Vehicle-mounted battery pack health state detection method, device, equipment and product

Abstract

The invention discloses a vehicle-mounted battery pack health state detection method, device, equipment and product, and relates to the technical field of lithium ion batteries, and the method comprises the steps: obtaining the current charging data of a vehicle; performing feature extraction on the current charging data to obtain current feature information; according to the current feature information, a battery pack health state detection model is adopted to determine the current health state of a battery pack of the vehicle; the battery pack health state detection model is a machine learning model built in advance according to a training sample set based on an Optuna framework; the training sample set comprises multiple groups of historical feature information and corresponding historical health states. According to the invention, the Optuna framework is utilized to optimize the machine learning model, the vehicle-mounted battery pack health state is detected, the complexity of the model is reduced, and the vehicle-mounted battery pack health state detection precision is improved.

Description

Technical Field

[0001] The present application relates to the technical field of lithium-ion batteries, and in particular to a method, device, equipment and product for detecting the health status of a vehicle-mounted battery pack. Background Art

[0002] In recent years, electric vehicles have developed rapidly due to their energy-saving and environmentally friendly advantages, and lithium-ion battery packs are one of their core components. However, lithium-ion battery packs are easily affected by the working environment. If there are drastic changes in the external environment, lithium-ion battery packs will undergo an irreversible performance degradation process, and in severe cases, they may even cause safety accidents such as fire and explosion, resulting in loss of personnel and property. For safety reasons, it is very important to monitor the status of the battery pack, and the health status of the lithium-ion battery pack is one of the important indicators. This indicator can reflect the degree of aging of the battery pack during use. The battery pack management system uses this as a basis to guide the charging and discharging decisions during the use of the battery pack, which can extend the service life of the battery pack to a certain extent and reduce the cost of using the battery pack.

[0003] At present, the detection methods of the health status of battery packs mainly include methods based on experimental measurement and methods based on equivalent models. Among them, the experimental measurement method aims to use professional experimental instruments to conduct experiments such as charging and discharging on the battery pack, and directly measure some physical quantities exhibited by the battery pack during the experiment. However, the experimental measurement method requires professionals to use experimental instruments with high precision to measure, and has very high requirements for accuracy, which is often difficult to meet under actual working conditions; the equivalent models of battery packs mainly include electrochemical models and equivalent circuit models. The methods based on electrochemical models generally have problems such as difficulty in solving and large amount of calculation due to complex partial differential equations, and the number of parameters is large, making it difficult to accurately obtain parameter values. The parameters of the equivalent circuit model of the battery pack are derived from specific experimental conditions. When the model is applied under other conditions, the accuracy will be significantly reduced. Summary of the invention

[0004] The purpose of this application is to provide a method, device, equipment and product for detecting the health status of a vehicle-mounted battery pack, so as to improve the accuracy of detecting the health status of a vehicle-mounted battery pack.

[0005] To achieve the above objectives, this application provides the following solutions:

[0006] In a first aspect, the present application provides a method for detecting the health status of a vehicle battery pack, comprising:

[0007] Get the current charging data of the vehicle;

[0008] Extracting features from the current charging data to obtain current feature information;

[0009] According to the current feature information, a battery pack health status detection model is used to determine the current health status of the vehicle's battery pack; the battery pack health status detection model is a machine learning model pre-built based on the Optuna framework according to a training sample set; the training sample set includes multiple sets of historical feature information and corresponding historical health status.

[0010] Optionally, before extracting features from the current charging data to obtain current feature information, the vehicle-mounted battery pack health status detection method further includes:

[0011] The current charging data is supplemented with missing data and deleted with erroneous data.

[0012] Optionally, the current characteristic information includes battery pack health characteristics.

