Joint prediction method and device for state of health of electric vehicle battery and charging time

Abstract

This invention provides a method and apparatus for jointly predicting the health status and charging time of an electric vehicle battery. The method includes: acquiring raw data of the electric vehicle and preprocessing the raw data to construct a feature dataset of the vehicle battery; determining the charging mode of the vehicle battery based on the charging range of the remaining battery capacity and the type of charging station used; calling a pre-built health status prediction model and combining it with the feature dataset of the vehicle battery to determine an estimated value of the vehicle battery's health status; and calling a pre-built charging time prediction model and training the charging time prediction model using the feature dataset and the estimated health status of the vehicle battery. This invention improves the accuracy of charging time prediction after battery aging, fundamentally solving the problem of decreased prediction accuracy caused by battery aging.

Description

Technical Field

[0001] This invention relates to the field of electric vehicle battery management technology, and in particular to a method and apparatus for jointly predicting the health status and charging time of an electric vehicle battery. Background Technology

[0002] Electric vehicles (EVs), as an important means of reducing carbon emissions and promoting sustainable development, have received widespread global attention due to their zero-emission advantage. In recent years, EV sales have grown significantly, but their widespread adoption still faces several bottlenecks, such as range anxiety, long charging times, and insufficient charging infrastructure. The inadequacy of charging infrastructure is particularly prominent, especially during peak charging periods when queues often form at charging stations, severely impacting users' travel planning and experience. Therefore, accurately predicting EV charging times is crucial for improving user satisfaction, rationally planning charging infrastructure, and enhancing overall charging efficiency.

[0003] Existing methods for predicting electric vehicle charging time can be broadly categorized into model-based and data-driven approaches. Model-based methods typically rely on modeling the physical characteristics of the battery, estimating charging time through electrochemical models, equivalent circuit models, or mathematical models. These methods offer the advantage of accurately reflecting the battery's internal mechanisms, but their application is often limited by complex computational requirements and dependence on battery parameters. Electrochemical models contain certain errors in describing internal battery reactions, affecting their prediction accuracy; equivalent circuit models are usually designed for specific batteries and have poor applicability under different operating conditions; while mathematical models can predict charging time relatively accurately, their high computational complexity makes them difficult to widely apply in real-world scenarios.

[0004] Data-driven approaches, particularly those combining machine learning and deep learning techniques, have become a research hotspot in recent years. These methods collect historical charging data from electric vehicles and utilize statistical and machine learning algorithms to build charging time prediction models, achieving high estimation accuracy. However, the accuracy of data-driven methods largely depends on the quality of the dataset used and the appropriateness of feature selection. Furthermore, many data-driven methods do not adequately consider other important factors affecting charging time, such as battery aging and ambient temperature, which can significantly impact charging time.

[0005] In summary, while existing technologies have made some progress in predicting electric vehicle charging times, many problems remain to be solved. For example, how to accurately estimate charging times when battery charging conditions are difficult to predict in real-world applications, and how to comprehensively consider complex influences such as battery aging and environmental factors in the prediction model, are directions that require further research and optimization. More accurate charging time predictions can not only provide users with a better charging experience but also effectively support the planning and optimization of charging infrastructure, thus contributing to the further promotion and popularization of electric vehicles.

[0006] There is currently no effective solution to the problem of poor accuracy in predicting charging time in existing related technologies. Summary of the Invention

[0007] This invention provides a method and apparatus for jointly predicting the health status and charging time of an electric vehicle battery, in order to solve the shortcomings of existing related technologies in terms of poor accuracy in predicting charging time.

