A health assessment system, method and electronic device for a power battery

Abstract

This specification discloses a health assessment system, method and electronic device for a power battery, including: a data access module for accessing and storing actual battery state data; a data preprocessing module for preprocessing the actual battery state data; a data calculation module for segmenting charging segments of the actual battery state data and finding charging working conditions; a model evaluation module for using a trained capacity correction model to process the actual charging working conditions and standard working conditions corresponding to a number of actual charging SOC segments, obtaining a working condition correction coefficient based on the processing results, and accordingly correcting the actual charging power of the number of actual charging SOC segments; obtaining an actual charging capacity based on the number of actual charging SOC segments and their corrected charging power; calculating the rated charging capacity of the vehicle based on the initial battery state data in the initial stage of the vehicle operation; and evaluating the battery health SOH using the ratio of the actual charging capacity to the rated charging capacity.

Description

Technical Field

[0001] The present specification relates to the field of vehicle technology, and in particular to a health assessment system, method and electronic equipment for a power battery. Background Art

[0002] With the rapid growth of new energy vehicle ownership, power batteries, as the core components of new energy vehicles, are also attracting more and more attention. As we all know, with the use of power batteries, their capacity is gradually declining, and when the battery capacity declines significantly, it will significantly affect the user's driving experience. For example, the battery runs out of power after a short driving distance, and the number of charging times becomes more and more frequent. Therefore, it is of great significance to accurately evaluate the health of the battery.

[0003] At present, the health of batteries is generally evaluated based on the actual capacity of the power battery and the calibrated rated capacity. However, the charging scenarios of power batteries are complex and changeable. For example, the calculation of capacity value is related to the current size, temperature and other working conditions. However, the existing technology does not consider the impact of actual working conditions, resulting in large errors in battery health evaluation. Summary of the invention

[0004] This specification provides a health assessment system, method and electronic device for a power battery. By dividing the actual state data of the battery into charging segments, a number of actual charging SOC segments are obtained and their corresponding charging conditions are found. On the one hand, the charging capacity is corrected by referring to the charging conditions using a capacity correction model, thereby eliminating the impact of capacity errors caused by different charging conditions and improving the accuracy of the assessment. On the other hand, the actual charging capacity of the battery is comprehensively calculated using a number of charging SOC segments of the cumulative charging period, so that the health assessment of the power battery in this specification does not only rely on one charging segment, thereby eliminating the error fluctuations caused by a single segment and making the assessment results more accurate.

[0005] In order to solve the above technical problems, this specification provides a power battery health assessment system, the system comprising:

[0006] A data access module is used to access and store the actual status data of the battery;

[0007] A data preprocessing module, used to preprocess the actual state data of the battery when the cumulative charging time of the power battery reaches a cumulative time threshold, wherein the preprocessing includes: data analysis, data conversion, and data cleaning;

[0008] A data calculation module is used to divide the actual battery status data into charging segments to obtain a plurality of actual charging SOC segments; and to search the charging conditions corresponding to the plurality of actual charging SOC segments from the actual battery status data;

[0009] A model evaluation module, which is used to process the actual charging conditions and standard conditions corresponding to the several actual charging SOC segments by using the trained capacity correction model, and obtain the working condition correction coefficients corresponding to the several actual charging SOC segments under the standard condition based on the processing results; correct the actual charging power corresponding to the several actual charging SOC segments by using the working condition correction coefficients corresponding to the several actual charging SOC segments; obtain the actual charging capacity based on the several actual charging SOC segments and their corrected charging power; calculate the rated charging capacity of the vehicle based on the initial battery state data at the initial stage of the vehicle operation; and evaluate the battery health SOH by using the ratio of the actual charging capacity to the rated charging capacity.

[0010] Preferably, the actual battery state data includes: vehicle identification, reporting time, vehicle state, charge and discharge state, total voltage, total current, SOC, temperature;

[0011] The data preprocessing module specifically includes:

[0012] A data parsing module, which is used to parse the actual battery state data from unstructured data into structured data;

[0013] A data conversion module, which is used to perform unit conversion and service data conversion on the actual battery state data;

[0014] A data cleaning module, which is used to repair and / or delete abnormal points in the actual battery state data.

[0015] Preferably, when repairing abnormal points, the data cleaning module is specifically used to repair the actual battery state data at the intermediate moment to the actual battery state data at the front and back moments when the first time differences between the intermediate moment and the front and back moments are both less than the first difference threshold;

[0016] When repairing abnormal points, the data cleaning module is specifically used to:

[0017] Delete the first outliers in the actual battery state data that exceed the corresponding value range;

[0018] Traverse the actual battery state data in turn by means of derivation to obtain the adjacent change amount between the actual battery state data and the adjacent data before and after; and use the change amount threshold to detect the adjacent change amount to determine the second outlier or the outlier interval corresponding to the second outlier; use a preset value as a judgment criterion to process the second outlier or the outlier interval; and / or divide the actual battery state data according to the interval division criterion to obtain several intervals; determine the outlier interval from the several intervals; use a preset value as a judgment criterion to process the outlier interval.

[0019] Preferably, the data calculation module includes:

[0020] A segment splitting module, configured to calculate the third time difference between two adjacent actual battery state data. If the third time difference is greater than or equal to the third difference threshold, the two adjacent actual battery state data are respectively used as segment endpoints for splitting; if the third time difference is less than the third difference threshold, the two adjacent actual battery state data are attributed to the same actual charging SOC segment;

[0021] A segment patching module, configured to, for each actual charging SOC segment, determine whether the fourth time difference between two adjacent actual battery state data within the actual charging SOC segment is between the fourth difference threshold and the third difference threshold; if so, perform linear interpolation at a preset interval time between the two adjacent actual battery state data, and calculate the SOC value corresponding to the linear interpolation; wherein, the fourth time difference threshold is less than the third time difference threshold;

[0022] A feature calculation module, configured to remove the head and tail segments of the interpolated several actual charging SOC segments; and calculate the actual charging power corresponding to each of the removed several actual charging SOC segments.

[0023] Preferably, the model evaluation module includes:

[0024] A model construction module, configured to collect the charging capacity sample data generated during charging by all vehicles of the same vehicle model and the same battery pack model as the vehicle in a fixed threshold period, and the corresponding working condition sample data; use the charging capacity sample data and the working condition sample data to perform fitting training on a machine learning algorithm to obtain the capacity correction model.

[0025] Preferably, the model evaluation module further includes:

[0026] A capacity correction module, configured to input the actual charging conditions corresponding to the plurality of actual charging SOC segments and the correction conditions into the capacity correction model for processing, respectively obtaining first predicted charging capacities of the plurality of actual charging SOC segments under their respective actual charging conditions, and second predicted charging capacities under the standard condition; comparing the second predicted charging capacities with the first predicted charging capacities corresponding to the plurality of actual charging SOC segments respectively to obtain condition correction coefficients corresponding to the plurality of actual charging SOC segments respectively; using the condition correction coefficients corresponding to the plurality of actual charging SOC segments respectively to correct the actual charging amounts corresponding to the plurality of actual charging SOC segments respectively to obtain corrected charging amounts corresponding to the plurality of actual charging SOC segments respectively; and further configured to use the weighted ratio of the plurality of actual charging SOC segments and their corrected charging amounts as the actual charging capacity;

[0027] A rated capacity calculation module, configured to collect the initial charging state data of the battery and obtain a plurality of initial charging SOC segments corresponding to the initial charging state data of the battery and their initial charging amounts; and determine the rated charging capacity by using the plurality of initial charging SOC segments and their initial charging amounts;

[0028] A battery health assessment module, configured to assess the battery health SOH by using the ratio of the actual charging capacity to the rated charging capacity.