[0013] Optionally, the current charging data includes the charged power and voltage of the battery pack at each sampling point in the constant current charging section; the battery pack health characteristics include the number of each interval and the peak characteristics in each interval;

[0014] Extracting features from the current charging data to obtain current feature information specifically includes:

[0015] According to the charged power and voltage of the battery pack at each sampling point in the constant current charging section, the capacity increment curve of the vehicle constant current charging section is obtained;

[0016] Dividing the capacity increment curve according to the number and distribution of peaks in the capacity increment curve to obtain a plurality of intervals;

[0017] Numbering each of the intervals in sequence;

[0018] Extract the peak features in each interval; the peak features include the maximum value of the capacity increment and the voltage value corresponding to the maximum value.

[0019] Optionally, the current charging data includes the cumulative mileage of the vehicle during charging, the charging time of the constant current charging section, the voltage, current and charge state of the battery pack at each sampling point in the constant current charging section;

[0020] The current characteristic information includes the charging time of the vehicle constant current charging section, the initial charging voltage, the average charging current, the initial state of charge, the final state of charge and the accumulated mileage;

[0021] The initial charging voltage is the voltage of the battery pack at the first sampling point in the constant current charging section;

[0022] The average charging current is the average value of the current at all sampling points of the battery pack in the constant current charging section;

[0023] The initial state of charge is the state of charge of the battery pack at the first sampling point in the constant current charging section;

[0024] The final state of charge is the state of charge of the battery pack at the last sampling point in the constant current charging section.

[0025] Optionally, the process of acquiring the training sample set includes:

[0026] Obtain historical charging data of the vehicle during multiple historical charging processes;

[0027] For any historical charging process of the vehicle, feature extraction is performed on historical charging data of the historical charging process to obtain historical feature information corresponding to the historical charging process;

[0028] Obtaining the historical charged power and the historical state of charge of the battery pack at each sampling point in the constant current charging section during the historical charging process;

[0029] According to the historical state of charge, dividing the constant current charging section in the historical charging process into a plurality of charging sections;

[0030] Calculating the battery pack capacity of each charging segment in the historical charging process according to the historical charged power and the historical state of charge;

[0031] According to the battery pack capacity of each charging segment in each historical charging process, the historical health state corresponding to each historical charging process is calculated.

[0032] Optionally, according to the battery pack capacity of each charging segment in each historical charging process, a historical health state corresponding to each historical charging process is calculated, specifically including:

[0033] For any historical charging process of the vehicle, outliers are removed from the battery pack capacity in the historical charging process to obtain a candidate battery pack capacity corresponding to the historical charging process;

[0034] averaging the candidate battery pack capacities to obtain an estimated battery pack capacity corresponding to the historical charging process;

[0035] The estimated battery pack capacity corresponding to all historical charging processes is subjected to random noise removal to obtain the actual battery pack capacity corresponding to each historical charging process;

[0036] The historical health state corresponding to each historical charging process is calculated according to the actual battery pack capacity corresponding to each historical charging process.

[0037] In a second aspect, the present application provides a vehicle-mounted battery pack health status detection device, comprising:

[0038] A data acquisition module, used to acquire the current charging data of the vehicle;

[0039] A feature extraction module, used to extract features from the current charging data to obtain current feature information;

[0040] A battery pack health status detection module is used to determine the current health status of the vehicle's battery pack based on the current feature information and using a battery pack health status detection model; the battery pack health status detection model is a machine learning model pre-built based on the Optuna framework according to a training sample set; the training sample set includes multiple sets of historical feature information and corresponding historical health status.

[0041] In a third aspect, the present application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned vehicle battery pack health status detection method.

[0042] In a fourth aspect, the present application provides a computer program product, including a computer program, which implements the above-mentioned vehicle battery pack health status detection method when executed by a processor.

