Dynamic prediction method for power battery charging curve based on bias adjustment

Abstract

The application relates to the technical field of power batteries, and discloses a power battery charging curve dynamic prediction method and system based on deviation adjustment, which comprises the following steps: acquiring vehicle historical charging data; determining charging mode classification, ideal charging curves corresponding to each classification, and historical charging segment feature data of each classification according to the vehicle historical charging data, and constructing a data set; training a classification model using the data set to obtain a trained classification model; obtaining a predicted charging mode classification according to real-time charging segment feature data through the trained classification model, and obtaining a corresponding ideal charging curve as an initial prediction curve; calculating a deviation adjustment term through a deviation adjustment formula according to the real-time charging segment feature data, and adjusting the initial prediction curve based on the deviation adjustment term to obtain a corrected prediction curve. The application can accurately predict the charging duration, dynamically optimize the charging process, reduce the charging time, and improve the utilization efficiency of charging piles and vehicle-mounted power batteries.

Description

Technical Field

[0001] The present invention relates to the technical field of power batteries, and in particular to a method for dynamically predicting a power battery charging curve based on deviation adjustment. Background Art

[0002] Currently, most research on predicting charging times for electric vehicle batteries focuses on charging network locations, focusing on aspects such as charging station resource allocation, grid load balancing, and optimizing the user charging experience. However, these studies primarily target highly intelligent public charging networks, which are typically concentrated in areas with well-developed infrastructure, such as large cities or highway service areas. For vehicles that frequently travel across regions, relying solely on specific charging network locations for predictions presents significant limitations. If vehicles could autonomously predict their charging times based on real-time charging data from local power batteries, regardless of charging network locations, this would not only help drivers plan their trips more effectively and reduce unnecessary waiting time, but would also significantly improve travel efficiency and user experience, enabling more intelligent charging behavior management.

[0003] The problem of predicting vehicle charging time can actually be regarded as a combination of the problem of charging mode classification and the problem of automatic fitting of charging curves. First of all, the morphology of the charging curves under different charging modes (such as constant current mode and constant voltage mode) is highly similar. Therefore, classification of specific charging modes is a prerequisite for achieving accurate prediction of power battery charging time. Secondly, from the actual usage scenarios of new energy vehicles, real-time prediction of charging time still faces many challenges. The main reason is that during the charging process, due to the combined influence of factors such as the current battery temperature, SOC and charging rate, the charging power is prone to dynamic adjustment or fluctuation, resulting in the staged irregular characteristics of the charging curve. These problems directly affect the accuracy of charging time prediction. Therefore, in order to effectively deal with the dynamic fluctuations of current and voltage during the charging process, proposing a corresponding efficient and adaptive deviation adjustment model is another prerequisite for achieving accurate prediction of battery charging time.

[0004] Therefore, there is an urgent need for a dynamic prediction method for power battery charging curve based on deviation adjustment, which can accurately predict the charging time and dynamically optimize the charging process, reduce the charging time, and improve the utilization efficiency of charging piles and vehicle-mounted power batteries. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention provides a dynamic prediction method for power battery charging curve based on deviation adjustment, which can accurately predict the charging time and dynamically optimize the charging process, reducing the charging time and improving the utilization efficiency of charging piles and vehicle-mounted power batteries.

[0006] The application provides a power battery charging curve dynamic prediction method based on deviation adjustment, comprising the following steps:

[0007] S1, historical charging data of a vehicle is acquired; wherein the historical charging data comprises time, voltage, current, temperature and SOC value;

[0008] S2, charging mode classification, ideal charging curve corresponding to each classification and historical charging segment feature data of each classification are determined according to the historical charging data of the vehicle, and a data set is constructed according to the charging mode classification and the historical charging segment feature data of each classification;

[0009] S3, a classification model is constructed, the classification model is trained using the data set, and a trained classification model is obtained;

[0010] S4, real-time charging segment feature data is acquired, a predicted charging mode classification is obtained through the trained classification model, and an ideal charging curve corresponding to the predicted charging mode classification is obtained as an initial prediction curve;

[0011] S5, a deviation adjustment term is calculated according to the real-time charging segment feature data through a deviation adjustment formula, the initial prediction curve is adjusted based on the deviation adjustment term, and a corrected prediction curve is obtained.

[0012] Further, in S3, the classification model is an LSTM model, the input of the classification model is the historical charging segment feature data, and the output of the classification model is the charging mode classification corresponding to the historical charging segment feature data.