[0029] Preferably, the rated capacity calculation module is further configured to process the initial charging conditions and the standard condition corresponding to the plurality of initial charging SOC segments by using the capacity correction model, and obtain condition correction coefficients corresponding to the plurality of initial charging SOC segments under the standard condition respectively based on the processing results, and use the condition correction coefficients corresponding to the plurality of initial charging SOC segments respectively to correct the initial charging amounts corresponding to the plurality of initial charging SOC segments respectively; and determine the rated charging capacity based on the plurality of actual charging SOC segments and their corrected charging amounts.

[0030] In a second aspect of the present disclosure, a method for assessing the health of a power battery is disclosed, and the method includes:

[0031] Obtaining the actual state data of the battery;

[0032] When the cumulative charging time period of the power battery reaches a cumulative duration threshold, preprocessing the actual state data of the battery, where the preprocessing includes: data parsing, data conversion, and data cleaning;

[0033] Performing charging segment splitting on the actual state data of the battery to obtain a plurality of actual charging SOC segments; and searching for the charging conditions corresponding to the plurality of actual charging SOC segments from the actual state data of the battery;

[0034] Process the actual charging conditions and standard conditions corresponding to the several actual charging SOC segments by using the trained capacity correction model, and obtain the working condition correction coefficients corresponding to the several actual charging SOC segments under the standard conditions based on the processing results; correct the actual charging powers corresponding to the several actual charging SOC segments by using the working condition correction coefficients corresponding to the several actual charging SOC segments; obtain the actual charging capacity based on the several actual charging SOC segments and their corrected charging powers; calculate the rated charging capacity of the vehicle based on the initial battery state data at the initial stage of the vehicle operation; and evaluate the battery health SOH by using the ratio of the actual charging capacity to the rated charging capacity.

[0035] Preferably, the actual battery state data includes: vehicle identification, reporting time, vehicle state, charge and discharge state, total voltage, total current, SOC, temperature;

[0036] The preprocessing of the actual battery state data specifically includes:

[0037] Parse the actual battery state data from unstructured data into structured data;

[0038] Perform unit conversion and business data conversion on the actual battery state data;

[0039] Repair and / or delete outliers in the actual battery state data.

[0040] Preferably, the repair and / or deletion of outliers in the actual battery state data specifically includes:

[0041] When repairing outliers, when the first time differences between the intermediate time and the front and rear times are both less than the first difference threshold, repair the actual battery state data at the intermediate time to the actual battery state data at the front and rear times;

[0042] When deleting outliers, delete the first outliers in the actual battery state data that exceed the corresponding value range;

[0043] Traverse the actual battery state data in turn by using the method of taking derivatives to obtain the adjacent change amounts between the actual battery state data and the adjacent front and rear data; and detect the adjacent change amounts by using the change amount threshold to determine the second outliers or the outlier intervals corresponding to the second outliers; process the second outliers or the outlier intervals by using a preset value as a judgment criterion; and / or divide the actual battery state data according to the interval division criterion to obtain several intervals; determine the outlier intervals from the several intervals; and process the outlier intervals by using a preset value as a judgment criterion.

[0044] Preferably, the charging segment of the actual battery state data is segmented to obtain a plurality of actual charging SOC segments, including:

[0045] Calculate the third time difference between two adjacent actual battery state data. If the third time difference is greater than or equal to the third difference threshold, the two adjacent actual battery state data are respectively used as segment endpoints for segmentation; if the third time difference is less than the third difference threshold, the two adjacent actual battery state data are attributed to the same actual charging SOC segment;

[0046] For each actual charging SOC segment, determine whether the fourth time difference between two adjacent actual battery state data within the actual charging SOC segment is between the fourth difference threshold and the third difference threshold; if so, perform linear interpolation at a preset interval time between the two adjacent actual battery state data, and calculate the SOC value corresponding to the linear interpolation; wherein, the fourth time difference threshold is less than the third time difference threshold;

[0047] Remove the head and tail segments of the interpolated plurality of actual charging SOC segments; and calculate the actual charging power corresponding to each of the remaining plurality of actual charging SOC segments.

[0048] Preferably, the capacity correction model is trained by the following method:

[0049] Collect the charging capacity sample data generated during charging of all vehicles of the same vehicle model and the same battery pack model as the vehicle in a fixed threshold period, and the corresponding working condition sample data; use the charging capacity sample data and the working condition sample data to perform fitting training on a machine learning algorithm to obtain the capacity correction model.

[0050] Preferably, the trained capacity correction model is used to process the actual charging working conditions and standard working conditions corresponding to the plurality of actual charging SOC segments, and based on the processing results, obtain the working condition correction coefficients corresponding to the plurality of actual charging SOC segments under the standard working conditions; use the working condition correction coefficients corresponding to the plurality of actual charging SOC segments to correct the actual charging power corresponding to the plurality of actual charging SOC segments; based on the plurality of actual charging SOC segments and their corrected charging power, obtain the actual charging capacity, specifically including:

[0051] Input the actual charging conditions corresponding to the several actual charging SOC segments and the correction conditions into the capacity correction model for processing, respectively obtaining the first predicted charging capacity of the several actual charging SOC segments under their respective actual charging conditions and the second predicted charging capacity under the standard condition; compare the second predicted charging capacity with the first predicted charging capacity corresponding to each of the several actual charging SOC segments respectively to obtain the condition correction coefficients corresponding to each of the several actual charging SOC segments; use the condition correction coefficients corresponding to each of the several actual charging SOC segments to correct the actual charging power corresponding to each of the several actual charging SOC segments to obtain the corrected charging power corresponding to each of the several actual charging SOC segments; and is also used to take the weighted ratio of the several actual charging SOC segments and their corrected charging powers as the actual charging capacity.

[0052] Calculating the rated charging capacity of the vehicle based on the battery initial state data in the initial stage of the vehicle operation specifically includes:

[0053] Collect the battery initial charging state data and obtain several initial charging SOC segments corresponding to the battery initial charging state data and their initial charging powers; determine the rated charging capacity by using the several initial charging SOC segments and their initial charging powers.

[0054] Preferably, calculating the rated charging capacity of the vehicle based on the battery initial state data in the initial stage of the vehicle operation specifically includes:

[0055] Use the capacity correction model to process the initial charging conditions and the standard condition corresponding to the several initial charging SOC segments, and obtain the condition correction coefficients corresponding to each of the several initial charging SOC segments under the standard condition based on the processing results, and use the condition correction coefficients corresponding to each of the several initial charging SOC segments to correct the initial charging powers corresponding to each of the several initial charging SOC segments; obtain the rated charging capacity based on the several actual charging SOC segments and their corrected charging powers.

[0056] In the third aspect of the present disclosure, an electronic device is disclosed, including the health assessment system of the power battery as described in any of the foregoing technical solutions.

[0057] Through one or more embodiments of this specification, this specification has the following beneficial effects or advantages:

[0058] In this specification, by segmenting the actual state data of the battery during charging, several actual charging SOC segments are obtained and their corresponding charging working conditions are found. On the one hand, the capacity correction model is used to correct the charging capacity with reference to the charging working condition, calculate the working condition correction coefficient and feedback it to the actual charging power, and combine it with the SOC to obtain the corrected actual charging capacity, so as to eliminate the influence of capacity errors caused by different charging working conditions and improve the accuracy of evaluation. On the other hand, the actual charging capacity of the battery is comprehensively calculated using several charging SOC segments in the historical cumulative charging period, so that the evaluation of the battery charging capacity in this specification does not solely depend on one charging segment, thereby eliminating the error fluctuations of a single segment and making the evaluation result more accurate.