[0043] According to the specific embodiments provided in this application, this application has the following technical effects:

[0044] The present application provides a method, device, equipment and product for detecting the health status of a vehicle-mounted battery pack. By extracting feature information and building a machine learning model, the complexity of the battery pack health status detection model is reduced; the hyperparameters of the machine learning model are optimized in combination with the Optuna framework, thereby further improving the accuracy of the vehicle-mounted battery pack health status detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 This is an application environment diagram of a vehicle-mounted battery pack health status detection method in one embodiment of the present application;

[0047] Figure 2 A flowchart of a method for detecting the health status of a vehicle battery pack provided in one embodiment of the present application;

[0048] Figure 3 This is a flowchart of missing data supplementation processing in one embodiment of the present application;

[0049] Figure 4 It is an average current distribution diagram of a constant current charging section of multiple sets of historical charging data in one embodiment of the present application;

[0050] Figure 5 It is a statistical diagram of average current in constant current charging section of multiple groups of historical charging data in one embodiment of the present application;

[0051] Figure 6 A graph showing the relationship between the amount of electricity charged and the voltage in a set of historical charging data in an embodiment of the present application;

[0052] Figure 7 A capacity increment curve diagram of a set of historical charging data in an embodiment of the present application;

[0053] Figure 8 A battery pack capacity distribution diagram of each charging segment in multiple groups of historical charging data in an embodiment of the present application;

[0054] Fig. 9 This is a schematic diagram of the principle of a box diagram in one embodiment of the present application;

[0055] Fig.10 A box diagram of the battery pack capacity of a portion of the historical charging process in an embodiment of the present application;

[0056] Fig.11 A scatter plot of the estimated battery pack capacity corresponding to all historical charging processes in an embodiment of the present application;

[0057] Fig.12 The actual battery pack capacity corresponding to each historical charging process in an embodiment of the present application;

[0058] Fig.13 This is a diagram showing the health status detection result of the vehicle-mounted battery pack in one embodiment of the present application;

[0059] Fig.14 is the relative error of the vehicle battery pack health status detection model in one embodiment of the present application;

[0060] Fig.15 A schematic diagram of functional modules of a vehicle-mounted battery pack health status detection device provided in one embodiment of the present application;

[0061] Fig.16 A schematic diagram of the structure of a computer device provided in one embodiment of the present application.

[0062] Reference numerals: 1501 - data acquisition module, 1502 - feature extraction module, 1503 - battery pack health status detection module, 1504 - preprocessing module. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0064] This application proposes a method, device, equipment and product for detecting the health status of a vehicle battery pack. The Optuna framework is used to optimize the machine learning model to detect the health status of the vehicle battery pack, thereby reducing the complexity of the model and improving the accuracy of the model.

[0065] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0066] The vehicle battery pack health status detection method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the vehicle charging data to the server 104. After receiving the vehicle charging data, the server 104 extracts the features of the charging data to obtain feature information; according to the feature information, the battery pack health status detection model is used to determine the health status of the battery pack of the vehicle. The server 104 can feedback the obtained vehicle battery pack health status to the terminal 102. In addition, in some embodiments, the vehicle-mounted battery pack health status detection method can also be implemented separately by the server 104 or the terminal 102, such as the terminal 102 can directly perform battery pack health status detection on the vehicle charging data, or the server 104 can obtain the vehicle charging data from the data storage system and detect the battery pack health status.

[0067] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.

[0068] In an exemplary embodiment, Figure 2As shown, a method for detecting the health status of a vehicle battery pack is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used as an example to illustrate, including the following steps 201 to 203. Among them:

[0069] Step 201, obtaining the current charging data of the vehicle. The current charging data may include the charged power and voltage of the battery pack at each sampling point in the constant current charging section. The current charging data may also include the accumulated mileage when the vehicle is charging, the charging time of the constant current charging section, the current and the state of charge of the battery pack at each sampling point in the constant current charging section.

[0070] Step 202: extract features from current charging data to obtain current feature information.

[0071] In an exemplary embodiment, the current characteristic information includes the charging time, initial charging voltage, average charging current, initial state of charge, final state of charge, and accumulated mileage of the constant current charging section of the vehicle. The initial charging voltage is the voltage of the battery pack at the first sampling point in the constant current charging section. The average charging current is the average value of the current of the battery pack at all sampling points in the constant current charging section. The initial state of charge is the state of charge of the battery pack at the first sampling point in the constant current charging section. The final state of charge is the state of charge of the battery pack at the last sampling point in the constant current charging section.