[0008] In a first aspect, the present invention provides a method for jointly predicting the health status and charging time of an electric vehicle battery, comprising: Obtain the raw data of electric vehicles and preprocess the raw data to construct a feature dataset of the vehicle battery; The charging mode of the vehicle battery is determined based on the charging range of the remaining power of the vehicle battery and the type of charging station used. A pre-built health status prediction model is invoked, and the estimated health status of the vehicle battery is determined by combining it with the feature dataset of the vehicle battery. A pre-built charging time prediction model is invoked and trained using the feature dataset and health status estimate of the vehicle battery. The preset charging rules are input into the trained charging time prediction model to obtain the charging time of the electric vehicle; the charging rules include the remaining battery power at the start and end of charging, the charging mode, and the estimated health status.

[0009] According to the present invention, a method for jointly predicting the battery health status and charging time of an electric vehicle is provided, which involves acquiring raw data of the electric vehicle and preprocessing the raw data, including: Obtain the raw data of the electric vehicle, remove irrelevant data from the raw data, and extract key data fields related to the electric vehicle's battery; The abnormal data in the key data fields are processed; the abnormal data includes duplicate values, missing values, current out of range, abnormal mileage jumps, disordered time sequence, and charging interruption data. The vehicle status is determined by the charging status field of the vehicle battery, and the charging data segment is filtered out to calculate the remaining battery power and charging time at the start and end of the vehicle battery charging.

[0010] According to the present invention, a method for jointly predicting the health status and charging time of an electric vehicle battery is provided, which constructs a feature dataset of the vehicle battery, including: Extract the remaining battery capacity characteristics of the vehicle battery and determine the frequency of the remaining battery capacity distribution in different intervals during charging; the remaining battery capacity characteristics include the remaining battery capacity at the start of charging, the remaining battery capacity at the end of charging, and the increment of the remaining battery capacity during charging; Extract the cumulative mileage from the raw data of the electric vehicle; Determine the number of days the electric vehicle was used and the number of days it was left idle; The ambient temperature and weather temperature distribution of the current environment of the electric vehicle are obtained; The voltage and temperature inconsistencies of the individual automotive battery cells were determined.

[0011] According to the present invention, a method for jointly predicting the health status and charging time of an electric vehicle battery is provided, which determines the voltage inconsistency and temperature inconsistency of the individual battery cells, including: The standard deviation of the voltage at the moment when the remaining charge of the car battery is 85% is used to characterize the voltage inconsistency of the individual cells. The temperature inconsistency of the individual cells is characterized by the standard deviation of the temperature at which the remaining charge of the car battery is 75%.

[0012] According to the present invention, a method for jointly predicting the health status and charging time of an electric vehicle battery determines the charging mode of the vehicle battery based on the charging range of the remaining battery capacity and the type of charging station used, including: A charging threshold is set for the vehicle battery; when the remaining charge of the vehicle battery exceeds the charging threshold, the charging process is constant voltage charging, and the charging current gradually decreases. The charging mode of the vehicle battery is classified by combining the charging threshold of the vehicle battery with the type of charging pile used; the type of charging pile includes ordinary charging piles and fast charging piles.

[0013] According to the present invention, a method for jointly predicting the health status and charging time of an electric vehicle battery, after determining the charging mode of the vehicle battery based on the charging range of the remaining battery capacity and the type of charging station used, includes: The current capacity of the vehicle battery is determined using the charging data of the vehicle battery and the increment of the remaining battery power. The actual health status of the vehicle battery is determined based on the ratio of its current capacity to its factory-manufactured capacity.

[0014] According to the present invention, a method for jointly predicting the health status and charging time of an electric vehicle battery is provided, which trains the charging time prediction model using the feature dataset and health status estimate of the vehicle battery, including: The feature dataset and health status estimate of the vehicle battery are input into the charging time prediction model to obtain the time prediction value; Using the charging time corresponding to the car battery as a label, the charging time prediction function is fine-tuned with the goal of minimizing the loss function of the charging time prediction model.

[0015] According to the present invention, a method for jointly predicting the battery health status and charging time of an electric vehicle includes inputting a preset charging rule into a trained charging time prediction model to obtain the charging time of the electric vehicle, comprising: The current health status of the vehicle battery is estimated by using the health status prediction model to obtain an estimated health status value. Set the remaining battery power and charging mode at the start and end of charging, and input the estimated health status of the car battery to obtain the charging time of the electric vehicle for this charge.

