Parameter identification method and device, computer equipment, storage medium and program product

Abstract

The invention relates to a parameter identification method and device, computer equipment, a storage medium and a program product. The method comprises the following steps: acquiring a data sample of a battery on a vehicle, and identifying parameters of a battery model of the battery based on an evaluation model and the data sample to obtain a parameter identification result. The evaluation model for evaluating the difference degree between the prediction data sequence and the real output data sequence is constructed based on the data sample meeting the preset condition, so that the obtained evaluation model can still obtain a more stable parameter identification result under the influence of the noise signal, and the robustness of the offline parameter identification method is improved.

Description

Technical Field

[0001] This application relates to the technical field of battery model determination, and particularly to a parameter identification method, device, computer device, storage medium, and program product. Background Art

[0002] Battery models can estimate the state of charge, state of health, and other performance of batteries. Currently, researchers have combined parameter identification methods to identify the parameters of battery models and obtained parameter identification results, and the stability of battery models is closely related to the parameter identification results.

[0003] Currently, in scenarios with low real-time requirements, an offline parameter identification method is usually used to identify the parameters of battery models to obtain parameter identification results. However, the stability of the parameter identification results obtained by current parameter identification algorithms is poor. Summary of the Invention

[0004] Based on this, it is necessary to provide a parameter identification method, device, computer device, storage medium, and program product that can improve the stability of parameter identification results for the above technical problems.

[0005] In a first aspect, this application provides a parameter identification method. The method includes:

[0006] Obtain data samples of the battery on the vehicle; the data samples include an input data sequence of the battery within a preset time period and a corresponding true output data sequence of the input data sequence, and the noise signal of the input data sequence satisfies a preset condition;

[0007] Based on the evaluation model and the data samples, identify the parameters of the battery model of the battery to obtain parameter identification results; the evaluation model is a model constructed based on the battery model of the battery and the data samples, and the predicted data sequence is obtained by inputting the input data sequence into the battery model.

[0008] By obtaining data samples including the input data sequence of the battery within a preset time period and the corresponding true output data sequence of the input data sequence, and the noise signal in the input data sequence satisfies a preset condition, an evaluation model for evaluating the difference degree between the predicted data sequence and the true output data sequence is constructed based on the data samples that meet the preset conditions, so that the obtained evaluation model can still obtain more stable parameter identification results under the influence of the noise signal, and the robustness of the offline parameter identification method is improved.

[0009] In one embodiment, based on the evaluation model and the data samples, identifying the parameters of the battery model of the battery to obtain parameter identification results includes:

[0010] Construct a distributed robust optimization model based on an evaluation model and data samples; the distributed robust optimization model is used to characterize the expected error of the evaluation results of each data sample obtained based on the evaluation model; each data sample is a sample obtained under the same working condition;

[0011] According to the distributed robust optimization model, identify the parameters of the battery model of the battery to obtain a parameter identification result.

[0012] The method provided by the embodiments of the present application uses a distributed robust optimization model for characterizing the expected error of the evaluation results of each data sample obtained based on the evaluation model to identify the parameters of the battery model of the battery, thereby further improving the stability of the parameter identification result.

[0013] In one embodiment, according to the distributed robust optimization model, identifying the parameters of the battery model of the battery to obtain a parameter identification result includes:

[0014] According to the first parameters of the battery model obtained after the previous iteration of the battery model, use the distributed robust optimization model to determine the noise signal of the current iteration from the noise signals of the input data sequence;

[0015] Based on the noise signal of the current iteration, use the distributed robust optimization model to determine the second parameters of the battery model obtained after the current iteration;

[0016] Based on the number of iterations and the preset number of iterations, identify the parameters of the battery model of the battery to obtain a parameter identification result.

[0017] The method provided by the embodiments of the present application uses the distributed robust optimization model to iterate the battery model to obtain a parameter identification result, so that on the one hand, it finds the noise that makes the distributed robust optimization model perform the worst, and on the other hand, under the action of the worst-case noise, it finds the parameters that make the distributed robust optimization model have the highest accuracy, improving the stability of the obtained parameter identification result.

[0018] In one embodiment, based on the number of iterations and the preset number of iterations, identifying the parameters of the battery model of the battery to obtain a parameter identification result includes:

[0019] When the number of iterations reaches the preset number of iterations, use the second parameters of the battery model obtained after the most recent iteration as the parameter identification result.

[0020] The method provided by the embodiments of the present application, when the number of iterations reaches the preset number of iterations, uses the second parameters of the battery model obtained after the most recent iteration as the parameter identification result, further improving the stability of the obtained parameter identification result.

[0021] In one embodiment, the method further includes:

[0022] Performing equivalent processing on the battery to obtain an equivalent circuit of the battery, and constructing a battery model based on the equivalent circuit;

[0023] Constructing an evaluation model based on the battery model and the data sample.

[0024] The method provided by the embodiments of the present application constructs an evaluation model based on the battery model and the data sample, thereby laying a foundation for determining the parameter identification result based on the evaluation model, and further improving the stability of the obtained parameter identification result.

[0025] In one embodiment, constructing an evaluation model based on the battery model and the data sample includes:

[0026] Inputting the input data sequence in the data sample into the battery model to obtain a predicted data sequence;

[0027] Determining the difference between the predicted data sequence and the true output data sequence;

[0028] Constructing an evaluation model according to the difference.