[0059] In this specification, when constructing the model, all vehicles of the same model as this vehicle, the same battery pack model, and within the same cycle threshold range are used to construct the working condition correction model, which solves the accuracy problem caused by inaccurate correction coefficients that may be caused by insufficient data volume of a single vehicle.

[0060] In this specification, the rated charging capacity of the battery is estimated based on the actual charging data, rather than directly using the nominal capacity set for the power battery, which can avoid the SOH estimation error caused by individual differences in the battery due to processes and other factors.

[0061] The above description is only an overview of the technical solution of this specification. In order to be able to more clearly understand the technical means of this specification, it can be implemented in accordance with the content of the specification. And in order to make the above and other purposes, features and advantages of this specification more obvious and understandable, the following specifically illustrates the specific implementation manners of this specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of this specification. And throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0063] Figure 1 Shows a schematic diagram of a health assessment system for a power battery according to an embodiment of this specification;

[0064] Figure 2 Shows an example diagram of removing SOC abnormal points within a reasonable range according to an embodiment of this specification;

[0065] Figure 3 Shows a flowchart of a health assessment method for a power battery according to an embodiment of this specification. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0066] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art.

[0067] Explanation of keywords:

[0068] SOC (State Of Charge), the available state of the remaining charge in the battery, is generally expressed as a percentage.

[0069] SOH, the battery health, characterizes the ability of the current battery to store electrical energy compared to a new battery, that is, the state of the battery from the start to the end of its life, thereby quantitatively describing the performance state of the current battery, and is generally expressed as a percentage.

[0070] Refer to Figure 1 , the embodiments of this specification disclose a health assessment system for power batteries. Based on the data acquisition technology of the vehicle network, the actual battery state data presented during the charging of the power battery installed in this vehicle is collected, combined with the basic data of the vehicle / (or) battery, and the health assessment of the power battery of this vehicle is realized by using big data analysis technology. Since the health assessment system of this specification needs to obtain the relevant data of other vehicles of the same model when assessing the power battery, and the computing power of the in-vehicle equipment of this vehicle is limited. Therefore, the health assessment system of this specification can be implemented in the cloud server, and data is transmitted mutually with all vehicles based on the vehicle network technology. Of course, on the premise that the computing power of the in-vehicle equipment can bear it, the health assessment system can be implemented in the in-vehicle equipment, and the relevant data of other vehicles can be obtained from the cloud server by using the vehicle network technology.

[0071] The health assessment system of this specification includes: a data access module 101, a data preprocessing module 102, a data calculation module 103, and a model evaluation module 104.

[0072] Among them, this vehicle provides the data required for the health assessment of the power battery, including the actual battery state data and the vehicle / battery basic data; the actual battery state data includes data such as vehicle identification, reporting time, vehicle state, charge / discharge state, total voltage, total current, charging current, SOC, temperature, etc.; the vehicle / battery basic data includes data such as vehicle brand, vehicle model, vehicle battery pack model, vehicle battery pack rated capacity, vehicle battery pack rated voltage, etc. Optionally, the vehicle reports the actual battery state data in real time during the charging of the power battery; or reports the actual battery state data at the start or end of each charging; or accumulates the actual battery state data of the power battery until the cumulative charging period of the power battery reaches the cumulative duration threshold and then reports the actual battery state data. The reporting method is to report to the data access module 101 based on the vehicle network. There is no limit on the reporting time of the vehicle / battery basic data, which can be reported when initially connecting to the data access module 101 or when needed. Since the actual battery state data reported in this embodiment is the actual battery state data corresponding to the cumulative charging period of the power battery, therefore, the actual battery state data in this embodiment may be the state data generated by multiple chargings of the power battery, and the working conditions of the power battery during each charging may be different.

[0073] Since the capacity estimated by a single charge can only indicate the health of a single charge, and the error fluctuation of a single charge is relatively large, which will affect the accuracy of the battery health assessment. Therefore, this embodiment uses the actual battery state data corresponding to the cumulative charging period to evaluate the health of the power battery, so that the health assessment of the power battery does not solely depend on a single charging segment, thereby eliminating the error fluctuation of a single segment and making the assessment result more accurate.

[0074] The data access module 101 is used to access and store the actual battery state data presented by the power battery during the cumulative charging period.

[0075] The data preprocessing module 102 is used to preprocess the actual battery state data when the cumulative charging period of the power battery reaches the cumulative duration threshold. The preprocessing includes: data parsing, data conversion, and data cleaning.

[0076] The data calculation module 103 is used to segment the actual battery state data into several actual charging SOC segments; and find the charging working conditions corresponding to the several actual charging SOC segments from the actual battery state data.

[0077] The model evaluation module 104 is configured to process the actual charging conditions and standard conditions corresponding to the several actual charging SOC segments by using the trained capacity correction model, and obtain the condition correction coefficients corresponding to the several actual charging SOC segments under the standard conditions based on the processing results; correct the actual charging power corresponding to the several actual charging SOC segments by using the condition correction coefficients corresponding to the several actual charging SOC segments; obtain the actual charging capacity based on the several actual charging SOC segments and their corrected charging power; calculate the rated charging capacity of the vehicle based on the initial battery state data at the initial stage of the vehicle operation; and evaluate the battery health SOH by using the ratio of the actual charging capacity to the rated charging capacity.

[0078] In this specification, by segmenting the actual battery state data into charging segments, several actual charging SOC segments are obtained and their corresponding charging conditions are found. On the one hand, the capacity correction model is used to correct the charging capacity with reference to the charging conditions, calculate the condition correction coefficient and feedback it to the actual charging power, and combine with the SOC to obtain the corrected actual charging capacity, so as to eliminate the influence of capacity errors caused by different charging conditions and improve the accuracy of the evaluation. On the other hand, the actual charging capacity of the battery is comprehensively calculated by using several charging SOC segments in the cumulative charging period, so that the evaluation of the battery charging capacity in this specification does not solely depend on a single charging segment, thereby eliminating the error fluctuations of a single segment and making the evaluation result more accurate.

[0079] For the convenience of illustration and explanation of this specification, the foregoing modules will be specifically introduced below.

[0080] In the data access module 101, based on the unified data access interface, the uploaded data is archived and stored according to the unified data standard, which is convenient for subsequent data calculation or analysis.

[0081] Optionally, when accessing the actual battery status data, the actual battery status data is accessed and stored in real time; or the actual battery status data is accessed and stored at the start or end of each charging. In addition, considering that in actual situations, it is inaccurate to evaluate the battery health using only the capacity estimated from a single charge, as it is easily affected by fluctuations in the capacity data error. The battery health generally shows a slow downward trend and does not experience a cliff-like drop. Therefore, to improve the evaluation accuracy, this specification accesses and stores the actual battery status data with a cumulative duration threshold for the battery health, and performs subsequent estimation processing, so that the battery charge capacity evaluation in this specification does not solely rely on a single charging segment, thereby eliminating the error fluctuations in a single segment and making the evaluation result more accurate. Specifically, when the cumulative charging period of the power battery does not reach the cumulative duration threshold, the battery health is not estimated and can be evaluated as 100%. When the cumulative charging period reaches the cumulative duration threshold, the actual battery status data is accessed and stored.

[0082] The data preprocessing module 102 specifically includes: a data parsing module 1021, a data conversion module 1022, and a data cleaning module 1023. The execution order of the three is determined according to the actual situation.