[0072] In another exemplary embodiment, the current characteristic information includes a battery pack health characteristic. The battery pack health characteristic includes the number of each interval and the peak characteristic within each interval. According to the charged power and voltage of the battery pack at each sampling point in the constant current charging section, the capacity increment curve of the vehicle constant current charging section is obtained. According to the number and distribution of peaks in the capacity increment curve, the capacity increment curve is divided to obtain multiple intervals. Each interval is numbered in turn. The peak characteristics in each interval are extracted, and the peak characteristics include the maximum value of the capacity increment and the voltage value corresponding to the maximum value.

[0073] In another exemplary embodiment, the current characteristic information also includes the battery pack health characteristics as well as the charging time, initial charging voltage, average charging current, initial state of charge, final state of charge and accumulated mileage of the vehicle constant current charging segment.

[0074] In an exemplary embodiment, Figure 2 As shown, before extracting features from the current charging data to obtain current feature information, the vehicle-mounted battery pack health status detection method may further include:

[0075] Step 204: supplement missing data and delete erroneous data for the current charging data.

[0076] In an exemplary embodiment, data with a cumulative mileage of 0 in the charging data is considered as error data and deleted. Data with extremely large voltage and extremely large current is considered as error data and deleted.

[0077] The data with abnormal sampling time interval in the charging data is regarded as missing data. In an exemplary embodiment, the charging data sampling time interval is 30 seconds, so it is necessary to interpolate and supplement the data with a time interval between adjacent sampling points greater than 60 seconds. This application uses Newton interpolation to supplement missing data. The process of missing data supplementation is as follows: Figure 3 As shown, Figure 3 In the equation, T represents the set of all sampling moments of charging data, and the length of set T is n. Starting from moment T[0], the time interval between the current moment and the next moment is calculated in sequence, that is, T[i]-T[i-1]. If the time interval is greater than 60 seconds, it is determined that data is missing, and the interpolation target x is set to moment T[i-1]+30, and Newton interpolation is performed on the charging data; otherwise, it is determined that there is no data missing, and the situation at the next moment is continued to be judged. This is done until all n sampling moments are judged.

[0078] The data sampling time format of the charging data sequence is a timestamp. For example, one of the sampling times is in the form of 2018-03-2618:56:45, which means 18:56:45 on March 26, 2018. The sampling time in the form of a timestamp cannot be interpolated, so the timestamp needs to be processed. For the sampling time of the charging data of each charging process, the sampling time of the first data point is set to 0 seconds, and the sampling time of the subsequent sampling points is the difference between the timestamp of the sampling point and the timestamp of the first sampling point, and is converted into a value in seconds.

[0079] Step 203, based on the current feature information, a battery pack health status detection model is used to determine the current health status of the vehicle's battery pack. The battery pack health status detection model is a machine learning model pre-built based on the Optuna framework based on a training sample set. The training sample set includes multiple sets of historical feature information and corresponding historical health status.

[0080] The process of obtaining the training sample set includes:

[0081] The historical charging data of the vehicle in multiple historical charging processes are obtained, and for any historical charging process of the vehicle, feature extraction is performed on the historical charging data in the historical charging process to obtain historical feature information corresponding to the historical charging process.

[0082] In an exemplary embodiment, the historical characteristic information includes the charging time, charging initial voltage, average charging current, initial state of charge, final state of charge, and accumulated mileage of the constant current charging segment in the vehicle's historical charging data. The average current distribution of the constant current charging segment in multiple sets of historical charging data is as follows: Figure 4 The average current statistics of the constant current charging section in multiple sets of historical charging data are shown in Figure 5 shown.