[0029] The method provided by the embodiments of the present application constructs an evaluation model based on the difference between the predicted data sequence and the true output data sequence, thereby laying a foundation for determining the parameter identification result based on the evaluation model, and further improving the stability of the obtained parameter identification result.

[0030] In one embodiment, the noise signal of the input data sequence satisfying the preset condition includes that the noise signal of the input data sequence is continuous noise, and / or the signal-to-noise ratio of the input data sequence is greater than the preset signal-to-noise ratio.

[0031] In the embodiments of the present application, by constraining the noise signal of the input data sequence to be continuous noise, and / or the signal-to-noise ratio of the input data sequence is greater than the preset signal-to-noise ratio, an evaluation model is constructed based on the data sample including the noise signal satisfying the preset condition, so that the stability of the parameter identification result obtained based on the evaluation model can be improved under any noise condition.

[0032] In a second aspect, the present application further provides a parameter identification device. The device includes:

[0033] An acquisition module, configured to acquire a data sample of the battery on the vehicle; the data sample includes an input data sequence of the battery within a preset time period and a true output data sequence corresponding to the input data sequence, and the noise signal of the input data sequence satisfies the preset condition;

[0034] An identification module, configured to identify parameters of a battery model of a battery based on an evaluation model and data samples to obtain a parameter identification result; the evaluation model is a model constructed based on the battery model of the battery and data samples, and the evaluation model is used to evaluate the difference degree between a predicted data sequence and an actual output data sequence, where the predicted data sequence is obtained by inputting an input data sequence into the battery model.

[0035] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of the method provided in the above embodiments are implemented.

[0036] In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, and when the computer program is executed by a processor, the steps of the method provided in the above embodiments are implemented.

[0037] In a fifth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the steps of the method provided in the above embodiments are implemented.

[0038] The above description is only an overview of the technical solutions of the present application. In order to be able to understand the technical means of the present application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present application more obvious and understandable, the following specifically illustrates the specific embodiments of the present application. Description of the Drawings

[0039] By reading the detailed description of the preferred embodiments below, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present application. And in all the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0040] Figure 1 is the internal structure diagram of a computer device in an embodiment;

[0041] Figure 2 is a schematic flowchart of a parameter identification method provided by an embodiment of the present application;

[0042] Figure 3 is a schematic flowchart of another parameter identification method provided by an embodiment of the present application;

[0043] Figure 4 is a schematic flowchart of yet another parameter identification method provided by an embodiment of the present application;

[0044] Figure 5It is a schematic flowchart of the evaluation model construction method provided by the embodiments of the present application;

[0045] Figure 6 It is a schematic structural diagram of an equivalent circuit provided by the embodiments of the present application;

[0046] Figure 7 It is a schematic flowchart of the evaluation model construction method provided by the embodiments of the present application;

[0047] Figure 8 It is one of the schematic structural diagrams of a parameter identification device provided by the embodiments of the present application;

[0048] Figure 9 It is another schematic structural diagram of a parameter identification device provided by the embodiments of the present application;

[0049] Figure 10 It is yet another schematic structural diagram of a parameter identification device provided by the embodiments of the present application;

[0050] Figure 11 It is the fourth schematic structural diagram of a parameter identification device provided by the embodiments of the present application. Detailed implementation manners

[0051] The embodiments of the technical solutions of the present application will be described in detail below with reference to the accompanying drawings. The following embodiments are only used to illustrate the technical solutions of the present application more clearly, and therefore are only examples and cannot be used to limit the protection scope of the present application.

[0052] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above accompanying drawings are intended to cover non-exclusive inclusion.

[0053] In the description of the embodiments of the present application, technical terms such as "first" and "second" are only used to distinguish different objects and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity, specific order or primary-secondary relationship of the indicated technical features. In the description of the embodiments of the present application, "a plurality of" means more than two unless otherwise specifically defined.

[0054] References to "embodiments" in this specification mean that specific features, structures, or characteristics described in connection with the embodiments can be included in at least one embodiment of the present application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive of other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0055] In the description of the embodiments of the present application, the term "and / or" is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this text generally represents an "or" relationship between the associated objects before and after.

[0056] In the description of the embodiments of the present application, the term "plurality" refers to two or more (including two). Similarly, "multiple groups" refers to two or more groups (including two groups), and "multiple pieces" refers to two or more pieces (including two pieces).

[0057] The battery model can estimate the state of charge, state of health and other performance of the battery. Currently, researchers have combined parameter identification methods to identify the parameters of the battery model and obtained parameter identification results, and the stability of the battery model is closely related to the parameter identification results.

[0058] Currently, in scenarios with low real-time requirements, an offline parameter identification method is usually used to identify the parameters of the battery model to obtain parameter identification results.

[0059] However, for a given battery model, when there is a noise signal in the real data input signal, if the noise signal is small or there is continuous noise information, the parameter identification results obtained by identifying the parameters of the battery model may change greatly. Therefore, the stability of the parameter identification results obtained by the current parameter identification algorithm is poor.

[0060] To solve the above technical problems, in an embodiment of the present application, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as Figure 1As shown. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner, and the wireless manner can be achieved through WIFI, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a parameter identification method. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or a button, trackball, or touchpad provided on the housing of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0061] Those skilled in the art can understand that Figure 1 the structure shown in [figure reference] is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0062] In one embodiment, as Figure 2 shown, Figure 2 is a schematic flowchart of a parameter identification method provided by an embodiment of this application. This method is applied to the computer device shown in Figure 1 and includes the following steps S201 - S202:

[0063] S201, obtain data samples of the battery on the vehicle; the data samples include an input data sequence of the battery within a preset time period and a corresponding true output data sequence of the input data sequence, and the noise signal of the input data sequence satisfies a preset condition.