[0083] The data parsing module 1021 is used to parse the actual battery status data from unstructured data into structured data. Optionally, it is also used to parse the vehicle / battery basic data from unstructured data into structured data, and merge the structured data corresponding to the actual battery status data and the structured data corresponding to the vehicle / battery basic data, so as to prepare for subsequent evaluation.

[0084] The data conversion module 1022 is used to perform unit conversion and business data conversion on the actual battery status data. Unit conversion is to convert the units in the actual battery status data into a unified unit representation. For example, the unit of current is uniformly converted to ampere (A), the unit of voltage is uniformly converted to volt (V), the unit of temperature is uniformly converted to degree Celsius (°C), the unit of time is uniformly converted to second (s), and the unit of SOC is uniformly converted to percentage. Business data conversion is to convert data with different representation methods into a unified representation method based on the conversion coefficient between data. For example, the SOC displayed on some vehicles may be the dashboard SOC, which does not represent the SOC of the power battery. In this case, it is necessary to convert the dashboard SOC to the SOC of the power battery according to the coefficient between the dashboard SOC and the power battery SOC in the vehicle static data.

[0085] The data cleaning module 1023 is used to repair and / or delete abnormal points in the actual battery status data.

[0086] During the implementation of anomaly point repair, the data cleaning module 1023 is specifically configured to repair the actual battery state data at the intermediate moment to the actual battery state data at the front and rear moments when the first time differences between the intermediate moment and the front and rear moments are both less than the first difference threshold. Anomaly point repair refers to repairing the intermediate actual state data. Specifically, the abnormal actual battery state data is regarded as intermediate data, and its adjacent front and rear data are used to repair it, so as to eliminate the influence of abnormal data. Taking the battery charge and discharge state as an example, if the current charge and discharge state at the current moment is not "parking charging", but the charge and discharge states at the previous moment and the next moment are both "parking charging", and the time intervals between the current moment and the front and rear moments are both less than the first difference threshold, then the charge and discharge state at the current moment can be repaired to "parking charging".

[0087] Anomaly point deletion is mainly the deletion of outliers. The outliers in this embodiment include: the first outliers that exceed the reasonable value range; the second outliers that do not exceed the reasonable value range, but the data difference or change amount between them and the adjacent data exceeds the threshold.

[0088] Therefore, for the first outliers, the data cleaning module 1023 is specifically configured to delete the first outliers that exceed the corresponding value range in the actual battery state data. For example, if the value of SOC at a certain moment is not within the range of [0, 100%], directly delete the SOC at that moment.

[0089] For the second outliers, since they do not exceed the reasonable value range, they need to be further processed, and based on this, it is determined whether to delete the second outliers.

[0090] The data cleaning module 1023 is specifically further configured to sequentially traverse the actual battery state data by using the derivative method to obtain the adjacent change amount between the actual battery state data and the adjacent front and rear data; and use the change amount threshold to detect the adjacent change amount to determine the second outliers or the outlier intervals corresponding to the second outliers; further, use a preset value as a judgment criterion to process the second outliers or the outlier intervals.

[0091] Specifically, for each actual battery state data, two adjacent change amounts between the actual battery state data and the adjacent front and rear data are determined by using the derivative method. For example, when taking the derivative, the ratio of the data difference between the actual battery state data and the previous adjacent data to the time difference is used to obtain the adjacent change amount.

[0092] Detect two adjacent variation amounts using a variation threshold, and determine whether to retain the actual battery state data according to the detection result. Specifically, if the detection result shows that both of the two adjacent variables are less than the variation threshold, it indicates that the change of the actual battery state data from the adjacent data before and after is normal, and the actual battery state data is retained. If the detection result shows that both of the two adjacent variation amounts are greater than or equal to the variation threshold, it indicates that the actual battery state data is the second outlier and is an isolated point, and the actual battery state data is deleted.

[0093] If one of the two adjacent variation amounts is less than the variation threshold, it indicates that both the actual battery state data and one of the adjacent data are the second outliers. Then continue to traverse the subsequent actual battery state data in a derivative manner, and determine the outlier interval accordingly. The outlier interval is an interval composed of several adjacent second outliers. If the total number of second outliers in the outlier interval is less than a preset value, it indicates that these second outliers are all abnormal points and are deleted. If the total number of second outliers in the outlier interval is greater than the preset value, it indicates that these second outliers are normal data points formed during charging. It may be due to an increase in the charging current that causes jump points, but charging can still be normal. Therefore, these outlier data points are retained as normal data.

[0094] The above is one way to process the second outliers. Of course, there are other ways in this specification to confirm or process the second outliers. Either one or both can be used.

[0095] As an optional embodiment, the data cleaning module 1023 is further configured to divide the actual battery state data into several intervals according to an interval division standard, and determine the outlier interval from the several intervals; and process the outlier interval using a preset value as a judgment criterion.

[0096] Specifically, there are two division methods in this embodiment when dividing intervals. The first division method: First, find all the second outliers by using the judgment result between the data difference of adjacent data before and after and the difference threshold. Use the second outliers as the interval division standard for interval division, so that several intervals can be divided. Further, if several second outliers are coherent, they can be divided into the same outlier interval.

[0097] Specifically, when dividing the interval, it is necessary to consider whether the time difference between the second outlier and the adjacent data before and after is less than the second time difference threshold. If it is less, the two data points corresponding to this time difference are used as interval division points for interval division. Taking SOC as an example, since it gradually increases over time, the data of SOC and time change relatively. Therefore, while considering the data difference, it is necessary to use the second time difference threshold to limit the time difference between adjacent data before and after, so as to ensure that no other data is inserted between adjacent data before and after.

[0098] It can be seen that when using the second outlier as the interval division standard in this embodiment, the judgment result between the time difference between the moment when the second outlier is located and the moment when its adjacent data is located and the second difference threshold is used to divide the several intervals. And since different intervals may correspond to different charging conditions, this operation can avoid mixing the actual state data of batteries charged in different batches into the same interval, thereby avoiding double confusion of data and conditions.

[0099] The above is the implementation method of first finding the second outlier and then dividing the interval. In the second division method, the division can be performed without searching for the second outlier. Specifically, the difference threshold corresponding to the data difference and the second difference threshold corresponding to the time difference are used as the interval division standard for interval division. Specifically, it is sequentially judged whether the data difference between two adjacent data is greater than the difference threshold, and whether the time difference between two adjacent data is less than the second difference threshold; if the judgment results of both are yes, the two adjacent data are respectively used as interval endpoints for interval division to obtain several intervals. Further, the second outlier and its outlier interval are determined therefrom using the judgment criteria of the second outlier.

[0100] Further, when processing the outlier interval using the preset value as the judgment criterion, if the total number of second outliers in the outlier interval is less than the preset value, the interval is deleted, otherwise it is retained.

[0101] For example, for the second outlier that does not exceed the reasonable value range, taking SOC as an example. The data difference between the second outlier SOC and its adjacent previous SOC is greater than the difference threshold SOCDiffThre, but the time difference between the two is less than the second difference threshold durationThre. Then, an interval division is performed between the second outlier SOC and its adjacent previous SOC. The processing method for the second outlier SOC and its adjacent subsequent SOC is similar, and finally, an outlier interval is obtained. Among them, if the second outlier is an isolated point, it is directly deleted. If an outlier interval is formed, it is determined whether to delete it according to whether the number of points inside the outlier interval is less than the preset value pointThre. For example, if it is less than the preset value pointThre, the outlier interval is deleted; otherwise, it is retained. Optionally, SOCDiffThre is taken as 10%, durationThre is taken as 50s, and pointThre is taken as 10, but this does not form a limitation.