[0083] In another exemplary embodiment, the historical characteristic information includes historical battery pack health characteristics. The capacity increment analysis method is applied to the constant current charging segment in the historical charging data of the vehicle to obtain the historical battery pack health characteristics. The historical battery pack health characteristics include the number of each interval in the capacity increment curve of the historical charging data and the historical peak characteristics in each interval. The historical peak characteristics include the maximum value of the capacity increment of the historical charging data and the voltage value corresponding to the maximum value.

[0084] When the vehicle is parked for charging, the initial state of charge of the battery pack is not necessarily 0, and the final state of charge of the constant current charging section is not necessarily high, which makes it easy for the capacity increment curve of the constant current charging section to lack a lot of information. Therefore, the historical charging data with a change of more than 50% between the final state of charge and the initial state of charge of the constant current charging section of the battery pack is taken as valid data. The historical charging data with a small change in the state of charge of the constant current charging section is filtered out.

[0085] like Figure 6 As shown in the figure, due to the short sampling time interval and large battery pack capacity, the voltage changes at adjacent sampling points are small or even unchanged. By fitting the scatter plot with support vector regression, only one power value corresponds to the same voltage, avoiding the situation where the denominator is too small or even 0 when calculating the capacity increment.

[0086] The charge-in capacity-voltage relationship curve is filtered to obtain the capacity increment curve. Figure 7 As shown, Figure 6After Savitzky-Golay filtering, the capacity increment curve obtained has three obvious peaks. However, when the initial state of charge of the constant current charging section of the vehicle battery pack is too large, peaks 1 and 2 may disappear. When the state of charge at the end of the constant current charging section of the vehicle battery pack is too small, peaks 3 and 2 may disappear. To extract health features from the capacity increment curve, the following scheme is adopted: (1) The voltage range of the capacity increment curve is divided into three intervals: (380, 400) V, (400, 420) V, and (435, 450) V; (2) The capacity increment curve is polled in the three voltage intervals to obtain the capacity increment maximum value and its corresponding voltage value; (3) If the capacity increment maximum value is not found in a certain voltage interval, it means that the capacity increment curve is missing in this voltage interval, and its capacity increment maximum value and the corresponding voltage value are set to 0; (4) To describe the voltage interval information where the capacity increment maximum value falls, the three voltage intervals are numbered 1, 2, and 3 in sequence; (5) In the case of missing capacity increment maximum value, the corresponding voltage interval number is set to 0. The capacity increment maximum value in the three voltage intervals, the voltage value corresponding to the capacity increment maximum value, and the voltage interval number are selected as health features.

[0087] In another exemplary embodiment, the historical characteristic information also includes historical battery pack health characteristics and the charging time, initial charging voltage, average charging current, initial state of charge, final state of charge and accumulated mileage of the constant current charging segment in the vehicle's historical charging data.

[0088] Obtain the historical charge input power and historical state of charge of the battery pack at each sampling point in the constant current charging segment during the historical charging process. Divide the constant current charging segment in the historical charging process into several charging segments according to the historical state of charge. Calculate the battery pack capacity of each charging segment in the historical charging process according to the historical charge input power and historical state of charge. Calculate the historical health status corresponding to each historical charging process according to the battery pack capacity of each charging segment in each historical charging process.

[0089] The calculation formula of battery pack health status SOH is as follows:

[0090]

[0091] Among them, SOH is the health status of the battery pack; C N is the capacity of the current battery pack at full charge; C nom It is the rated capacity of the battery pack when it leaves the factory.

[0092] The calculation formula of the battery pack's state of charge SOC is as follows:

[0093]

[0094] Among them, SOC is the state of charge of the battery pack; C rest The remaining power in the battery pack at the current moment.

[0095] During the charging process of the battery pack, the SOC will change within a short period of time. The formula for calculating the change in SOC is as follows:

[0096]

[0097] Among them, ΔSOC is the change in the battery pack state of charge; SOC1 is the battery pack state of charge after the change; SOC2 is the battery pack state of charge before the change; C rest1 is the remaining power in the battery pack corresponding to SOC1; C rest2 is the remaining power in the battery pack corresponding to SOC2; ΔC is the change in the remaining power in the battery pack during this period of time, and its calculation formula is as follows:

[0098] ΔC=∫Idt.