[0064] Among them, one data sample may include an input data sequence x of the battery within a preset time period and a corresponding true output data sequence y of the input data sequence. The preset time period can be a continuous time period or a discontinuous time period. Exemplarily, taking the preset time period as a continuous one hour as an example, then the true input data sequence x of the battery obtained within one hour and the corresponding true output data sequence y of this input data sequence are used as one data sample. The input data sequence can be a current sequence. If the input data sequence of the battery obtained within one hour includes 600 current data, then these 600 current data form an input data sequence.

[0065] The input data sequence in a data sample corresponds to an actual vehicle observation data sequence, and the actual vehicle observation data sequence refers to the true output data sequence y. Exemplarily, the actual vehicle observation data sequence can be a voltage data sequence. The actual vehicle observation data sequence can be used as the label in a data sample.

[0066] In the embodiments of the present application, a data sample of the battery on the vehicle can be obtained, that is, one data sample forms a sample set. Multiple data samples can also be obtained, that is, multiple data samples form a sample set. The multiple data samples can be multiple samples obtained under the same working conditions, and the same working conditions mean that the temperature of the same battery is the same and the degree of battery aging is the same.

[0067] If the noise signal of the input data sequence x is denoted as ξ, the noise signal ξ of the input data sequence x satisfying the preset condition can mean that the noise is within a smaller ∈-ball, and / or the noise signal is a continuous noise signal. Among them, the noise being within a smaller ∈-ball means that the noise signal ξ satisfies It can be a certain measure, such as the 1-norm or 2-norm or Wasserstein distance, etc., that is, the noise signal ξ satisfies For a continuous input data sequence x, the pattern of the noise signal ξ includes but is not limited to discontinuous noise, and a noise sequence is superimposed on a continuous time series segment.

[0068] Traditional offline parameter identification methods can often only process a few frames of significantly abnormal noise that appears in the real data signal sequence, but are powerless for continuous noise or minute noise. Unfortunately, in the data signal sequence in practical applications, such noise is ubiquitous and its distribution is uncontrollable, and such noise may greatly affect the parameter identification result. The stability of the parameter identification result obtained by the existing parameter identification algorithms is poor. In the embodiments of the present application, in order to improve the stability of the parameter identification result, by obtaining data samples in which the noise signal ξ satisfies the preset condition, and then through the data samples satisfying the preset condition and the evaluation model, the parameters of the battery model of the battery are identified to obtain the parameter identification result, improving the stability of the parameter identification result.

[0069] S202. Based on the evaluation model and the data sample, identify the parameters of the battery model of the battery to obtain the parameter identification result.

[0070] Among them, the evaluation model is a model constructed based on the battery model of the battery and the data sample. The evaluation model is used to evaluate the difference degree between the predicted data sequence and the true output data sequence. The predicted data sequence is obtained by inputting the input data sequence into the battery model. The battery model can include but is not limited to the electrochemical model, equivalent circuit model, and data-driven model of the battery.

[0071] Suppose the established battery model is represented by f(θ; x), where θ includes identification parameters such as resistance and capacitance, and x is the current data sequence of the battery on the vehicle. An evaluation model can be constructed based on the battery model and data samples. After inputting the input data sequence into the battery model f(θ; x), a predicted data sequence can be obtained. Among them, the initial parameters of θ can be set in advance. Input the input data sequence into the battery model f(θ; x) with the initial parameters set, and the battery model f(θ; x) can obtain the predicted data sequence corresponding to the input data sequence.

[0072] In the embodiments of the present application, the evaluation model can be defined by the mean square error (MSE), root mean square error (RMSE), mean absolute deviation (MAE), etc. of the data samples. Taking MSE as an example, the defined evaluation model l can be expressed by the following formula (1):

[0073]

[0074] Among them, M represents the number of data in an input data sequence. When the input data sequence is the charging current sequence under the charging condition, M represents the number of charging current data in the charging current sequence. When the input data sequence is the discharging current sequence under the discharging condition, M represents the number of discharging current data in the discharging current sequence.

[0075] In a possible implementation manner, the parameters of the battery model of the battery can be identified based on the evaluation model constructed by MSE and the data samples to obtain the parameter identification result.

[0076] In another possible implementation manner, the parameters of the battery model of the battery can be identified based on the evaluation model constructed by RMSE and the data samples to obtain the parameter identification result.

[0077] The method provided by the embodiments of the present application realizes constructing an evaluation model for evaluating the difference degree between the predicted data sequence and the true output data sequence based on the data samples that meet the preset conditions by obtaining the data samples including the input data sequence of the battery within the preset time period and the true output data sequence corresponding to the input data sequence, and the noise signal in the input data sequence meets the preset conditions, so that the obtained evaluation model can still obtain a more stable parameter identification result under the influence of the noise signal, and improve the robustness of the offline parameter identification method.

[0078] Refer to Figure 3 , Figure 3It is a schematic flowchart of another parameter identification method provided by an embodiment of the present application. This embodiment relates to a possible implementation manner of identifying the parameters of a battery model based on an evaluation model and data samples to obtain a parameter identification result. Based on the above embodiment, the above S202 may include the following steps S301-S302:

[0079] S301, constructing a distributed robust optimization model based on the evaluation model and data samples; the distributed robust optimization model is used to characterize the expected error of the evaluation results of each data sample obtained based on the evaluation model; each data sample is a sample obtained under the same working condition.