[0013] Further, in S5, the deviation adjustment term is calculated according to the real-time charging segment feature data through the deviation adjustment formula, the initial prediction curve is adjusted based on the deviation adjustment term, and the corrected prediction curve is obtained, wherein the deviation adjustment formula is as follows:

[0014] s i+1 =s i +ωc i Δt+δ=f(t,c i Δt,s i ,σ);

[0015] Wherein, s i represents the SOC value at the current moment, s i+1 represents the SOC value at the next moment, c i represents the current at the current moment, △t represents the sampling time interval, δ represents the deviation adjustment term, σ represents the statistical value of the static feature, ω represents the coefficient of the linear relationship between the current and the SOC value, and f(t, c i Δt, s i, σ) represents the coefficient of the linear relationship between current and SOC value and the relationship between current and static characteristics, and t represents time t; among them, the deviation adjustment item is divided into linear deviation adjustment item, logarithmic deviation adjustment item, proportional deviation adjustment item and deviation adjustment item based on deep learning model.

[0016] Furthermore, when the deviation adjustment item is a linear deviation adjustment item, the calculation formula of the deviation adjustment item is as follows:

[0017]

[0018] Among them, δ represents the deviation adjustment term, c i Indicates the current charging current. It represents the expected value of the predicted current, and k represents the adjustment coefficient.

[0019] Furthermore, when the deviation adjustment term is a logarithmic deviation adjustment term, the calculation formula of the deviation adjustment term is as follows:

[0020]

[0021] Among them, δ represents the deviation adjustment term, c i Indicates the current charging current. It represents the expected value of the predicted current, and k represents the adjustment coefficient.

[0022] Furthermore, when the deviation adjustment item is a proportional deviation adjustment item, the calculation formula of the deviation adjustment item is as follows:

[0023]

[0024] Among them, δ represents the deviation adjustment term, c i Indicates the current charging current. Represents the expected value of the predicted current.

[0025] Furthermore, when the bias adjustment item is a bias adjustment item based on a deep learning model, the calculation formula of the bias adjustment item is as follows:

[0026] δ=f(t,c i ,s i ,σ);

[0027] Among them, δ represents the deviation adjustment term, s i Indicates the SOC value at the current moment, c i represents the current at the current moment, σ represents the statistical value of the static feature, t represents the time t, and f represents the deep learning model based on the time series prediction model.

[0028] The application also provides a power battery charging curve dynamic prediction system based on deviation adjustment, which is used for realizing the power battery charging curve dynamic prediction method based on deviation adjustment as any one of the above.

[0029] A data acquisition module is configured to acquire vehicle historical charging data.

[0030] A data set construction module is connected with the data acquisition module and is configured to determine charging mode classification, ideal charging curves corresponding to each classification and historical charging segment feature data of each classification according to the vehicle historical charging data, and construct a data set according to the charging mode classification and the historical charging segment feature data of each classification.

[0031] A classification model construction module is connected with the data set construction module and is configured to construct a classification model, train the classification model using the data set and obtain a trained classification model.

[0032] An initial curve prediction module is connected with the classification model construction module and is configured to acquire real-time charging segment feature data, obtain a predicted charging mode classification through the trained classification model and obtain a corresponding ideal charging curve as an initial prediction curve according to the predicted charging mode classification.

[0033] A deviation adjustment module is connected with the initial curve prediction module and is configured to calculate a deviation adjustment term according to the real-time charging segment feature data through a deviation adjustment formula, adjust the initial prediction curve based on the deviation adjustment term and obtain a corrected prediction curve.

[0034] The embodiments of the application have the following technical effects:

[0035] 1. Through accurate classification of charging modes and real-time deviation adjustment, the application can more accurately predict charging duration, dynamically optimize the charging process, reduce charging time and improve the utilization efficiency of charging piles and vehicle-mounted power batteries.

[0036] 2. The application provides four deviation adjustment methods, namely linear, logarithmic, proportional and deep learning (such as LSTM), for different computing power conditions. In the adjustment methods based on linearity, logarithm and proportionality, the calculation complexity is low and they are suitable for scenarios with limited computing power. When the computing power is sufficient, the dynamic deviation adjustment method based on the deep learning model can further improve the prediction accuracy and real-time adjustment ability of the charging curve and flexibly adapt to diversified charging environments.