[0099] Where I is the current charged into the battery pack.

[0100] Substituting into the above formula, we get:

[0101]

[0102] In reality, the charging data sequence is a discrete point, so the above formula is discretized to obtain the calculation formula for the capacity of the current battery pack at full charge:

[0103]

[0104] Among them, Q1 is the power of the battery pack corresponding to SOC1; Q2 is the power of the battery pack corresponding to SOC2; ΔQ is the charge of the battery pack during this period of time.

[0105] The above calculation method is based on the ampere-hour integration method during the charging process. Applying it to the capacity calculation of the battery pack requires dividing the vehicle's historical charging data into charging segments. During the battery pack charging process, the state of charge of the battery pack will remain unchanged for a period of time. When the state of charge changes to a new value, the sampling point is used as the starting point to poll the historical charging data sequence. When the state of charge changes again, the sampling point is used as the end point. All sampling points between the above starting point and the end point constitute a charging segment. The constant current charging segment of any historical charging process is divided into several charging segments. The battery pack capacity of each charging segment is calculated based on the historical charged power and the historical state of charge. In an exemplary embodiment, the battery pack capacity distribution of each charging segment in multiple sets of historical charging data is as follows: Figure 8 shown.

[0106] For any historical charging process of the vehicle, outliers are removed from the battery pack capacity in the historical charging process to obtain the candidate battery pack capacity corresponding to this historical charging process. The candidate battery pack capacities are averaged to obtain the estimated battery pack capacity corresponding to this historical charging process. Random noise is removed from the estimated battery pack capacities corresponding to all historical charging processes to obtain the actual battery pack capacity corresponding to each historical charging process. Based on the actual battery pack capacity corresponding to each historical charging process, the historical health status corresponding to each historical charging process is calculated.

[0107] In an exemplary embodiment, first Figure 8 The battery pack capacities that are obviously outliers are deleted, and then outlier detection is performed on the remaining battery pack capacities. Fig. 9 As shown in Figure 2, a box plot is introduced to handle outliers. U and B L The upper and lower limits of the box plot are respectively. Points beyond the upper and lower limits are outliers and need to be removed. The box plot will sort the observations from small to large. Assuming the sequence length is s, the upper quartile Q3 is located at The lower quartile Q1 is the value at When the sequence length s is an odd number, the median Q2 is the value at When the sequence length s is an even number, the median Q2 is the value at and The average value of the value at; the interquartile range IQR = Q3-Q1. The battery pack capacity box diagram of part of the historical charging process is as follows Fig.10 shown.

[0108] In an exemplary embodiment, Fig.11 As shown in the figure, there is a strong random fluctuation in the estimated battery pack capacity change, which is caused by many reasons. Random fluctuations bring random factors to capacity changes, while battery pack aging is not related to random factors, which has a negative impact on the accuracy of the battery pack health status detection model. Therefore, it is necessary to remove the noise caused by random factors.

[0109] In an exemplary embodiment, a moving average method is used to remove random noise. The principle is to select m sample points on both sides of the sample point with sample point i as the center, form a sliding window with a length of 2m+1, and calculate the average of the sample points in the window as the value of sample point i. The sliding window slides backward in the sample sequence, replacing the sample value point by point. The formula for calculating the average value in the sliding window is as follows:

[0110]

[0111] Among them, y(i) is the value at sample point i in the original data; Y(i) is the value at sample point i after moving average; y(im) is the value at sample point im in the original data; y(i+m) is the value at sample point i+m in the original data; 2m+1 is the length of the sliding window.

[0112] The estimated battery pack capacity corresponding to all historical charging processes is subjected to random noise removal, and the actual battery pack capacity corresponding to each historical charging process is obtained as follows: Fig.12 shown.

