stored in or on a memory to display graphical information of a GUI (Graphical User Interface) on an external input / output device (such as a display device coupled to an interface). In other embodiments, if necessary, multiple processors can be used together with multiple memories, and / or multiple buses can be used together with multiple memories. Similarly, multiple electronic devices can be connected (for example, as a server array, a group of blade servers, or a multi-processor system), and each device provides some necessary operations. Figure 7 A processor 301 is taken as an example.

[0078] The memory 302 is a computer-readable storage medium that can be used to store software programs, computer executable programs and modules, such as program instructions / modules corresponding to the power battery simulation method based on AI parameter identification in the embodiment of the present invention. The processor 301 executes various functional applications and data processing of the device by running the software programs, instructions and modules stored in the memory 302, that is, realizing the above-mentioned power battery simulation method based on AI parameter identification.

[0079] The memory 302 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and at least one application required for a function; the data storage area may store data created according to the use of the terminal, etc. In addition, the memory 302 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 302 may further include a memory remotely arranged relative to the processor 301, and these remote memories may be connected to the device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0080] The electronic device may further include: an input device 303 and an output device 304. The processor 301, the memory 302, the input device 303 and the output device 304 may be connected via a bus or other means. Figure 7 The example of connecting through bus is taken in the following.

[0081] The input device 303 can receive input digital or character information, and the output device 304 can include a display device, an auxiliary lighting device (e.g., LED), and a tactile feedback device (e.g., a vibration motor), etc. The display device can include, but is not limited to, a liquid crystal display (LCD), a light emitting diode (LED) display, and a plasma display. In some embodiments, the display device can be a touch screen.

[0082] Example 3

[0083] This embodiment provides a computer-readable storage medium, on which computer instructions are stored, and the computer instructions are used to enable the computer to execute the above method. The computer instructions on the computer-readable storage medium are used to enable the computer to execute the above method, and thus have at least the same advantages as the above method.

[0084] The medium in the present invention can adopt any combination of one or more computer-readable media. The medium can be a computer-readable signal medium or a computer-readable storage medium. The medium can be, for example, but not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples of the medium (non-exhaustive list) include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, the medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, a device or a device or used in combination with it.

[0085] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0086] The program code contained on the computer-readable medium can be transmitted using any appropriate medium, including but not limited to wireless, wire, optical cable, RF (Radio Frequency), etc., or any suitable combination of the above.

[0087] Computer program code for performing the operations of the present invention may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0088] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this application can be executed in parallel, sequentially or in different orders, as long as the expected results of the technical solution disclosed in this application can be achieved, and this document is not limited here.

[0089] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A power battery simulation method based on AI parameter identification, characterized in that: include: Build a simulation model of the power battery; Determining the primary parameters of the simulation model according to the type of the simulation model; If the simulation model is an equivalent circuit model, the primary parameters include at least open circuit voltage, battery internal resistance, charge transfer resistance and charge transfer capacitance; if the simulation model is an electrochemical model, the primary parameters include at least SEI film characteristic parameters, positive and negative electrode diffusion coefficients and electrochemical reaction kinetic characteristic parameters; If the simulation model is a thermoelectrically coupled electrochemical model, the primary parameters include at least SEI film characteristic parameters, positive and negative electrode diffusion coefficients, electrochemical reaction kinetic characteristic parameters, heat transfer coefficient, flow channel structure and fluid medium physical property parameters; Determining secondary parameters and tertiary parameters for correcting the primary parameters according to an application scenario of the simulation model; Inputting the primary parameters into the simulation model to obtain an output value of the simulation model; The primary parameter is modified according to the secondary parameter and the tertiary parameter to minimize the deviation of the output value.

2. The method according to claim 1, characterized in that Build a simulation model of the power battery, including: In response to a user's selected operation, a model required for power battery modeling is determined; the model includes a battery connection type, an electrical model, a coupling model, and a high-order model; Determine the interface according to the type of the model; the interface includes an electrical interface, a fluid interface, a thermal interface, a force interface and other software interfaces of joint simulation; The interfaces are connected to form a simulation model of a power battery.

3. The method according to claim 1, characterized in that According to the application scenario of the simulation model, determining the secondary parameters and tertiary parameters for correcting the primary parameters includes: If the application scenario is a normal temperature and pressure scenario, determine a secondary parameter for correcting the primary parameter; wherein the secondary parameter is used to fit a reference value of the primary parameter; If the application scenario is an extreme scenario and / or includes a scenario in a non-electrical field, determine the secondary parameters and tertiary parameters used to correct the primary parameters; wherein the secondary parameters and tertiary parameters are both used to fit the reference values ​​of the primary parameters; the tertiary parameters are parameters affected by extreme scenarios or scenarios in a non-electrical field.

4. The method according to claim 3, characterized in that Determining a secondary parameter for correcting the primary parameter includes: If the simulation model is an equivalent circuit model, the secondary parameters include at least SEI film characteristic parameters, positive and negative electrode diffusion coefficients and electrochemical reaction kinetic characteristic parameters in the electrochemical model; If the simulation model is an electrochemical model, the secondary parameters include at least the open circuit voltage, battery internal resistance, charge transfer resistance and charge transfer capacitance in the equivalent circuit model; If the simulation model is a thermoelectrically coupled electrochemical model, the secondary parameters include at least electrical characteristic parameters of the equivalent circuit, mechanical properties related to the battery structure, and heat generation coefficient of the equivalent resistor.

5. The method according to claim 4, characterized in that Determining a tertiary parameter for correcting the primary parameter includes: If the application scenario is an extreme scenario and the simulation model is an equivalent circuit model, the three-level parameters include at least: correction coefficients of battery voltage, internal resistance, diffusion resistance and diffusion capacitance under extreme conditions; If the application scenario includes a non-electrical scenario, and the simulation model is an equivalent circuit model, the three-level parameters include at least: physical parameters of battery structural materials under extreme conditions, physical parameters of fluids, and environmental impact factors; If the application scenario is an extreme scenario and the simulation model is an electrochemical model, the three-level parameters include at least: electrochemical reaction kinetic correction coefficient under extreme conditions, diffusion coefficient, and SEI film characteristic parameter correction coefficient; If the application scenario includes a non-electrical field scenario, and the simulation model is an electrochemical model, the three-level parameters include at least: physical properties of battery structural materials under extreme conditions, fluid physical properties and environmental impact factors.

6. The method according to claim 5, characterized in that The first-level parameter is modified according to the second-level parameter and the third-level parameter to minimize the deviation of the output value, including: The primary parameters are corrected according to the secondary parameters and the tertiary parameters through a neural network model or a transfer learning algorithm to minimize the deviation of the output value.

7. An electronic device, characterized in that: include: at least one processor, and a memory communicatively coupled to at least one of the processors; Wherein, the memory stores instructions that can be executed by at least one of the processors, and the instructions are executed by at least one of the processors so that at least one of the processors can execute the power battery simulation method based on AI parameter identification as described in any one of claims 1-6.

8. A computer-readable storage medium, characterized in that: The medium stores computer instructions, and the computer instructions are used to enable the computer to execute the power battery simulation method based on AI parameter identification according to any one of claims 1 to 6.

Images