[0102] As Figure 2 shown, it shows an example diagram of removing SOC outliers within a reasonable range. It can be seen that according to the above logic, the actual battery state data will be divided into 3 interval segments. Among them, the number of second outliers in interval 2 is less than 10 points, and interval 2 is deleted.

[0103] The data calculation module 103 includes: a segment splitting module 1031, a segment patching module 1032, and a feature calculation module 1033.

[0104] The segment splitting module 1031 is used to calculate the third time difference between two adjacent actual battery state data. If the third time difference is greater than or equal to the third difference threshold timeDiffThr, the two adjacent actual battery state data are respectively used as segment endpoints for splitting; if the third time difference is less than the third difference threshold timeDiffThr, the two adjacent actual battery state data are attributed to the same actual charging SOC segment. Optionally, timeDiffThr is 5 minutes, but this does not form a limitation.

[0105] The segment splitting is to obtain the fragmentation processing of the actual battery state data, so as to obtain the actual charging SOC segments generated by charging under different working conditions. By using the third difference threshold corresponding to the time difference as the division criterion for segment division, it is possible to avoid double confusion of data and working conditions. In the actual charging SOC segment field, there are corresponding working conditions such as the total voltage, total current, and temperature.

[0106] Specifically, first, filter the data with the battery charge and discharge status of "parking charging" in the actual battery status data. For the filtered actual battery status data, calculate the time difference between adjacent data after sorting by time, and segment the actual charging SOC based on the time difference. Further, based on the result of the segment division, label each actual charging SOC segment with a unique charging segment identifier. For example, label the charging segment based on the vehicle unique identifier + charging start time. It should be noted that if the SOC in the actual battery status data is used for segment division, it can be directly divided into several actual charging SOC segments according to time. If the charging current in the actual battery status data is used for segment division, several current segments will be obtained, and then based on the mapping relationship between the charging current and the SOC, they will be converted into several actual charging SOC segments represented by the SOC.

[0107] The segment repair module 1032 is used to determine whether the fourth time difference between two adjacent actual battery status data within each actual charging SOC segment is between the fourth difference threshold and the third difference threshold; if so, perform linear interpolation at a preset interval time within the two adjacent actual battery status data, and calculate the SOC value corresponding to the linear interpolation; wherein, the fourth time difference threshold is less than the third time difference threshold.

[0108] For each actual charging SOC segment, there may be some missing data. For example, due to the above-mentioned outlier deletion operation, it may lead to missing data in the charging SOC segment. Therefore, for the accuracy of the battery health assessment, it is necessary to repair the data of each actual charging SOC segment and supplement the missing values of each actual charging SOC segment.

[0109] When using the linear interpolation method for data repair, the specific steps are as follows:

[0110] (1) Insert data points; if in the actual charging SOC segment, the time of the current actual battery status data is T nOw , and the time of the next actual battery status data is T next , if the fourth difference threshold insertDurationThre < T next – T now < the third time difference threshold timeDiffThr, then insert a piece of data every other acquisition time interval t. Except for the timestamp, the values of the remaining fields of this inserted data are temporarily all empty. If T next – T now < t, stop inserting.

[0111] (2) Supplement the missing values of the inserted data; assume that the field to be repaired is the field SOC. First, find the time when the adjacent SOC values are not empty, assume it is Tbegin With T end , the corresponding SOC values are SOC begin and SOC end , then for the intermediate T m The patched value SOC of the SOC at the moment m is obtained using the following formula:

[0112] Furthermore, the patched value and the patching moment are filled into the corresponding inserted data.

[0113] As an alternative embodiment, since there may be various abnormalities for the segmented charging segments. Therefore, the data calculation module 103 of this embodiment further includes: a quality inspection module for performing quality inspection on a number of actual charging SOC segments using quality inspection indicators. Furthermore, it is also used to identify the detected abnormal segments for subsequent modules to selectively filter. Optionally, the quality inspection indicators include but are not limited to the following indicators: the charging duration of the segment, the charged SOC of the segment, the SOC charged per unit duration of the segment, the SOC missing rate of the segment, the charging current missing rate of the segment.

[0114] The feature calculation module 1033 is used to remove the head and tail segments of the interpolated several actual charging SOC segments; and calculate the actual charging power corresponding to each of the removed several actual charging SOC segments.

[0115] Specifically, since the head and tail of the actual charging SOC segment have a greater impact on the capacity calculation. For example, since the SOC of the power battery is displayed according to a percentage scale. For example, when the power battery is charging, the SOC will be displayed as 5%, 10%, 15%... and so on as the power increases. And the power battery has a remaining SOC before charging. Since the power battery is displayed based on a 5-percentage-point measurement standard, when the power battery is displayed as having a remaining SOC of 5%, its actual remaining may be any value between 0% and 10%. For example, it may be 4% or it may be 9%. Then, if 5% is used as the head in the actual charging SOC segment sequence for calculation, a capacity error will occur. The same is true for the tail of the actual charging SOC segment sequence. Therefore, in order to eliminate the capacity error caused by the head and tail, it is necessary to remove the head and tail from the actual charging SOC segment sequence. Another example is that two actual charging SOC segment sequences both change from SOC1 to SOC n , and the first sequence is [SOC1, SOC2, SOC2,..., SOC m , SOC m , SOC n, the second sequence is [SOC1, SOC1, SOC1, SOC1, SOC1, SOC2, SOC2, …, SOC m , SOC m , SOC n , SOC n ; taking actual values as an example, the first sequence is [5%, 10%, 10%, …, 80%, 80%, 85%], and the second sequence is [5%, 5%, 5%, 5%, 5%, 10%, 10%, …, 80%, 80%, 85%, 85%]. It can be seen that both sequences participate in the calculation with 5% as the head, and only the head and tail differences result in a capacity error. Therefore, it is necessary to reconfirm the start and end SOC values to eliminate the influence of the head and tail in the actual charging SOC segment.

[0116] Taking the first sequence as an example to receive the steps of removing the head and tail: The first sequence changes with time in the time series as [SOC1, SOC2, SOC2, …, SOC m , SOC m , SOC n , then the charging segment after removing the head and tail starts from the first SOC2 and ends at the last SOC m .

[0117] Furthermore, when calculating the actual charging power corresponding to each of the several actual charging SOC segments after removal, the charging current in the screening interval is used for ampere-hour integration calculation to obtain the actual charging power. The calculation formula is as follows: Q Ah =∫I t ·dt≈∑i t ·Δt, where Δt is the sampling time interval in the actual charging SOC segment, and I t is the discrete current sampling value in the actual charging SOC segment.

[0118] The model evaluation module 104 includes: a model construction module 1041, a capacity correction module 1042, a rated capacity calculation module 1043, and a battery health assessment module 1044.

[0119] The model construction module 1041 is used to collect the charging capacity sample data generated during charging by all vehicles of the same vehicle model and the same battery pack model as the vehicle in a fixed threshold period, as well as the corresponding working condition sample data; use the charging capacity sample data and the working condition sample data to perform fitting training on the machine learning algorithm to obtain the capacity correction model.

[0120] The state of health assessment of power batteries generally uses the ratio of the actual charging capacity to the rated charging capacity for evaluation. On the one hand, since the actual charging capacity in this specification needs to be calculated using the capacity values of the charging SOC segments during charging in different batches, and the charging SOC segments during charging in different batches will be affected by different charging conditions and have large differences when calculating the capacity values. Therefore, in order to eliminate the errors caused by different charging conditions, it is necessary to perform capacity correction for different conditions. On the other hand, due to the differences in manufacturing processes, even for the same vehicle model and the same battery pack model, the actual rated capacities of the batteries are different, which results in a certain difference between the actual rated capacity presented by the battery and the calibrated rated capacity publicly announced by the vehicle factory. Therefore, for the accuracy of the battery state of health assessment, the estimation of the rated capacity is also a necessary step in evaluating the battery state of health.