[0080] In the case where the number of data samples is multiple, a distributed robust optimization model can be constructed based on the evaluation model and each data sample. Taking the MSE model as an example of the evaluation model, the constructed distributed robust optimization model can refer to the following formula (2). Among them, the distributed robust optimization model is the DRO model, and the full English name of DRO is Distributed robust optimization. The DRO model is shown in the following formula (2) for example:

[0081]

[0082] Formula (2) is a two-layer optimization problem. Its inner-layer optimization can be understood as finding a noise within a range such that the expected error of the battery model with a given parameter θ in the sample set is the largest; the outer-layer optimization problem then considers that under the condition of noise, finding the optimal parameter θ such that the expected error of the DRO model in the sample set is the smallest. The whole problem is a game problem, which can be understood as, on the one hand, finding the noise that makes the DRO model perform the worst, and on the other hand, under the action of the worst-performing noise, finding the parameter θ that makes the DRO model have the highest accuracy.

[0083] The distributed robust optimization model is used to characterize the expected error of the evaluation results of each data sample obtained based on the evaluation model. The evaluation result of a data sample refers to the mean square error obtained by the evaluation model based on a data sample, and the expected error refers to the average value of the mean square errors of multiple data samples.

[0084] Since the purpose of the parameter identification method provided by the embodiment of the present application is to obtain a more stable parameter identification result, therefore, in order to obtain a more stable parameter identification result, the parameter identification results obtained under the same working condition should be the same in theory. Therefore, in the case of constructing a distributed robust optimization model using multiple data samples, the working conditions of each data sample need to be the same, so as to further improve the stability of the parameter identification results obtained under the same working condition.

[0085] S302. Identify the parameters of the battery model of the battery according to the distributionally robust optimization model to obtain the parameter identification result.

[0086] Formula (2) is a two-layer optimization problem. Its inner-layer optimization can be understood as finding a noise within a range such that the expected error of the battery model with a given parameter θ in the sample set is maximized; the outer-layer optimization problem then considers finding the optimal parameter θ such that the expected error of the model in the sample set is minimized when the model is under noise. The whole problem is a game problem, which can be understood as, on the one hand, finding the noise that makes the model perform worst, and on the other hand, finding the parameter θ that makes the model have the highest accuracy under the action of the worst-case noise. That is, through the adversarial training between the inner layer and the outer layer, the parameter identification result is obtained.

[0087] The distributionally robust optimization model can be solved by using, but not limited to, the Cutting Plane Method (CPM), the Primal Dual Method (PDM), and the "alternating optimization" method to obtain the parameter identification result.

[0088] The method provided in the embodiment of the present application identifies the parameters of the battery model of the battery according to the distributionally robust optimization model for characterizing the expected error of the evaluation results of each data sample obtained based on the evaluation model to obtain the parameter identification result, thereby further improving the stability of the parameter identification result.

[0089] Refer to Figure 4 , Figure 4 FIG. is a schematic flowchart of another parameter identification method provided in the embodiment of the present application. This embodiment relates to a possible implementation manner of how to identify the parameters of the battery model of the battery according to the distributionally robust optimization model to obtain the parameter identification result. On the basis of the above embodiment, the above S302 may include the following steps S401-S403:

[0090] S401. Determine the noise signal of the current iteration from the noise signals of the input data sequence by using the distributionally robust optimization model according to the first parameters of the battery model obtained after the previous iteration of the battery model.

[0091] This embodiment relates to an example of solving a distributed robust optimization model using the "alternating optimization" method. Assume that the parameter θ of the battery model is set to θ0. During the first iteration of the battery model, θ0 can be fixed, and the noise signal 1 under the given parameter θ0 is determined using the distributed robust optimization model. Given the noise signal 1, θ1 is determined using the distributed robust optimization model. During the second iteration, θ1 is fixed first, the noise signal 2 under the given parameter θ1 is determined using the distributed robust optimization model, and given the noise signal 2, θ2 is determined using the distributed robust optimization model. During the third iteration, θ2 is fixed first, the noise signal 3 under the given parameter θ2 is determined using the distributed robust optimization model, and given the noise signal 3, θ3 is determined using the distributed robust optimization model. And so on, the battery model is iterated multiple times.

[0092] Among them, the first parameter of the battery model obtained after the previous iteration refers to the parameter θ before each iteration. Exemplarily, if the current iteration is the first iteration, the first parameter is θ0; if the current iteration is the second iteration, the first parameter is θ1; if the current iteration is the third iteration, the first parameter is θ2. And so on, which will not be elaborated here. The noise signal of the current iteration refers to the noise signal determined in the current iteration. If the current iteration is the first iteration, the noise signal of the current iteration is the noise signal 1; if the current iteration is the second iteration, the noise signal of the current iteration is the noise signal 2.

[0093] S402: Based on the noise signal of the current iteration, use the distributed robust optimization model to determine the second parameter of the battery model obtained after the current iteration.

[0094] Exemplarily, if the current iteration is the first iteration, the second parameter is θ1; if the current iteration is the second iteration, the first parameter is θ2; if the current iteration is the third iteration, the first parameter is θ3. And so on, which will not be elaborated here.