[0113] In an exemplary embodiment, the present application builds an LSTM-Attention model based on the Optuna framework to detect the health status of the battery pack. Optuna is a lightweight framework that automatically optimizes hyperparameters. It is highly modular and can dynamically set the search space for hyperparameters. Optuna uses an efficient optimization algorithm with low dependencies, and can efficiently prune hopeless experiments; it can share optimization information between different threads to speed up the optimization process. Based on the records of recommended parameter values ​​and target values ​​that need to be evaluated, the Optuna sampler will continuously narrow the search space and eventually find an optimal search space.

[0114] This application sets the number of LSTM layers to 2 and the batch size to 32. Other parameters that need to be set for the model and their lower and upper bounds are shown in Table 1:

[0115] Table 1 Model preset parameters and their upper and lower bounds

[0116] parameter The Nether Upper Bound Number of neurons in LSTM layer 10 100 Number of neurons in the fully connected layer 10 100 Learning Rate 0.0001 0.01 Loss rate 0.2 0.5 Time step 1 20

[0117] Optuna is used to optimize the hyperparameters of the LSTM-Attention model, and the optimized model is used to detect the health status of the vehicle battery pack, such as Fig.13 and Fig.14 As shown, the relative error of the model is controlled within 1%. In addition, the maximum relative error of the model is 0.83%, the average absolute error is 0.00173, and the root mean square error is 0.00231. It can be seen that the health status detection model of the vehicle battery pack of the present application has a high detection accuracy.

[0118] Based on the same inventive concept, the embodiment of the present application also provides a vehicle-mounted battery pack health status detection device for implementing the vehicle-mounted battery pack health status detection method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more vehicle-mounted battery pack health status detection device embodiments provided below can refer to the limitations of the vehicle-mounted battery pack health status detection method above, and will not be repeated here.

[0119] In an exemplary embodiment, Fig.15 As shown, the present application also provides a vehicle-mounted battery pack health status detection device, which includes: a data acquisition module 1501, a feature extraction module 1502 and a battery pack health status detection module 1503. Among them, the data acquisition module 1501 is used to obtain the current charging data of the vehicle. The feature extraction module 1502 is used to extract features from the current charging data to obtain current feature information. The battery pack health status detection module 1503 is used to determine the current health status of the vehicle's battery pack based on the current feature information and using a battery pack health status detection model. The battery pack health status detection model is a machine learning model that is pre-built based on the Optuna framework according to a training sample set. The training sample set includes multiple sets of historical feature information and corresponding historical health status.

[0120] In another exemplary embodiment of the present application, Fig.15 As shown, the vehicle-mounted battery pack health status detection device may also include:

[0121] The pre-processing module 1504 is used to supplement missing data and delete erroneous data for the current charging data.

[0122] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Fig.16 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. 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, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store vehicle charging data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for detecting the health status of a vehicle battery pack is implemented.

[0123] Those skilled in the art will understand that Fig.16The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components. In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0124] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0125] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0126] 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 used 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, and the collection, use and processing of relevant data must comply with relevant regulations.

[0127] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and 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-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. 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), magnetic 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 may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0128] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0129] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, 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, they should be considered to be within the scope of this specification.

[0130] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for detecting the health status of a vehicle battery pack, characterized in that: The vehicle-mounted battery pack health status detection method comprises: Get the current charging data of the vehicle; Extracting features from the current charging data to obtain current feature information; According to the current feature information, a battery pack health status detection model is used to determine the current health status of the vehicle's battery pack; the battery pack health status detection model is a machine learning model pre-built based on the Optuna framework according to a training sample set; the training sample set includes multiple sets of historical feature information and corresponding historical health status.

2. The vehicle-mounted battery pack health status detection method according to claim 1, characterized in that: Before extracting features from the current charging data to obtain current feature information, the vehicle-mounted battery pack health status detection method further includes: The current charging data is supplemented with missing data and deleted with erroneous data.