[0016] Secondly, the present invention also provides a device for jointly predicting the health status and charging time of an electric vehicle battery, comprising: The data acquisition module is used to acquire raw data of electric vehicles and preprocess the raw data to construct a feature dataset of the vehicle battery. The mode determination module is used to determine the charging mode of the vehicle battery based on the charging range of the remaining power of the vehicle battery and the type of charging pile used. The state prediction module is used to call a pre-built health state prediction model and combine it with the feature dataset of the vehicle battery to determine the estimated health state of the vehicle battery. The model training module is used to call a pre-built charging time prediction model and train the charging time prediction model using the feature dataset and health status estimate of the car battery. The time prediction module is used to input the preset charging rules into the trained charging time prediction model to obtain the charging time of the electric vehicle; the charging rules include the remaining battery power at the start and end of charging, the charging mode, and the estimated health status.

[0017] Thirdly, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the joint prediction method for the health status and charging time of an electric vehicle battery as described in the first aspect above.

[0018] In a fourth aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for jointly predicting the state of health of an electric vehicle battery and charging time as described in the first aspect above.

[0019] Fifthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the method for jointly predicting the state of health of an electric vehicle battery and charging time as described in the first aspect above.

[0020] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides a joint prediction method for electric vehicle battery health status and charging time. By constructing a feature dataset to first predict the battery health status as input for charging time prediction, and then predicting the charging time, the accuracy of charging time prediction after battery aging is improved. This fundamentally solves the problem of decreased prediction accuracy caused by battery aging, significantly improves the prediction accuracy throughout the vehicle's entire life cycle, and addresses the problem of poor accuracy in charging time prediction in existing related technologies. Moreover, through data preprocessing techniques, the charging process is abstracted into pre-obtainable charging mode features. Based on charging mode standards, charging modes are divided into different modes, avoiding direct reliance on the difficult-to-obtain charging current and simplifying the modeling process. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0022] Figure 1 This is a flowchart of the method for jointly predicting the health status and charging time of electric vehicle batteries provided by the present invention. Figure 2 This is a schematic diagram illustrating the process of predicting charging time in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the process of dividing charging modes in an embodiment of the present invention; Figure 4This is a structural block diagram of the electric vehicle battery health status and charging time joint prediction device provided by the present invention. Figure 5 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0024] Currently, while model-based methods can reflect the internal physical characteristics of batteries, they rely on precise battery mechanisms and parameters. These methods are typically designed for specific battery types, have poor applicability, and high computational complexity, making them difficult to widely implement in practical applications. Data-driven methods offer high accuracy in charging time prediction, but their accuracy heavily depends on the quality of the dataset and the selected feature parameters. Existing data-driven methods do not consider the impact of battery aging on charging time. Battery aging leads to reduced battery capacity and increased internal resistance, thus affecting charging time; therefore, it is essential to consider the impact of battery aging. On the other hand, charging time is closely related to the charging mode, and the charging current parameter is affected by multiple factors such as the current state of the battery and the type of charging station, often making it difficult to obtain accurately in advance. The charging time difference between fast and slow charging modes is significant; if the charging mode cannot be effectively distinguished, the reliability of the prediction will be further reduced.

[0025] To address the aforementioned technical problems, this invention provides a method for jointly predicting the health status and charging time of electric vehicle batteries. Figure 1 This is a flowchart of the joint prediction method for electric vehicle battery health status and charging time provided by the present invention, as shown below. Figure 1 As shown, the method includes the following steps: Step S101: Obtain the raw data of the electric vehicle and preprocess the raw data to construct a feature dataset of the car battery. Step S102: Determine the charging mode of the car battery based on the charging range of the remaining battery power and the type of charging station used. Step S103: Invoke the pre-built health status prediction model and combine it with the feature dataset of the car battery to determine the estimated value of the State of Health (SOH) of the car battery; Step S104: Call the pre-built charging time prediction model and train the charging time prediction model using the feature dataset and health status estimate of the car battery. Step S105: Input the preset charging rules into the trained charging time prediction model to obtain the charging time of the electric vehicle; the charging rules include the remaining battery charge (State of Charge, SOC) at the start and end of charging, the charging mode, and the estimated health status.