[0095] S403: Based on the number of iterations and the preset number of iterations, identify the parameters of the battery model of the battery to obtain a parameter identification result.

[0096] After each iteration, the number of iterations is incremented by 1. If the number of iterations is less than the preset number of iterations, the iteration can continue. For example, if the preset number of iterations is 100 and the number of iterations is 80, the iteration can continue. If the number of iterations reaches 100, the second parameter of the battery model obtained in the 100th iteration can be used as the parameter identification result.

[0097] The method provided by the embodiments of the present application iterates the battery model using a distributionally robust optimization model to obtain a parameter identification result, so that on the one hand, the noise that makes the distributionally robust optimization model perform the worst is found, and on the other hand, under the action of the worst-performing noise, the parameters that make the distributionally robust optimization model have the highest accuracy are found, improving the stability of the obtained parameter identification result.

[0098] In one embodiment, for the above S403, based on the number of iterations and the preset number of iterations, the parameters of the battery model of the battery are identified to obtain a parameter identification result, which can be implemented in the following manner:

[0099] When the number of iterations reaches the preset number of iterations, the second parameter of the battery model obtained after the most recent iteration is used as the parameter identification result.

[0100] When the preset number of iterations is 100, the second parameter of the battery model obtained after the 100th iteration can be used as the parameter identification result.

[0101] The method provided by the embodiments of the present application, when the number of iterations reaches the preset number of iterations, uses the second parameter of the battery model obtained after the most recent iteration as the parameter identification result, further improving the stability of the obtained parameter identification result.

[0102] Refer to Figure 5 , Figure 5 is a schematic flowchart of the evaluation model construction method provided by the embodiments of the present application. This method may include the following steps S501 - S502:

[0103] S501, perform equivalent processing on the battery to obtain an equivalent circuit of the battery, and construct a battery model based on the equivalent circuit.

[0104] The equivalent circuit may refer to Figure 6 shown, Figure 6 is a schematic structural diagram of an equivalent circuit provided by the embodiments of the present application. This equivalent circuit is a first-order RC equivalent circuit, and this equivalent circuit includes an ohmic internal resistance R0, a polarization internal resistance R p , a polarization capacitor C P , and a voltage source U OCV . Based on this equivalent circuit, an equivalent circuit model is constructed. This equivalent circuit model is a first-order RC equivalent circuit model, and the first-order RC equivalent circuit model is shown in the following formulas (3) and (4):

[0105]

[0106]

[0107] Among them, OCV(t) = OCV(SOC(t)), It is a look-up table function, which can be obtained from the OCV-SOC relationship. U(t) represents the voltage of the battery at time t, I(t) represents the current of the battery at time t, SOC(t) represents the state of charge of the battery at time t. OCV is the abbreviation of Open Circuit Voltage, referring to the open circuit voltage. OCV means the potential difference between the positive and negative electrodes of the battery when the battery is in a non-charging and discharging state, that is, in an open circuit state. OCV(t) represents the open circuit voltage of the battery at time t.

[0108] The equivalent circuit model can include but is not limited to any-order RC models, equivalent circuit models with hysteresis effects. Among them, the order of any-order RC models can be numerical values such as 0, 1, 2, 3, etc. When the order is equal to 1, the equivalent circuit model is a first-order RC equivalent circuit model. The first-order RC equivalent circuit model can be constructed based on the first-order RC equivalent circuit of the battery. The parameters to be identified in the first-order RC equivalent circuit model include the ohmic internal resistance R0, the polarization internal resistance R p , the polarization capacitance C P , the initial state of charge SOC 0 and the battery capacity Q0. Among them, R0, R p and C P can be further used as dependent variables of SOC 0 . According to Kirchhoff's current law and Kirchhoff's voltage law, the above first-order RC equivalent circuit model can be obtained.

[0109] S502. Construct an evaluation model based on the battery model and data samples.

[0110] The method provided in the embodiments of the present application constructs an evaluation model based on the battery model and data samples, thereby laying a foundation for determining the parameter identification result based on the evaluation model, and further improving the stability of the obtained parameter identification result.

[0111] Refer to Figure 7 , Figure 7 is a schematic flow chart of the evaluation model construction method provided in the embodiments of the present application. The method can include the following steps S701-S703:

[0112] S701. Input the input data sequence in the data sample into the battery model to obtain a predicted data sequence.

[0113] The input data sequence can include but is not limited to a current data sequence, a temperature data sequence, etc. The predicted data sequence includes but is not limited to a voltage data sequence, a state of charge sequence, etc.

[0114] S702. Determine the difference between the predicted data sequence and the true output data sequence.

[0115] S703. Construct an evaluation model according to the difference.

[0116] The constructed evaluation models include, but are not limited to, the MSE model, the RMSE model, the MAE model, etc.

[0117] The method provided by the embodiments of the present application constructs an evaluation model based on the difference between the predicted data sequence and the true output data sequence, thereby laying a foundation for determining the parameter identification result based on the evaluation model, and further improving the stability of the obtained parameter identification result.

[0118] In one embodiment, the noise signal of the input data sequence satisfying the preset conditions includes: the noise signal of the input data sequence is continuous noise, and / or the signal-to-noise ratio of the input data sequence is greater than the preset signal-to-noise ratio.