[0121] Different charging conditions have an important impact on the charge amount calculated by the ampere-hour integration method. Specifically, even if the same SOC is charged, if the charging conditions are different, the charge amount calculated based on the ampere-hour integration method is also different. Therefore, in order to estimate the actual charging capacity of the power battery, it is necessary to correct the actual charging power of each actual charging SOC segment using the standard conditions, and then estimate the actual charging capacity of the power battery accordingly. The standard conditions refer to the charging conditions most commonly encountered by the vehicle, which may be different for each vehicle. And the charging conditions in this embodiment include but are not limited to the charging current, the temperature of the charging battery pack, and the ambient temperature.

[0122] In addition, since the charging data of a single vehicle is very few and the individual differences are large. Therefore, in order to eliminate the individual differences and increase the robustness of the sample data, this specification uses the working condition sample data of each charging segment generated during the charging of all vehicles of the same vehicle model and the same battery pack model as this vehicle within a fixed threshold period (such as a duration period and / or a mileage period) to fit the charging capacity sample data and construct a capacity correction model.

[0123] The reason for selecting the charging capacity sample data generated during charging within a fixed threshold period is that during the entire charging life of the power battery, the change in its charging capacity will be affected by both doping capacity attenuation and actual working conditions, thus affecting the accuracy of working condition correction. Therefore, in order to avoid the influence of doping capacity attenuation, all vehicles select the charging capacity sample data generated during charging within the same fixed threshold period for model construction.

[0124] When building the model, the input of the model is the operating conditions of each charging SOC segment, such as charging current, charging battery pack temperature, ambient temperature, etc. Optionally, the median current within the segment is used as the representative value of the charging current operating condition, and the median temperature within the segment is used as the representative value of the charging temperature operating condition. Of course, in addition to the median, the average value, quantile value, weighted average value, geometric average value, etc. can also be used as the representative value of the operating condition. The output of the model is the battery capacity calculated based on the ampere-hour integration method. The initial model includes but is not limited to logistic regression, decision tree, and random forest. If the effect of the constructed model meets the requirements, it is output; otherwise, the parameters or the model are adjusted again and training is carried out again. Of course, for the accuracy of model construction, the fixed threshold period can also be adjusted.

[0125] The capacity correction module 1042 is configured to input the actual charging operating conditions corresponding to the plurality of actual charging SOC segments and the corrected operating conditions into the capacity correction model for processing, respectively obtaining the first predicted charging capacity of the plurality of actual charging SOC segments under their respective actual charging operating conditions, and the second predicted charging capacity under the standard operating condition; comparing the second predicted charging capacity with the first predicted charging capacity corresponding to each of the plurality of actual charging SOC segments respectively to obtain the operating condition correction coefficient corresponding to each of the plurality of actual charging SOC segments; using the operating condition correction coefficient corresponding to each of the plurality of actual charging SOC segments to correct the actual charging power corresponding to each of the plurality of actual charging SOC segments to obtain the corrected charging power corresponding to each of the plurality of actual charging SOC segments; and further configured to use the weighted ratio of the plurality of actual charging SOC segments and their corrected charging power as the actual charging capacity. In this embodiment, by using the standard operating condition to correct the capacity under different operating conditions, the error influence caused by different operating conditions is eliminated.

[0126] Optionally, before processing using the capacity correction model, it can be first determined whether the actual charging operating condition corresponding to the plurality of actual charging SOC segments is the standard operating condition. Among them, the standard operating condition is the operating condition with the largest number among the actual charging operating conditions corresponding to the plurality of actual charging SOC segments. If it is the standard operating condition, there is no need to use the correction model, and the corresponding charging power is directly calculated using the ampere-hour integration method, and the ratio of the charging power to the SOC segment is used as the corresponding charging segment capacity. If it is not the standard operating condition, the capacity correction model needs to be called for correction.

[0127] For the convenience of explaining and interpreting this specification, specific examples are used for illustration below. The charging SOC segments of this vehicle have n, and the charging start time t is used as the identifier. The corresponding operating conditions and actual charging power are shown in Table 1.

[0128] Table 1

[0129]

[0130] If the standard working condition is a, then during actual use, the actual charging SOC segments corresponding to the time period from t1 to t3 do not need to be corrected, and the actual charging amounts during the time period from t4 to tn all need to be corrected.

[0131] Taking the time t4 as an example, the steps are as follows:

[0132] (1) Based on the trained capacity correction model, input the working condition c at time t4 to obtain the first predicted charging capacity Cm of time t4 under the working condition c.

[0133] (2) Based on the trained capacity correction model, input the standard working condition a to obtain the second predicted charging capacity Cn of t4 under the standard working condition a.

[0134] (3) Calculate the working condition correction coefficient Cn / Cm at time t4, and based on this coefficient, correct the actual charging amount of time t4 under the standard working condition a to obtain the corrected charging amount of time t4 under the standard working condition a: Q4_correct = Q4 * Cn / Cm.

[0135] Correct the actual charging amounts of other times according to the above method. Thus, each time corresponds to its respective corrected charging amount. Although the charging amounts at times t1 to t3 are not corrected, during calculation, the actual charging amounts at times t1 to t3 can be regarded as the corrected charging amounts and included in the calculation. Thus, take the weighted ratio of the said several actual charging SOC segments and their corrected charging amounts as the actual charging capacity. Specifically, the following formula is used to calculate the actual charging capacity: where Q′ i is the corrected charging amount of the i-th one, and SoC i is the charged SOC of the i-th actual charging SOC segment.

[0136] Of course, the calculation method of the actual charging capacity is not limited to this. For example, the capacity of each charging SOC segment can be calculated first, and then the average value method is used to obtain the actual charging capacity. Specifically, use the formula to calculate the capacity of each charging SOC segment; where C i refers to the actual charging capacity of the i-th actual charging SOC segment, here Q′ i refers to the corrected charging amount of the i-th actual charging SOC segment, [SoC iend -SoC ibegin refers to the charged SOC of the i-th actual charging SOC segment.

[0137] The rated capacity calculation module 1043 is configured to collect the initial charging state data of the battery and obtain a plurality of initial charging SOC segments and their initial charging power corresponding to the initial charging state data of the battery; and determine the rated charging capacity by using the plurality of initial charging SOC segments and their initial charging power.

[0138] Since the data capacity of the power battery hardly decays in the initial stage of charging, the rated charging capacity can be calculated by using the initial charging state data of the battery. The initial stage of battery charging is defined based on capacity decay. For example, if the power battery does not decay within 400 kilometers of vehicle driving, the charging state data within 400 kilometers can be collected to calculate the rated charging capacity. Since the rated capacity of the battery is estimated based on the initial data of actual charging in this embodiment instead of directly using the nominal capacity set at the time of battery factory shipment, the SOH estimation error caused by individual differences in the battery due to processes and the like can be avoided.

[0139] Specifically, in order to reduce the data fluctuation error, the initial charging state data of the vehicle can be collected when the data accumulation duration reaches the standard or the number of cyclic charging times reaches the standard, and the initial charging state data is screened based on the quality inspection index, and a plurality of initial charging SOC segments that meet the following quality inspection index are screened out therefrom. For example, the charged SOC meets the threshold (such as greater than 20%), the SOC missing rate is below 10%, and the charging current missing rate is below 10%, but this is not restrictive. Outliers are removed from the plurality of initial charging SOC segments that are screened out, and the methods that can be adopted include, but are not limited to, the box plot method. If the number of the remaining plurality of initial charging SOC segments is greater than the threshold after removing the outliers, the rated charging capacity is calculated based on the remaining plurality of initial charging SOC segments by using the weighted average method or the method of obtaining the capacity average value. The specific method can refer to the foregoing embodiments and will not be elaborated herein. If the number of the remaining plurality of initial charging SOC segments is less than the threshold after removing the segments with abnormal capacity, the initial charging state data is continuously accumulated and processed again according to the above process until the rated charging capacity is determined.