[0037] 3. Through the dynamic deviation adjustment module, the application effectively solves the prediction error caused by fluctuations in charging parameters such as current and voltage, ensures that the prediction curve is highly matched with the actual charging behavior, and especially in complex working conditions such as battery aging and environmental temperature changes, the application can still maintain high-precision prediction.

[0038] 4. Dynamic deviation adjustment can respond to abnormalities that may occur during charging, such as overcharging, undercharging, or excessive current, in real time, and by correcting the charging curve in real time, the safety of the battery in use is ensured;

[0039] 5. For devices with low computing power, the application provides linear, logarithmic and proportional deviation adjustment methods, which significantly reduce the requirements for hardware performance, expand the application range of the algorithm, and enable stable operation in ordinary charging devices. BRIEF DESCRIPTION OF DRAWINGS

[0040] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the drawings needed in the specific embodiments or the prior art description will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0041] Figure 1 is a flowchart of the dynamic prediction method of the power battery charging curve based on deviation adjustment provided by the embodiment of the present application;

[0042] Figure 2 is a comparison diagram of the initial prediction curve and the actual charging process provided by the embodiment of the present application;

[0043] Figure 3 is a schematic diagram of current fluctuations occurring in the actual charging process provided by the embodiment of the present application;

[0044] Figure 4 is a comparison diagram of the prediction curve corrected by the deviation adjustment formula and the actual charging process provided by the embodiment of the present application;

[0045] Figure 5 is a comparison diagram of the actual charging process, the initial prediction curve, and the prediction curve corrected by the deviation adjustment formula provided by the embodiment of the present application;

[0046] Figure 6 is a structural schematic diagram of the dynamic prediction system of the power battery charging curve based on deviation adjustment provided by the embodiment of the present application. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solutions and advantages of the present application more clear, the technical solutions of the present application will be described clearly and completely below. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of the present application.

[0048] The charging behavior in the charging process of new energy vehicles is affected by multiple factors, including charging current, voltage, SOC (state of charge), ambient temperature, etc. These factors are closely related to the charging mode (such as constant current mode, constant voltage mode), and the change of the charging curve under different modes has significant regularity.

[0049] In the actual application of charging, the dynamic changes of the SOC of the battery, the current fluctuation, the ambient temperature, etc. have an important influence on the accuracy of the charging duration prediction and the safety of the charging process. However, existing researches are mostly focused on the optimization of the fixed mode of specific charging facilities, and the adaptability of the vehicle in the dynamic charging scene is ignored.

[0050] To solve the above problems, the present application provides a dynamic prediction method for the charging curve of a power battery based on deviation adjustment, Figure 1 is a flowchart of the dynamic prediction method for the charging curve of a power battery based on deviation adjustment provided by the embodiments of the present application, referring to Figure 1 , specifically comprising:

[0051] S1, obtaining historical charging data of a vehicle.

[0052] In some embodiments, all complete charging segments are extracted from the historical charging data of the vehicle, including the historical charging data of each segment, including time t, voltage V i , current c i , temperature T i and SOC value s i , etc.

[0053] S2, determining the charging mode classification, the ideal charging curve corresponding to each classification, and the historical charging segment feature data of each classification according to the historical charging data of the vehicle, and constructing a data set according to the charging mode classification and the historical charging segment feature data of each classification.

[0054] In some embodiments, the historical charging data of the vehicle is analyzed to summarize the possible charging modes of the vehicle, such as constant current mode (CC) and constant voltage mode (CV). These modes form the basic characteristics of the charging behavior and provide data support for the training of the subsequent classification model.

[0055] In some embodiments, according to the classification result, an initial prediction curve C under the current charging mode is generated. The prediction curve is based on the mathematical model of the classification mode, and the relationship between the charging current, voltage and SOC is combined to calculate the charging duration and charging rate.

[0056] The historical charging segment feature data is standardized, and the characteristic values such as current and voltage are scaled to the interval [0, 1] to eliminate the influence of the characteristic quantity dimension on model training. The missing data is processed by a low-rank matrix filling method to ensure the consistency and integrity of the charging data. The data set is constructed according to the charging mode classification and the historical charging segment feature data of each classification.

[0057] S3, a classification model is constructed, and the data set is used to train the classification model to obtain a trained classification model.

[0058] In some embodiments, the classification model is an LSTM model, the input of the classification model is the historical charging segment feature data, and the output of the classification model is the charging mode classification corresponding to the historical charging segment feature data. The LSTM model is trained using the data set, the input is the time sequence feature (time, current, etc.), the output is the predicted charging mode category, and the LSTM model parameters are optimized through the label in the historical data to improve the classification accuracy.