[0119] For traditional offline parameter identification methods, for a given battery model, when there is a noise signal in the true data input signal, even when the signal-to-noise ratio of the data input signal is relatively high, the stability of the parameter identification result obtained by the parameter identification algorithm is poor. In other words, if the noise signal is small, the parameter identification result obtained by parameter identifying the battery model may change greatly. In the case where the noise signal is a continuous noise signal, the stability of the parameter identification result obtained by traditional parameter identification algorithms is also poor. In the embodiments of the present application, by restricting the noise signal, that is, the noise signal of the input data sequence is continuous noise, and / or the signal-to-noise ratio of the input data sequence is greater than the preset signal-to-noise ratio, an evaluation model is constructed based on data samples including noise signals that meet the preset conditions. Even when there is a slight noise in the input data sequence, a parameter identification result with better stability can still be obtained, so that the stability of the parameter identification result can be improved under any noise conditions.

[0120] In the embodiments of the present application, by restricting the noise signal of the input data sequence to be continuous noise, and / or the signal-to-noise ratio of the input data sequence is greater than the preset signal-to-noise ratio, an evaluation model is constructed based on data samples including noise signals that meet the preset conditions, so that the stability of the parameter identification result obtained based on the evaluation model can be improved under any noise conditions.

[0121] For a clearer introduction of the embodiments of the present application, the following specifically exemplarily describes in conjunction with Figure 6 the shown equivalent circuit.

[0122] After obtaining the first-order RC equivalent circuit model shown in formulas (3) and (4) above, the first-order RC equivalent circuit model can be discretized to adapt to real vehicle data to complete parameter identification. The discretized model is shown in the following formulas (5) and (6):

[0123]

[0124]

[0125] We set the time interval of adjacent iterative data points to the sampling interval of the actual vehicle, and use the least squares method to evaluate the error between the predicted voltage sequence and the true voltage sequence, as shown in formula (7):

[0126]

[0127] where x = [R0, R1, C1, SOC 0 , Q0] is the parameter vector to be identified, K is the total number of sampled data points, is the true voltage at time k. Substituting the discretized state equation into U k , we then obtain the modeling of the least squares problem, that is, the evaluation model. After that, a distributed robust optimization model can be constructed based on the evaluation model, and the parameter vector to be identified can be determined based on the distributed robust optimization model.

[0128] It should be understood that although the steps in the flowcharts involved in the above embodiments are shown in sequence according to the arrows, these steps do not necessarily need to be executed in the order indicated by the arrows. Unless clearly stated in this article, the execution of these steps has no strict order limitation, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages. These steps or stages do not necessarily need to be executed at the same time, but can be executed at different times. The execution order of these steps or stages does not necessarily need to be sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0129] Based on the same inventive concept, the embodiments of the present application also provide a parameter identification device for implementing the parameter identification method involved above. The implementation solutions provided by this device to solve problems are similar to the implementation solutions described in the above method. Therefore, the specific limitations in one or more embodiments of the parameter identification device provided below can refer to the limitations on the parameter identification method in the above text, and will not be repeated here.

[0130] In one embodiment, as Figure 8 shown, Figure 8 is one of the structural schematic diagrams of a parameter identification device provided by an embodiment of the present application. The device 800 includes:[[]]

[0131] An acquisition module 801, configured to acquire data samples of the battery on the vehicle; the data samples include an input data sequence of the battery within a preset time period and a true output data sequence corresponding to the input data sequence, and the noise signal of the input data sequence satisfies a preset condition;

[0132] An identification module 802, configured to identify parameters of a battery model of a battery based on an evaluation model and data samples to obtain a parameter identification result; the evaluation model is a model constructed based on the battery model and data samples of the battery, and the evaluation model is used to evaluate the difference degree between a predicted data sequence and an actual output data sequence, and the predicted data sequence is obtained by inputting an input data sequence into the battery model

[0133] In one embodiment, as Figure 9 shown Figure 9 FIG. 9 is a second schematic structural diagram of a parameter identification device provided by an embodiment of the present application. The identification module 802 in the device 900 includes:

[0134] A construction unit 8021, configured to construct a distributionally robust optimization model based on an evaluation model and data samples; the distributionally robust optimization model is used to characterize the expected error of the evaluation results of the data samples obtained based on the evaluation model; each data sample is a sample obtained under the same working condition;

[0135] An identification unit 8022, configured to identify parameters of a battery model of a battery according to the distributionally robust optimization model to obtain a parameter identification result.

[0136] In one embodiment, as Figure 10 shown Figure 10 FIG. 10 is a third schematic structural diagram of a parameter identification device provided by an embodiment of the present application. The identification unit 8022 in the device 1000 includes:

[0137] A first determination subunit 80221, configured to determine a noise signal of the current iteration from a noise signal of an input data sequence by using the distributionally robust optimization model according to a first parameter of the battery model obtained after the previous iteration of the battery model;

[0138] A second determination subunit 80222, configured to determine a second parameter of the battery model obtained after the current iteration by using the distributionally robust optimization model based on the noise signal of the current iteration;

[0139] An identification subunit 80223, configured to identify parameters of a battery model of a battery based on the number of iterations and a preset number of iterations to obtain a parameter identification result.

[0140] In one embodiment, the identification subunit 80223 is specifically configured to, when the number of iterations reaches the preset number of iterations, use the second parameter of the battery model obtained after the last iteration as the parameter identification result.

[0141] In one embodiment, as Figure 11 shown Figure 11FIG. 4 is a schematic structural diagram of a parameter identification device provided by an embodiment of the present application. The device 1100 further includes:

[0142] A first construction module 1101, configured to perform equivalent processing on a battery to obtain an equivalent circuit of the battery, and construct a battery model based on the equivalent circuit;

[0143] A second construction module 1102, configured to construct an evaluation model based on the battery model and data samples.