[0026] For example, firstly, the raw data of the electric vehicle is analyzed, and outlier data is preprocessed. Then, charging segment data of the vehicle battery is extracted to calculate the health status and charging time of the charging segment. To more comprehensively describe the factors affecting aging and battery charging time, features such as State of Charge (SOC), cumulative mileage, storage days, and battery temperature are extracted from the raw vehicle battery data to construct a feature dataset. Next, the charging mode is determined based on the SOC's charging range and the type of charging station used. Then, a joint prediction framework is constructed, and a health status prediction model is used to output a health status estimate. Finally, this health status estimate is used as a key input feature and introduced into the charging time prediction model to achieve high-precision charging time prediction.

[0027] In the above process, by constructing a feature dataset to first predict the battery health status as input for charging time prediction, and then predicting the charging time, the accuracy of charging time prediction after battery aging is improved. This fundamentally solves the problem of decreased prediction accuracy caused by battery aging, significantly improves the prediction accuracy throughout the vehicle's entire life cycle, and addresses the problem of poor accuracy in charging time prediction in existing related technologies. Moreover, through data preprocessing techniques, the charging process is abstracted into pre-obtainable charging mode features. Based on charging mode standards, charging modes are divided into different modes, avoiding direct reliance on the difficult-to-obtain charging current and simplifying the modeling process.

[0028] Figure 2 This is a schematic diagram illustrating the charging time prediction process in an embodiment of the present invention, as shown below. Figure 2 As shown, in some embodiments, step S101 involves acquiring the raw data of the electric vehicle and preprocessing the raw data, including: acquiring the raw data of the electric vehicle and removing irrelevant data from the raw data, extracting key data fields related to the electric vehicle's battery; processing abnormal data in the key data fields; abnormal data includes duplicate values, missing values, current exceeding the range, abnormal mileage jumps, disordered time sequence, and charging interruption data; determining the vehicle status through the charging status field of the electric vehicle battery, filtering out charging data segments, and calculating the remaining battery power and charging time at the start and end times of charging the electric vehicle battery.

[0029] Based on this, step S101 involves constructing a feature dataset for the vehicle battery, including: extracting the remaining battery capacity features of the vehicle battery and determining the frequency of the remaining battery capacity distribution in different intervals during charging; the remaining battery capacity features include the remaining battery capacity at the start of charging, the remaining battery capacity at the end of charging, and the increment of the remaining battery capacity during charging; extracting the cumulative mileage from the original data of the electric vehicle; determining the number of days the electric vehicle has been used and the number of days it has been idle; obtaining the current ambient temperature and weather temperature distribution of the electric vehicle; and determining the voltage inconsistency and temperature inconsistency of the individual battery cells.

[0030] Specifically, extract the battery's SOC feature, that is, the SOC value at the start of charging. The SOC value at the end of charging and the increase in SOC during charging As a feature, the frequency of SOC distribution within three intervals (0-40%, 40-80%, and above 80%) during charging is calculated cumulatively, and the SOC distribution is... as follows:

[0031] in, , , This represents the frequency in three different intervals.

[0032] Days of use D d and number of days of suspension D s The relationship is as follows:

[0033]

[0034] in, Indicates the number of days the item was left untouched. Indicates the sampling time of the current data. Indicates the initial sampling time of vehicle data. Indicates the number of days used. N This represents the number of data samples taken at the current moment. Dividing by 8640 is because electric vehicles only record data while driving, and 8640 data points can be collected in 24 hours of driving.