[0059] S4, real-time charging segment feature data is obtained, the predicted charging mode classification is obtained through the trained classification model, and the corresponding ideal charging curve is obtained as the initial prediction curve according to the predicted charging mode classification.

[0060] In some embodiments, real-time charging segment feature data is obtained, which is input into the trained classification model, and the model output result is the charging mode to which the current real-time charging segment feature data belongs, such as constant current mode or constant voltage mode. The corresponding ideal charging curve is obtained as the initial prediction curve according to the predicted charging mode classification.

[0061] Figure 2 is a comparison diagram of an initial prediction curve and an actual charging process provided by an embodiment of the application, Figure 3 is a schematic diagram of current fluctuation in an actual charging process, see Figure 2 and Figure 3 It can be seen that the initial prediction curve is consistent with the actual charging curve in the overall trend, but it performs poorly in dealing with sudden situations (such as the current stage fluctuation shown in Figure 3 .

[0062] S5, a deviation adjustment term is calculated according to the real-time charging segment feature data through a deviation adjustment formula, and the initial prediction curve is adjusted based on the deviation adjustment term to obtain a corrected prediction curve.

[0063] In some embodiments, a deviation adjustment term is calculated according to the deviation adjustment formula, and the initial prediction curve is adjusted based on the deviation adjustment term to obtain a corrected prediction curve, wherein the deviation adjustment formula is as follows:

[0064] s i+1= s i + ωc i Δt + δ = f(t, c i Δt, s i , σ);

[0065] wherein s i denotes the SOC value at the current time, s i+1 denotes the SOC value at the next time, c i denotes the current at the current time, Δt denotes the sampling time interval, δ denotes the bias adjustment term, σ denotes the statistical value of the static characteristic (such as the ambient temperature, the degree of battery aging, etc.), ω denotes the coefficient of the linear relationship between the current and the SOC value, f(t, c i Δt, s i , σ) denotes the coefficient of the linear relationship between the current and the SOC value and the relationship between the current and the static characteristic, and t denotes the time t.

[0066] In some embodiments, the bias adjustment term is divided into a linear bias adjustment term, a logarithmic bias adjustment term, a proportional bias adjustment term, and a bias adjustment term based on a deep learning model.

[0067] In some embodiments, when the bias adjustment term is a linear bias adjustment term, the calculation formula of the bias adjustment term is as follows:

[0068]

[0069] wherein δ denotes the bias adjustment term, c i denotes the charging current at the current time, denotes the expected value of the predicted current, and k denotes the adjustment coefficient. In the formula of the linear bias adjustment term, k is a linear proportionality coefficient.

[0070] The linear bias adjustment term has the characteristics of simple calculation, is suitable for scenarios with extremely low computing power, and can quickly realize bias adjustment.

[0071] In some embodiments, when the bias adjustment term is a logarithmic bias adjustment term, the calculation formula of the bias adjustment term is as follows:

[0072]

[0073] wherein δ denotes the bias adjustment term, c i denotes the charging current at the current time, denotes the expected value of the predicted current, and k denotes the adjustment coefficient. In the formula of the linear bias adjustment term, k is a linear proportionality coefficient.

[0074] The logarithmic deviation adjustment term enhances the adjustment capability for large deviations by taking the logarithm of the current difference, and is more suitable for scenarios with large deviations and nonlinear adjustment than the linear method.

[0075] In some embodiments, when the deviation adjustment term is a proportional deviation adjustment term, the calculation formula of the deviation adjustment term is as follows:

[0076]

[0077] Wherein, δ represents the deviation adjustment term, c i represents the charging current at the current time, represents the expected value of the predicted current.

[0078] The proportional deviation adjustment term adjusts the current value at the next time according to the proportional relationship between the current and the predicted value, and is suitable for scenarios where the charging curve presents a proportional change rule, and has lower computational complexity than the deep learning method.

[0079] In some embodiments, when the deviation adjustment term is a deep learning model-based deviation adjustment term, the calculation formula of the deviation adjustment term is as follows:

[0080] δ=f(t,c i ,s i ,σ);

[0081] Wherein, δ represents the deviation adjustment term, s i represents the SOC value at the current time, c i represents the current at the current time, σ represents the statistical value of the static feature, t represents the t time, and f represents the deep learning model based on the time series prediction model.