[0144] In one embodiment, the second construction module 1102 is specifically configured to input an input data sequence in the data sample into the battery model to obtain a predicted data sequence; determine the difference between the predicted data sequence and the true output data sequence; and construct an evaluation model according to the difference.

[0145] In one embodiment, the noise signal of the input data sequence satisfying a preset condition includes that the noise signal of the input data sequence is continuous noise, and / or the signal-to-noise ratio of the input data sequence is greater than a preset signal-to-noise ratio.

[0146] Each module in the above parameter identification device can be implemented in whole or in part by software, hardware, and their combination. The above modules can be embedded in or independent of a processor in a computer device in the form of hardware, or stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0147] In one embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, the following steps are implemented:

[0148] Obtain data samples of the battery on the vehicle; the data samples include an input data sequence of the battery within a preset time period and a true output data sequence corresponding to the input data sequence, and the noise signal of the input data sequence satisfies a preset condition;

[0149] Based on the evaluation model and the data samples, identify the parameters of the battery model of the battery to obtain a parameter identification result; the evaluation model is a model constructed based on the battery model of the battery and the data samples, and the predicted data sequence is obtained by inputting the input data sequence into the battery model.

[0150] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0151] Construct a distributed robust optimization model based on the evaluation model and the data samples; the distributed robust optimization model is used to characterize the expected error of the evaluation results of each data sample obtained based on the evaluation model; each data sample is a sample obtained under the same working condition;

[0152] According to the distributed robust optimization model, the parameters of the battery model of the battery are identified to obtain the parameter identification result.

[0153] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0154] According to the first parameters of the battery model obtained after the previous iteration of the battery model, using the distributed robust optimization model, determine the noise signal of the current iteration from the noise signals of the input data sequence;

[0155] Based on the noise signal of the current iteration, use the distributed robust optimization model to determine the second parameters of the battery model obtained after the current iteration;

[0156] Based on the number of iterations and the preset number of iterations, identify the parameters of the battery model of the battery to obtain the parameter identification result.

[0157] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0158] In the case where the number of iterations reaches the preset number of iterations, use the second parameters of the battery model obtained after the most recent iteration as the parameter identification result.

[0159] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0160] Perform equivalent processing on the battery to obtain the equivalent circuit of the battery, and construct a battery model based on the equivalent circuit;

[0161] Construct an evaluation model based on the battery model and the data samples.

[0162] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0163] Input the input data sequence in the data samples into the battery model to obtain a predicted data sequence;

[0164] Determine the difference between the predicted data sequence and the true output data sequence;

[0165] Construct an evaluation model according to the difference.

[0166] In one embodiment, the noise signal of the input data sequence satisfying the preset conditions includes: the noise signal of the input data sequence is continuous noise, and / or, the signal-to-noise ratio of the input data sequence is greater than the preset signal-to-noise ratio.

[0167] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0168] Obtain data samples of the battery on the vehicle; the data samples include an input data sequence of the battery within a preset time period and a true output data sequence corresponding to the input data sequence, and the noise signal of the input data sequence satisfies a preset condition;

[0169] Based on the evaluation model and the data samples, identify the parameters of the battery model of the battery to obtain a parameter identification result; the evaluation model is a model constructed based on the battery model of the battery and the data samples, and the predicted data sequence is obtained by inputting the input data sequence into the battery model.

[0170] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0171] Construct a distributionally robust optimization model based on the evaluation model and the data samples; the distributionally robust optimization model is used to characterize the expected error of the evaluation results of each data sample obtained based on the evaluation model; each data sample is a sample obtained under the same working conditions;

[0172] According to the distributionally robust optimization model, identify the parameters of the battery model of the battery to obtain a parameter identification result.

[0173] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0174] According to the first parameters of the battery model obtained after the previous iteration of the battery model, use the distributionally robust optimization model to determine the noise signal of the current iteration from the noise signal of the input data sequence;

[0175] Based on the noise signal of the current iteration, use the distributionally robust optimization model to determine the second parameters of the battery model obtained after the current iteration;

[0176] Based on the number of iterations and the preset number of iterations, identify the parameters of the battery model of the battery to obtain a parameter identification result.

[0177] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0178] In the case where the number of iterations reaches the preset number of iterations, use the second parameters of the battery model obtained after the most recent iteration as the parameter identification result.

[0179] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0180] Perform equivalent processing on the battery to obtain an equivalent circuit of the battery, and construct a battery model based on the equivalent circuit;

[0181] Construct an evaluation model based on the battery model and the data samples.

[0182] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0183] Input the input data sequence in the data sample into the battery model to obtain a predicted data sequence;

[0184] Determine the difference between the predicted data sequence and the true output data sequence;

[0185] Construct an evaluation model according to the difference.

[0186] In one embodiment, the noise signal of the input data sequence satisfying the preset condition includes: the noise signal of the input data sequence is continuous noise, and / or, the signal-to-noise ratio of the input data sequence is greater than the preset signal-to-noise ratio.