3. The vehicle-mounted battery pack health status detection method according to claim 1, characterized in that: The current characteristic information includes battery pack health characteristics.

4. The vehicle-mounted battery pack health status detection method according to claim 3, characterized in that: The current charging data includes the charged power and voltage of the battery pack at each sampling point in the constant current charging section; the battery pack health characteristics include the number of each interval and the peak characteristics in each interval; Extracting features from the current charging data to obtain current feature information specifically includes: According to the charged power and voltage of the battery pack at each sampling point in the constant current charging section, the capacity increment curve of the vehicle constant current charging section is obtained; Dividing the capacity increment curve according to the number and distribution of peaks in the capacity increment curve to obtain a plurality of intervals; Numbering each of the intervals in sequence; Extract the peak features in each interval; the peak features include the maximum value of the capacity increment and the voltage value corresponding to the maximum value.

5. The vehicle-mounted battery pack health status detection method according to claim 1, characterized in that: The current charging data includes the accumulated mileage of the vehicle during charging, the charging time of the constant current charging section, the voltage, current and charge state of the battery pack at each sampling point in the constant current charging section; The current characteristic information includes the charging time of the vehicle constant current charging section, the initial charging voltage, the average charging current, the initial state of charge, the final state of charge and the accumulated mileage; The initial charging voltage is the voltage of the battery pack at the first sampling point in the constant current charging section; The average charging current is the average value of the current at all sampling points of the battery pack in the constant current charging section; The initial state of charge is the state of charge of the battery pack at the first sampling point in the constant current charging section; The final state of charge is the state of charge of the battery pack at the last sampling point in the constant current charging section.

6. The vehicle-mounted battery pack health status detection method according to claim 1, characterized in that: The process of obtaining the training sample set includes: Obtain historical charging data of the vehicle during multiple historical charging processes; For any historical charging process of the vehicle, feature extraction is performed on historical charging data of the historical charging process to obtain historical feature information corresponding to the historical charging process; Obtaining the historical charged power and the historical state of charge of the battery pack at each sampling point in the constant current charging section during the historical charging process; According to the historical state of charge, dividing the constant current charging section in the historical charging process into a plurality of charging sections; Calculating the battery pack capacity of each charging segment in the historical charging process according to the historical charged power and the historical state of charge; According to the battery pack capacity of each charging segment in each historical charging process, the historical health state corresponding to each historical charging process is calculated.

7. The vehicle-mounted battery pack health status detection method according to claim 6, characterized in that: According to the battery pack capacity of each charging segment in each historical charging process, the historical health status corresponding to each historical charging process is calculated, including: For any historical charging process of the vehicle, outliers are removed from the battery pack capacity in the historical charging process to obtain a candidate battery pack capacity corresponding to the historical charging process; averaging the candidate battery pack capacities to obtain an estimated battery pack capacity corresponding to the historical charging process; The estimated battery pack capacity corresponding to all historical charging processes is subjected to random noise removal to obtain the actual battery pack capacity corresponding to each historical charging process; The historical health state corresponding to each historical charging process is calculated according to the actual battery pack capacity corresponding to each historical charging process.

8. A vehicle-mounted battery pack health status detection device, applied to the vehicle-mounted battery pack health status detection method according to any one of claims 1 to 7, characterized in that: The vehicle-mounted battery pack health status detection device comprises: A data acquisition module, used to acquire the current charging data of the vehicle; A feature extraction module, used to extract features from the current charging data to obtain current feature information; A battery pack health status detection module is used to determine the current health status of the vehicle's battery pack based on the current feature information and using a battery pack health status detection model; the battery pack health status detection model is a machine learning model pre-built based on the Optuna framework according to a training sample set; the training sample set includes multiple sets of historical feature information and corresponding historical health status.

9. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the vehicle battery pack health status detection method according to any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for detecting the health status of a vehicle battery pack according to any one of claims 1 to 7 is implemented.

Images