[0035] Current ambient temperature T c and weather temperature distribution T wThe ERA5 reanalysis meteorological data website was used to obtain temperatures for the entire year of 2022 from various locations, with a sampling interval of 1 hour. The ambient temperature at the start of electric vehicle charging was taken as the current ambient temperature. T c The temperature was calculated cumulatively in three intervals (below 10℃, 10-30℃, and above 30℃) on an hourly basis, and the temperature distribution was... T w as follows:

[0036] in, Indicates temperature distribution. , and This represents the temperature in three different ranges.

[0037] Determining the voltage and temperature inconsistencies of individual automotive battery cells includes: characterizing the voltage inconsistency using the standard deviation of the voltage at 85% battery capacity; and characterizing the temperature inconsistency using the standard deviation of the temperature at 75% battery capacity. The specific formulas are as follows:

[0038]

[0039] in, This indicates inconsistent voltage. This indicates inconsistent temperature. This represents the voltage of each individual cell at the sampling point. This represents the average voltage of a single cell at the sampling point. Indicates the number of cells with different voltages. This indicates the sampling temperature between each individual battery cell at the sampling point. This represents the average temperature of a single cell at the sampling point. This indicates the number of temperature probes for a single cell. Since the number of temperature probes is less than the number of individual cells, this... and They are not the same.

[0040] Figure 3 This is a schematic diagram illustrating the process of dividing charging modes in an embodiment of the present invention, as shown below. Figure 3As shown, in some embodiments, step S102, determining the charging mode of the vehicle battery based on the charging range of the remaining battery power and the type of charging pile used, includes: setting a charging threshold for the vehicle battery; when the remaining battery power exceeds the charging threshold, the charging process is constant voltage charging, and the charging current gradually decreases; combining the charging threshold of the vehicle battery with the type of charging pile used, classifying the charging mode of the vehicle battery; the charging pile type includes ordinary charging piles and fast charging piles.

[0041] For example, based on the SOC charging range, the charging threshold is set to 90%. When the SOC is greater than 90%, constant voltage charging will occur, the current will continuously decrease, and the charging speed will slow down. Based on the SOC range and the type of charging station selected, the charging modes are divided into the following four modes, with the following rules:

[0042] in, L c Indicates the charging mode. This represents the SOC value at the end of the charging process.

[0043] Based on the above embodiments, step S102, after determining the charging mode of the car battery according to the charging range of the remaining power of the car battery and the type of charging pile used, includes: determining the current capacity of the car battery using the charging data of the car battery and the increment of the remaining power of the battery; and determining the actual value of the health status of the car battery according to the ratio of the current capacity of the car battery to the factory capacity.

[0044] For example, using partial charging data and the increment of SOC to calculate the current capacity of the battery pack can be described as follows:

[0045] in, Indicates the current capacity of the battery pack. express SOC at any moment express SOC at any moment This indicates the charging current during charging. Indicates that during charging from Time's up The increment of SOC at any given time. The ratio of the battery's current capacity to its original manufacturing capacity is the battery's SOH. Based on this, a predictive model for the SOH of electric vehicles is constructed. As input, a health status prediction model is built, such as the CatBoost model, whose input... x The relationship between the predicted output and the actual output is as follows:

[0046] in, For the output of the model, Indicates the weights of the decision tree. The computational factors of the decision tree are represented. This represents the output function of each decision tree. M This indicates the number of decision trees.

[0047] In some embodiments, step S104, training the charging time prediction model using the feature dataset and health status estimate of the vehicle battery, includes: inputting the feature dataset and health status estimate of the vehicle battery into the charging time prediction model to obtain the time prediction value; using the charging time corresponding to the vehicle battery as a label, and fine-tuning the charging time prediction function with the goal of minimizing the loss function of the charging time prediction model.