[0140] Certainly, in order to eliminate the influence of the charging condition on the capacity in the initial stage of battery charging, the rated capacity calculation module 1043 is further configured to process the initial charging condition and the standard condition corresponding to the plurality of initial charging SOC segments by using the trained capacity correction model, and obtain the working condition correction coefficient corresponding to each of the plurality of initial charging SOC segments under the standard condition based on the processing result, and correct the initial charging power corresponding to each of the plurality of initial charging SOC segments by using the working condition correction coefficient corresponding to each of the plurality of initial charging SOC segments; and obtain the rated charging capacity based on the plurality of actual charging SOC segments and their corrected charging power. The specific correction process can refer to the description of the foregoing embodiment for calculating the actual charging capacity and will not be elaborated herein.

[0141] The battery health assessment module 1044 is used to evaluate the state of health (SOH) of the battery by using the ratio of the actual charging capacity to the rated charging capacity.

[0142] In this embodiment, a complete set of battery health assessment systems is designed, covering multiple aspects such as data acquisition, data access, data preprocessing, data calculation, and model evaluation. In addition, during the evaluation process of this embodiment, on the one hand, the actual charging capacity of the battery is comprehensively calculated using several charging SOC segments during the cumulative charging period, so that the battery charging capacity evaluation in this specification does not solely rely on a single charging segment, thereby eliminating the error fluctuations caused by a single segment and making the evaluation results more accurate. On the other hand, considering the influence of different working conditions on the battery capacity calculation, a capacity correction model is designed to correct the charging capacity with reference to the charging working conditions, thereby eliminating the error influence caused by different working conditions and improving the accuracy of the evaluation.

[0143] In addition, when constructing the model, all vehicles of the same vehicle model, the same battery pack model, and within the same cycle threshold range as this vehicle are used to construct the working condition correction model, which solves the accuracy problem caused by inaccurate correction coefficients that may be caused by insufficient data volume of a single vehicle.

[0144] In addition, this embodiment estimates the rated charging capacity of the battery based on the actual charging data, rather than directly using the nominal capacity set by the power battery, which can avoid the SOH estimation error caused by individual differences in the battery due to processes and other factors.

[0145] Based on the same inventive concept as the foregoing embodiments, the following embodiments introduce an electronic device, which can be a cloud server or an in-vehicle device. The electronic device includes the health assessment system of the power battery described in any of the foregoing embodiments.

[0146] Based on the same inventive concept as the foregoing embodiments, the following embodiments introduce a method for assessing the health of a power battery. Refer to Figure 3 The method includes the following steps:

[0147] Step 301, obtain the actual state data of the battery.

[0148] Specifically, the actual state data of the battery and the vehicle / battery basic data; the actual state data of the battery includes data such as vehicle identification, reporting time, vehicle state, charge and discharge state, total voltage, total current, charging current, SOC, temperature, etc.; the vehicle / battery basic data includes data such as vehicle brand, vehicle model, vehicle battery pack model, vehicle battery pack rated capacity, vehicle battery pack rated voltage, etc.

[0149] When accessing the actual battery status data, the actual battery status data is accessed and stored in real time; or the actual battery status data is accessed and stored at the start or end of each charging. In addition, considering that in actual situations, it is inaccurate to evaluate the battery health using only the capacity estimated from a single charge, as it is easily affected by fluctuations in the capacity data error. The battery health generally shows a slow downward trend and does not experience a cliff-like drop. Therefore, to improve the evaluation accuracy, this specification accesses and stores the actual battery status data with a cumulative duration threshold for battery health, and performs subsequent estimation processing, so that the battery charge capacity evaluation in this specification does not solely rely on a single charging segment, thereby eliminating the error fluctuations in a single segment and making the evaluation result more accurate. Specifically, when the cumulative charging period of the power battery does not reach the cumulative duration threshold, the battery health is not estimated, and the battery health can be evaluated as 100%. When the cumulative charging period reaches the cumulative duration threshold, the actual battery status data is accessed and stored.

[0150] Step 302, when the cumulative charging period of the power battery reaches the cumulative duration threshold, preprocess the actual battery status data, and the preprocessing includes: data parsing, data conversion, and data cleaning.

[0151] During data parsing, the actual battery status data is parsed from unstructured data into structured data. Optionally, the vehicle / battery basic data is parsed from unstructured data into structured data, and the structured data corresponding to the actual battery status data and the structured data corresponding to the vehicle / battery basic data are merged to prepare for subsequent evaluation.

[0152] During data conversion, unit conversion and business data conversion are performed on the actual battery status data. Among them, unit conversion is to convert the units in the actual battery status data into a unified unit representation. Among them, business data conversion is to convert data with different representation methods into a unified representation method based on the conversion coefficient between the data.

[0153] During data cleaning, anomaly point repair and / or anomaly point deletion are performed on the actual battery status data.

[0154] During anomaly point repair, when the first time differences between the middle moment and the front and back moments are both less than the first difference threshold, the actual battery status data at the middle moment is repaired to the actual battery status data at the front and back moments;

[0155] During anomaly point deletion, mainly the outliers are deleted. The outliers in this embodiment include: the first outliers that exceed the reasonable value range; the second outliers that do not exceed the reasonable value range but the data difference or change amount from the adjacent data exceeds the threshold.

[0156] Therefore, for the first outlier, delete the first outlier in the actual battery state data that exceeds the corresponding value range.

[0157] For the second outlier, traverse the actual battery state data sequentially by taking the derivative to obtain the adjacent change amount between the actual battery state data and the adjacent data before and after; and use the change amount threshold to detect the adjacent change amount to determine the second outlier or the outlier interval corresponding to the second outlier; use a preset value as a judgment criterion to process the second outlier or the outlier interval; and / or

[0158] Divide the actual battery state data according to the interval division criterion to obtain several intervals; determine the outlier interval from the several intervals; use a preset value as a judgment criterion to process the outlier interval.

[0159] The specific implementation details have been introduced in detail in the foregoing system embodiment, so they will not be elaborated here.

[0160] Step 303: Segment the actual battery state data to obtain several actual charging SOC segments; search for the charging conditions corresponding to the several actual charging SOC segments from the actual battery state data.

[0161] Among them, segmenting is to perform fragmentation processing on the actual battery state data to obtain the actual charging SOC segments generated by charging under different conditions. By using the third difference threshold corresponding to the time difference as the division criterion for segmentation, it is possible to avoid double confusion of data and conditions. In the actual charging SOC segment field, there are corresponding working conditions such as the total voltage, total current, and temperature.

[0162] When segmenting, the third time difference threshold is used as the standard for segmentation. Specifically, calculate the third time difference between two adjacent actual battery state data. If the third time difference is greater than or equal to the third difference threshold, the two adjacent actual battery state data are respectively used as segment endpoints for segmentation; if the third time difference is less than the third difference threshold, the two adjacent actual battery state data are attributed to the same actual charging SOC segment.

[0163] For each actual charging SOC segment, there may be some data missing. For example, due to the above-mentioned operation of deleting outliers, it may cause data missing in the charging SOC segment. Therefore, for the accuracy of the battery health assessment, it is necessary to repair the data of each actual charging SOC segment and supplement the missing values of each actual charging SOC segment.