[0082] The deep deviation adjustment term can capture complex nonlinear characteristics and multivariate relationships, and is suitable for scenarios with sufficient computing power and high requirements for prediction accuracy.

[0083] In scenarios with insufficient computing power, the above linear, logarithmic and proportional deviation adjustment methods have the advantages of simple calculation, low resource occupation, and are efficient solutions.

[0084] When the computing power is allowed and higher accuracy is required, the deep learning deviation adjustment method based on LSTM can significantly improve the deviation correction capability and curve optimization effect, and is a more preferred solution.

[0085] Through the combination of the four deviation adjustment methods, the present application can flexibly adapt to the needs of different computing power and charging scenarios, ensuring computational efficiency and improving the matching degree of the charging curve and the actual charging behavior.

[0086] In some embodiments, the initial prediction curve is adjusted based on the bias adjustment term, and the adjustment can be performed using a direct superposition mapping method to obtain a real-time corrected charging prediction curve C adjusted =C+δ, the corrected prediction curve is more consistent with the current charging state, and the prediction accuracy and charging efficiency are significantly improved. Other mapping methods can also be used according to actual conditions, which are not limited herein.

[0087] Figure 4 is a comparison diagram of a prediction curve corrected by a bias adjustment formula provided by an embodiment of the present application and an actual charging process, referring to Figure 4 After adjusting the prediction curve by using the deep learning adjustment term in the bias adjustment term, the corrected prediction curve is more consistent with the actual charging process, and the duration required for this battery charging phase is accurately predicted.

[0088] Figure 5 is a comparison diagram of an actual charging process, an initial prediction curve, and a prediction curve corrected by a bias adjustment formula, referring to Figure 5 It can be seen that the power battery charging curve dynamic prediction method based on bias adjustment provided by the present method can effectively solve the problem of inaccurate curve fitting in the dynamic charging process, especially in dealing with overcharging prediction.

[0089] The present application realizes accurate prediction of charging duration in multiple scenarios by combining charging mode classification and dynamic bias adjustment. Various bias adjustment methods are provided to balance the application scenarios with limited computing power and high precision requirements. The optimized charging curve can reduce battery consumption, prolong battery life, and thus improve the safety of the charging process. In summary, the present application realizes intelligent optimization of charging behavior from the perspective of new energy vehicles through classification and adjustment, and has significant practical application value.

[0090] The present application also provides a power battery charging curve dynamic prediction system based on bias adjustment, which is used to realize the power battery charging curve dynamic prediction method based on bias adjustment, Figure 6 is a structural schematic diagram of a power battery charging curve dynamic prediction system based on bias adjustment provided by an embodiment of the present application, referring to Figure 6 The system comprises the following modules:

[0091] A data acquisition module is used to acquire vehicle historical charging data.

[0092] A data set construction module is connected with the data acquisition module, and is used to determine charging mode classification, ideal charging curves corresponding to each classification, and historical charging segment feature data of each classification according to the vehicle historical charging data, and to construct a data set according to the charging mode classification and the historical charging segment feature data of each classification.

[0093] The classification model building module is connected to the data set building module and is used to build a classification model and train the classification model using the data set to obtain a trained classification model;

[0094] The initial curve prediction module is connected to the classification model building module to obtain real-time charging segment feature data, obtain the predicted charging mode classification through the trained classification model, and obtain the corresponding ideal charging curve as the initial prediction curve based on the predicted charging mode classification;

[0095] The deviation adjustment module is connected to the initial curve prediction module and is used to calculate the deviation adjustment item through the deviation adjustment formula according to the real-time charging segment characteristic data, and adjust the initial prediction curve based on the deviation adjustment item to obtain a corrected prediction curve.

[0096] It should be noted that the terms used in the present invention are only for describing specific embodiments and are not intended to limit the scope of this application. As shown in the present specification, unless the context clearly indicates an exception, the words "one", "a", "a kind of" and / or "the" do not specifically refer to the singular and may also include the plural. The terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method or device comprising a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method or device. In the absence of further restrictions, the elements defined by the sentence "comprise a..." do not exclude the presence of other identical elements in the process, method or device comprising the elements.

[0097] It should also be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, they cannot be understood as limitations on the present invention. Unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", etc. should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be a communication between the internal parts of two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0098] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit the present application; although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions recorded in the above embodiments can still be modified, or some or all of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the technical solutions of the embodiments of the present application.