[0048] For example, after the above feature processing and SOH prediction process, using The charging time prediction model is trained by using the input of the input. X input This represents the model input for the charging time prediction model. This indicates the SOC value at the start of charging. This represents the SOC value at the end of charging. This indicates the increment of SOC during charging.

[0049] Based on the above embodiments, step S105 involves inputting the preset charging rules into the trained charging time prediction model to obtain the charging time of the electric vehicle, including: estimating the current health status of the vehicle battery through the health status prediction model to obtain a health status estimate; setting the remaining battery power and charging mode at the start and end of charging, and inputting the vehicle battery health status estimate to obtain the charging time of the electric vehicle for this charge.

[0050] In summary, this method incorporates battery state of harm (SOH) as a core input into charging time prediction, addressing the issue of decreased prediction accuracy due to battery aging and significantly improving prediction accuracy throughout the vehicle's lifecycle, especially during battery aging. Replacing precise charging current parameters with charging level reduces reliance on high-precision real-time data, enhancing the model's engineering feasibility and applicability. This method balances prediction accuracy and practicality, providing reliable technical support for charging scheduling optimization, vehicle operation management, and user experience improvement.

[0051] This invention provides a device for jointly predicting the health status and charging time of an electric vehicle battery. The device for jointly predicting the health status and charging time of an electric vehicle battery provided by this invention will be described below. The device for jointly predicting the health status and charging time of an electric vehicle battery described below can be referred to in correspondence with the method for jointly predicting the health status and charging time of an electric vehicle battery described above. Figure 4 This is a structural block diagram of the electric vehicle battery health status and charging time joint prediction device provided by the present invention, as shown in the figure. Figure 4 As shown, the device includes: The data acquisition module 401 is used to acquire the raw data of the electric vehicle and preprocess the raw data to construct a feature dataset of the car battery. The mode determination module 402 is used to determine the charging mode of the vehicle battery based on the charging range of the remaining power of the vehicle battery and the type of charging pile used. The state prediction module 403 is used to call a pre-built health state prediction model and combine it with the feature dataset of the car battery to determine the estimated health state of the car battery. The model training module 404 is used to call a pre-built charging time prediction model and train the charging time prediction model using the feature dataset and health status estimate of the car battery. The time prediction module 405 is used to input the preset charging rules into the trained charging time prediction model to obtain the charging time of the electric vehicle; the charging rules include the remaining battery power at the start and end of charging, the charging mode, and the estimated health status.

[0052] For example, firstly, the data acquisition module 401 analyzes the raw data of the electric vehicle and preprocesses any abnormal data. Then, it extracts charging segment data from the vehicle battery to calculate the health status and charging time of the charging segment. To more comprehensively describe the factors affecting aging and battery charging time, it extracts features such as SOC (State of Charge) characteristics, cumulative mileage, storage days, and battery temperature from the raw data of the vehicle battery to construct a feature dataset. The mode determination module 402 then determines the charging mode based on the charging range of the SOC and the type of charging station used. Next, the state prediction module 403 constructs a joint prediction framework and outputs a health status estimate using the health status prediction model. Finally, the model training module 404 trains the charging time prediction model, and the time prediction module 405 uses this health status estimate as a key input feature, introducing it into the charging time prediction model to ultimately achieve high-precision charging time prediction.

[0053] In the above process, by constructing a feature dataset to first predict the battery health status as input for charging time prediction, and then predicting the charging time, the accuracy of charging time prediction after battery aging is improved. This fundamentally solves the problem of decreased prediction accuracy caused by battery aging, significantly improves the prediction accuracy throughout the vehicle's entire life cycle, and addresses the problem of poor accuracy in charging time prediction in existing related technologies. Moreover, through data preprocessing techniques, the charging process is abstracted into pre-obtainable charging mode features. Based on charging mode standards, charging modes are divided into different modes, avoiding direct reliance on the difficult-to-obtain charging current and simplifying the modeling process.