[0187] In one embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0188] Obtain a data sample of the battery on the vehicle; the data sample includes the input data sequence of the battery within a preset time period and the true output data sequence corresponding to the input data sequence, and the noise signal of the input data sequence satisfies the preset condition;

[0189] Based on the evaluation model and the data sample, identify the parameters of the battery model of the battery to obtain a parameter identification result; the evaluation model is a model constructed based on the battery model of the battery and the data sample, and the predicted data sequence is obtained after inputting the input data sequence into the battery model.

[0190] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0191] Construct a distributed robust optimization model based on the evaluation model and the data sample; the distributed robust optimization model is used to characterize the expected error of the evaluation results of each data sample obtained based on the evaluation model; each data sample is a sample obtained under the same working condition;

[0192] According to the distributed robust optimization model, identify the parameters of the battery model of the battery to obtain a parameter identification result.

[0193] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0194] According to the first parameter of the battery model obtained after the previous iteration of the battery model, use the distributed robust optimization model to determine the noise signal of the current iteration from the noise signal of the input data sequence;

[0195] Based on the noise signal of the current iteration, use the distributed robust optimization model to determine the second parameter of the battery model obtained after the current iteration;

[0196] Based on the number of iterations and a preset number of iterations, parameter identification of the battery model of the battery is performed to obtain a parameter identification result.

[0197] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0198] In the case where the number of iterations reaches the preset number of iterations, the second parameter of the battery model obtained after the most recent iteration is used as the parameter identification result.

[0199] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0200] The battery is equivalently processed to obtain an equivalent circuit of the battery, and a battery model is constructed based on the equivalent circuit;

[0201] An evaluation model is constructed based on the battery model and data samples.

[0202] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0203] The input data sequence in the data sample is input into the battery model to obtain a predicted data sequence;

[0204] Determine the difference between the predicted data sequence and the true output data sequence;

[0205] An evaluation model is constructed according to the difference.

[0206] In one embodiment, the noise signal of the input data sequence satisfying the preset condition includes: the noise signal of the input data sequence is continuous noise, and / or, the signal-to-noise ratio of the input data sequence is greater than the preset signal-to-noise ratio.

[0207] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0208] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memory can include Read-Only Memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., without limitation.

[0209] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification.

[0210] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A parameter identification method, characterized in that, The method includes: Obtaining data samples of the battery on the vehicle; the data samples include an input data sequence of the battery within a preset time period and a true output data sequence corresponding to the input data sequence, and the noise signal of the input data sequence satisfies a preset condition; Based on an evaluation model and the data samples, identifying parameters of the battery model of the battery to obtain a parameter identification result; the evaluation model is a model constructed based on the battery model of the battery and the data samples, and the predicted data sequence is obtained by inputting the input data sequence into the battery model.

2. The method according to claim 1, wherein The identifying parameters of the battery model of the battery based on the evaluation model and the data samples to obtain a parameter identification result includes: Constructing a distributionally robust optimization model based on the evaluation model and the data samples; the distributionally robust optimization model is used to characterize the expected error of the evaluation results of the data samples obtained based on the evaluation model; each of the data samples is a sample obtained under the same working condition; According to the distributionally robust optimization model, identifying parameters of the battery model of the battery to obtain a parameter identification result.

3. The method according to claim 2, wherein The identifying parameters of the battery model of the battery according to the distributionally robust optimization model to obtain a parameter identification result includes: According to the first parameters of the battery model obtained after the previous iteration of the battery model, using the distributionally robust optimization model to determine the noise signal of the current iteration from the noise signal of the input data sequence; Based on the noise signal of the current iteration, using the distributionally robust optimization model to determine the second parameters of the battery model obtained after the current iteration; Based on the number of iterations and a preset number of iterations, identifying parameters of the battery model of the battery to obtain a parameter identification result.

4. The method according to claim 3, wherein The identifying parameters of the battery model of the battery based on the number of iterations and a preset number of iterations to obtain a parameter identification result includes: When the number of iterations reaches the preset number of iterations, using the second parameters of the battery model obtained after the most recent iteration as the parameter identification result.

5. The method according to any one of claims 1-4, characterized in that, The method further includes: Performing equivalent processing on the battery to obtain an equivalent circuit of the battery, and constructing the battery model based on the equivalent circuit; Constructing the evaluation model based on the battery model and the data samples.

6. The method according to claim 5, wherein The constructing the evaluation model based on the battery model and the data samples includes: Inputting the input data sequence in the data samples into the battery model to obtain the predicted data sequence; Determining the difference between the predicted data sequence and the true output data sequence; Constructing the evaluation model according to the difference.

7. The method according to any one of claims 1-6, characterized in that, The noise signal of the input data sequence satisfying the preset condition includes: the noise signal of the input data sequence is continuous noise, and / or, the signal-to-noise ratio of the input data sequence is greater than a preset signal-to-noise ratio.

8. A parameter identification device, characterized in that, The device includes: An acquisition module, configured to acquire data samples of a battery on a vehicle; the data samples include an input data sequence of the battery within a preset time period and a true output data sequence corresponding to the input data sequence, and a noise signal of the input data sequence satisfies a preset condition; An identification module, configured to identify parameters of a battery model of the battery based on an evaluation model and the data samples to obtain a parameter identification result; the evaluation model is a model constructed based on the battery model of the battery and the data samples, and the evaluation model is used to evaluate a difference degree between a predicted data sequence and the true output data sequence, and the predicted data sequence is obtained by inputting the input data sequence into the battery model.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1-7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, the steps of the method according to any one of claims 1-7 are implemented.

11. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, the steps of the method according to any one of claims 1-7 are implemented.

Images