Claims

1. A dynamic prediction method for power battery charging curve based on deviation adjustment, characterized in that: The steps include: S1. Acquire historical charging data of the vehicle; wherein the historical charging data includes time, voltage, current, temperature and SOC value; S2. Determine charging mode classifications, ideal charging curves corresponding to each classification, and historical charging segment feature data for each classification based on historical vehicle charging data, and construct a data set based on the charging mode classifications and the historical charging segment feature data for each classification; S3. Build a classification model, and use the data set to train the classification model to obtain a trained classification model; S4. Acquire real-time charging segment feature data, obtain a predicted charging mode classification using the trained classification model, and obtain a corresponding ideal charging curve as an initial prediction curve based on the predicted charging mode classification; S5. Calculating a deviation adjustment term using a deviation adjustment formula according to the real-time charging segment characteristic data, and adjusting the initial prediction curve based on the deviation adjustment term to obtain a corrected prediction curve; The deviation adjustment formula is as follows: ； in, Indicates the SOC value at the current moment, Indicates the SOC value at the next moment, represents the current at the current moment, represents the sampling time interval, represents the deviation adjustment term, Represents the statistical value of static features, The coefficient representing the linear relationship between current and SOC value, The coefficient representing the linear relationship between current and SOC value, and the relationship between current and static characteristics, t represents time t; among them, the deviation adjustment items are divided into linear deviation adjustment items, logarithmic deviation adjustment items, proportional deviation adjustment items, and deviation adjustment items based on deep learning models.

2. The method for dynamic prediction of power battery charging curve based on deviation adjustment according to claim 1, characterized in that: In S3, the classification model is an LSTM model, the input of the classification model is historical charging segment feature data, and the output of the classification model is the charging mode classification corresponding to the historical charging segment feature data.

3. The method for dynamic prediction of power battery charging curve based on deviation adjustment according to claim 1, characterized in that: When the deviation adjustment item is a linear deviation adjustment item, the calculation formula of the deviation adjustment item is as follows: ； in, represents the deviation adjustment term, Indicates the current charging current. represents the expected value of the predicted current, Indicates the adjustment factor.

4. The method for dynamic prediction of power battery charging curve based on deviation adjustment according to claim 1, characterized in that: When the deviation adjustment item is a logarithmic deviation adjustment item, the calculation formula of the deviation adjustment item is as follows: ； in, represents the deviation adjustment term, Indicates the current charging current. represents the expected value of the predicted current, Indicates the adjustment factor.

5. The method for dynamic prediction of power battery charging curve based on deviation adjustment according to claim 1, characterized in that: When the deviation adjustment item is a proportional deviation adjustment item, the calculation formula of the deviation adjustment item is as follows: ； in, represents the deviation adjustment term, Indicates the current charging current. Represents the expected value of the predicted current.

6. The method for dynamic prediction of power battery charging curve based on deviation adjustment according to claim 1, characterized in that: When the deviation adjustment item is a deviation adjustment item based on a deep learning model, the calculation formula of the deviation adjustment item is as follows: ； in, represents the deviation adjustment term, Indicates the SOC value at the current moment, represents the current at the current moment, Represents the statistical value of the static feature, t represents time t, and f represents the deep learning model based on the time series prediction model.

7. A system for dynamic prediction of a power battery charging curve based on deviation adjustment, for implementing the method for dynamic prediction of a power battery charging curve based on deviation adjustment as described in any one of claims 1 to 6 above, characterized in that: The system includes the following modules: A data acquisition module is used to obtain historical charging data of the vehicle; a data set construction module, connected to the data acquisition module, for determining charging mode classifications, ideal charging curves corresponding to each classification, and historical charging segment characteristic data of each classification based on historical vehicle charging data, and constructing a data set based on the charging mode classifications and the historical charging segment characteristic data of each classification; A classification model construction module, connected to the data set construction module, is used to construct a classification model, and train the classification model using the data set to obtain a trained classification model; an initial curve prediction module, connected to the classification model building module, for obtaining real-time charging segment feature data, obtaining a predicted charging mode classification through the trained classification model, and obtaining a corresponding ideal charging curve as an initial prediction curve based on the predicted charging mode classification; The deviation adjustment module is connected to the initial curve prediction module and is used to calculate the deviation adjustment item through the deviation adjustment formula according to the real-time charging segment characteristic data, and adjust the initial prediction curve based on the deviation adjustment item to obtain a corrected prediction curve.

Images