[0054] Figure 5 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 5 As shown, the electronic device may include: a processor 501, a communication interface 502, a memory 503, and a communication bus 504, wherein the processor 501, the communication interface 502, and the memory 503 communicate with each other via the communication bus 504. The processor 501 can call logical instructions in the memory 503 to execute a joint prediction method for the electric vehicle battery health status and charging time, the method including: Obtain the raw data of electric vehicles, preprocess the raw data, and construct a feature dataset of car batteries; The charging mode of the car battery is determined based on the charging range of the remaining battery power and the type of charging station used. The pre-built health status prediction model is invoked, and the estimated health status of the vehicle battery is determined by combining it with the feature dataset of the vehicle battery. The pre-built charging time prediction model is invoked and trained using the feature dataset and health status estimate of the car battery. The preset charging rules are input into the trained charging time prediction model to obtain the charging time of the electric vehicle. The charging rules include the remaining battery power at the start and end of charging, the charging mode, and the estimated health status.

[0055] Furthermore, the logical instructions in the aforementioned memory 503 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0056] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the joint prediction method for the health status and charging time of an electric vehicle battery provided by the above methods. The method includes: Obtain the raw data of electric vehicles, preprocess the raw data, and construct a feature dataset of car batteries; The charging mode of the car battery is determined based on the charging range of the remaining battery power and the type of charging station used. The pre-built health status prediction model is invoked, and the estimated health status of the vehicle battery is determined by combining it with the feature dataset of the vehicle battery. The pre-built charging time prediction model is invoked and trained using the feature dataset and health status estimate of the car battery. The preset charging rules are input into the trained charging time prediction model to obtain the charging time of the electric vehicle. The charging rules include the remaining battery power at the start and end of charging, the charging mode, and the estimated health status.

[0057] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for jointly predicting the state of health of an electric vehicle battery and charging time provided by the methods described above, the method comprising: Obtain the raw data of electric vehicles, preprocess the raw data, and construct a feature dataset of car batteries; The charging mode of the car battery is determined based on the charging range of the remaining battery power and the type of charging station used. The pre-built health status prediction model is invoked, and the estimated health status of the vehicle battery is determined by combining it with the feature dataset of the vehicle battery. The pre-built charging time prediction model is invoked and trained using the feature dataset and health status estimate of the car battery. The preset charging rules are input into the trained charging time prediction model to obtain the charging time of the electric vehicle. The charging rules include the remaining battery power at the start and end of charging, the charging mode, and the estimated health status.

[0058] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0059] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for jointly predicting the state of health of an electric vehicle battery and its charging time, characterized in that, include: Obtain the raw data of electric vehicles and preprocess the raw data to construct a feature dataset of the vehicle battery; The charging mode of the vehicle battery is determined based on the charging range of the remaining power of the vehicle battery and the type of charging station used. A pre-built health status prediction model is invoked, and the estimated health status of the vehicle battery is determined by combining it with the feature dataset of the vehicle battery. A pre-built charging time prediction model is invoked and trained using the feature dataset and health status estimate of the vehicle battery. The preset charging rules are input into the trained charging time prediction model to obtain the charging time of the electric vehicle; the charging rules include the remaining battery power at the start and end of charging, the charging mode, and the estimated health status.

2. The method for jointly predicting the health status and charging time of an electric vehicle battery according to claim 1, characterized in that, Acquire raw data of the electric vehicle and preprocess the raw data, including: Obtain the raw data of the electric vehicle, remove irrelevant data from the raw data, and extract key data fields related to the electric vehicle's battery; The abnormal data in the key data fields are processed; the abnormal data includes duplicate values, missing values, current out of range, abnormal mileage jumps, disordered time sequence, and charging interruption data. The vehicle status is determined by the charging status field of the vehicle battery, and the charging data segment is filtered out to calculate the remaining battery power and charging time at the start and end of the vehicle battery charging.

3. The method for jointly predicting the health status and charging time of an electric vehicle battery according to claim 1, characterized in that, Construct a feature dataset for automotive batteries, including: Extract the remaining battery capacity characteristics of the vehicle battery and determine the frequency of the remaining battery capacity distribution in different intervals during charging; the remaining battery capacity characteristics include the remaining battery capacity at the start of charging, the remaining battery capacity at the end of charging, and the increment of the remaining battery capacity during charging; Extract the cumulative mileage from the raw data of the electric vehicle; Determine the number of days the electric vehicle was used and the number of days it was left idle; The ambient temperature and weather temperature distribution of the current environment of the electric vehicle are obtained; The voltage and temperature inconsistencies of the individual automotive battery cells were determined.

4. The method for jointly predicting the health status and charging time of an electric vehicle battery according to claim 3, characterized in that, Determining the voltage and temperature inconsistencies of the individual battery cells in the vehicle includes: The standard deviation of the voltage at the moment when the remaining charge of the car battery is 85% is used to characterize the voltage inconsistency of the individual cells. The temperature inconsistency of the individual cells is characterized by the standard deviation of the temperature at which the remaining charge of the car battery is 75%.

5. The method for jointly predicting the health status and charging time of an electric vehicle battery according to claim 1, characterized in that, Based on the charging range of the remaining battery capacity and the type of charging station used, the charging mode of the vehicle battery is determined, including: A charging threshold is set for the vehicle battery; when the remaining charge of the vehicle battery exceeds the charging threshold, the charging process is constant voltage charging, and the charging current gradually decreases. The charging mode of the vehicle battery is classified by combining the charging threshold of the vehicle battery with the type of charging pile used; the type of charging pile includes ordinary charging piles and fast charging piles.

6. The method for jointly predicting the health status and charging time of an electric vehicle battery according to claim 3, characterized in that, After determining the charging mode of the vehicle battery based on its remaining charge range and the type of charging station used, the process includes: The current capacity of the vehicle battery is determined using the charging data of the vehicle battery and the increment of the remaining battery power. The actual health status of the vehicle battery is determined based on the ratio of its current capacity to its factory-manufactured capacity.

7. The method for jointly predicting the health status and charging time of an electric vehicle battery according to claim 2, characterized in that, The charging time prediction model is trained using the feature dataset and health status estimate of the vehicle battery, including: The feature dataset and health status estimate of the vehicle battery are input into the charging time prediction model to obtain the time prediction value; Using the charging time corresponding to the car battery as a label, the charging time prediction function is fine-tuned with the goal of minimizing the loss function of the charging time prediction model.

8. The method for jointly predicting the health status and charging time of an electric vehicle battery according to claim 1, characterized in that, The preset charging rules are input into the trained charging time prediction model to obtain the charging time of the electric vehicle, including: The current health status of the vehicle battery is estimated by using the health status prediction model to obtain an estimated health status value. Set the remaining battery power and charging mode at the start and end of charging, and input the estimated health status of the car battery to obtain the charging time of the electric vehicle for this charge.

9. A device for jointly predicting the health status and charging time of an electric vehicle battery, characterized in that, include: The data acquisition module is used to acquire raw data of electric vehicles and preprocess the raw data to construct a feature dataset of the vehicle battery. The mode determination module is used to determine the charging mode of the vehicle battery based on the charging range of the remaining power of the vehicle battery and the type of charging pile used. The state prediction module is used to call a pre-built health state prediction model and combine it with the feature dataset of the vehicle battery to determine the estimated health state of the vehicle battery. The model training module is used to call a pre-built charging time prediction model and train the charging time prediction model using the feature dataset and health status estimate of the car battery. The time prediction module is used to input the preset charging rules into the trained charging time prediction model to obtain the charging time of the electric vehicle; the charging rules include the remaining battery power at the start and end of charging, the charging mode, and the estimated health status.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method for jointly predicting the health status and charging time of an electric vehicle battery as described in any one of claims 1 to 8.

Images