[0164] In the specific repair process, for each actual charging SOC segment, it is determined whether the fourth time difference between the actual state data of two adjacent batteries within the actual charging SOC segment is between the fourth difference threshold and the third difference threshold; if so, linear interpolation is performed at a preset interval time within the actual state data of the two adjacent batteries, and the SOC value corresponding to the linear interpolation is calculated; wherein, the fourth time difference threshold is less than the third time difference threshold.

[0165] Since the head and tail of the actual charging SOC segment have a greater impact on the capacity calculation, therefore, the head and tail segments of the interpolated actual charging SOC segments are removed, thereby eliminating the influence of the head and tail of the actual charging SOC segment. Thereafter, the actual charging power corresponding to each of the removed actual charging SOC segments is calculated.

[0166] Step 304, using the trained capacity correction model to process the actual charging conditions and standard conditions corresponding to the actual charging SOC segments, and obtaining the condition correction coefficients corresponding to the actual charging SOC segments under the standard conditions based on the processing results; using the condition correction coefficients corresponding to the actual charging SOC segments to correct the actual charging power corresponding to the actual charging SOC segments; obtaining the actual charging capacity based on the actual charging SOC segments and their corrected charging power; calculating the rated charging capacity of the vehicle based on the initial battery state data in the initial stage of the vehicle operation; and evaluating the battery health SOH using the ratio of the actual charging capacity to the rated charging capacity.

[0167] In the process of calculating the actual charging capacity, the actual charging conditions and the corrected conditions corresponding to the actual charging SOC segments are respectively input into the capacity correction model for processing, and the first predicted charging capacity of the actual charging SOC segments under their respective actual charging conditions and the second predicted charging capacity under the standard conditions are respectively obtained; the second predicted charging capacity is compared with the first predicted charging capacity corresponding to each of the actual charging SOC segments to obtain the condition correction coefficients corresponding to the actual charging SOC segments; using the condition correction coefficients corresponding to the actual charging SOC segments to correct the actual charging power corresponding to the actual charging SOC segments to obtain the corrected charging power corresponding to the actual charging SOC segments; and also used to take the weighted ratio of the actual charging SOC segments and their corrected charging power as the actual charging capacity.

[0168] During the process of calculating the rated charging capacity, the initial charging state data of the battery is collected, and a number of initial charging SOC segments corresponding to the initial charging state data of the battery and their initial charging powers are obtained; the rated charging capacity is determined by using the number of initial charging SOC segments and their initial charging powers.

[0169] As an alternative embodiment, in order to eliminate the influence of different working conditions, the initial charging working conditions and standard working conditions corresponding to the number of initial charging SOC segments are processed by using the capacity correction model, and based on the processing results, the working condition correction coefficients corresponding to the number of initial charging SOC segments under the standard working conditions are obtained, and the initial charging powers corresponding to the number of initial charging SOC segments are corrected by using the working condition correction coefficients corresponding to the number of initial charging SOC segments; the rated charging capacity is obtained based on the number of actual charging SOC segments and their corrected charging powers.

[0170] The state of health SOH of the battery is evaluated by the ratio of the actual charging capacity and the rated charging capacity determined in the foregoing manner. Among them, the larger the ratio, the higher the health degree.

[0171] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. Based on the above description, the structure required to construct such a system is obvious. In addition, this specification is not directed to any particular programming language. It should be understood that the content of this specification described herein can be implemented using various programming languages, and the description of a particular language above is for the purpose of disclosing the best mode of this specification.

[0172] In the specification provided herein, a large number of specific details are set forth. However, it can be understood that the embodiments of this specification can be practiced without these specific details. In some instances, well-known methods, structures, and technologies have not been shown in detail so as not to obscure the understanding of this specification.

[0173] Similarly, it should be understood that in order to streamline this disclosure and assist in understanding one or more of the various inventive aspects, in the foregoing description of the exemplary embodiments of this specification, the various features of this specification are sometimes grouped together into a single embodiment, figure, or description thereof. However, the disclosed method should not be construed as reflecting the intention that the claimed specification requires more features than are expressly recited in each claim. Rather, as reflected in the following claims, the inventive aspects lie in less than all the features of the preceding disclosed single embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of this specification.

[0174] Those skilled in the art can understand that the modules in the devices in the embodiments can be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments can be combined into one module or unit or component, and in addition, they can be divided into multiple sub-modules or sub-units or sub-components. Except that at least some of such features and / or processes or units are mutually exclusive, any combination can be used to combine all the features disclosed in this specification (including the accompanying claims, abstract and drawings) and all the processes or units of any method or device so disclosed. Unless otherwise clearly stated, each feature disclosed in this specification (including the accompanying claims, abstract and drawings) can be replaced by an alternative feature that provides the same, equivalent or similar purpose.

[0175] In addition, those skilled in the art can understand that although some of the embodiments herein include certain features included in other embodiments rather than other features, the combination of the features of different embodiments means that it is within the scope of this specification and forms different embodiments. For example, in the following claims, any one of the claimed embodiments can be used in any combination.

[0176] Each component embodiment of this specification can be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art should understand that a microprocessor or a digital signal processor (DSP) can be used in practice to implement some or all of the functions of some or all of the components of the gateway, proxy server, and method according to the embodiments of this specification. This specification can also be implemented as a device or device program (for example, a computer program and a computer program product) for executing part or all of the methods described herein. Such a program for implementing this specification can be stored on a computer-readable medium, or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0177] It should be noted that the above embodiments illustrate the present specification rather than limit the present specification, and those skilled in the art can design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in the claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present specification may be implemented by means of hardware including several different elements and by means of a suitably programmed computer. In the unit claims listing several devices, several of these devices may be embodied by the same item of hardware. The use of the words first, second, and third, etc. does not denote any order. These words may be interpreted as names.

Claims

1. A health assessment system for a power battery, the system comprising: A data access module for accessing and storing the actual battery state data; A data preprocessing module for preprocessing the actual battery state data when the cumulative charging period of the power battery reaches a cumulative duration threshold, the preprocessing including: data parsing, data conversion, and data cleaning; A data calculation module for segmenting the actual battery state data into several actual charging SOC segments; and finding the charging working conditions corresponding to the several actual charging SOC segments from the actual battery state data; A model evaluation module for using a trained capacity correction model to process the actual charging working conditions and standard working conditions corresponding to the several actual charging SOC segments, and obtaining the working condition correction coefficients corresponding to the several actual charging SOC segments under the standard working conditions based on the processing results; using the working condition correction coefficients corresponding to the several actual charging SOC segments to correct the actual charging power corresponding to the several actual charging SOC segments; obtaining the actual charging capacity based on the several actual charging SOC segments and their corrected charging power; calculating the rated charging capacity of the vehicle battery based on the initial battery charging state data at the initial stage of the vehicle operation; and evaluating the battery health SOH using the ratio of the actual charging capacity to the rated charging capacity; A capacity correction module for respectively inputting the actual charging working conditions and standard working conditions corresponding to the several actual charging SOC segments into the capacity correction model for processing, and respectively obtaining the first predicted charging capacity of the several actual charging SOC segments under their respective actual charging working conditions and the second predicted charging capacity under the standard working conditions; comparing the second predicted charging capacity with the first predicted charging capacity corresponding to each of the several actual charging SOC segments to obtain the working condition correction coefficients corresponding to the several actual charging SOC segments; using the working condition correction coefficients corresponding to the several actual charging SOC segments to correct the actual charging power corresponding to the several actual charging SOC segments to obtain the corrected charging power corresponding to the several actual charging SOC segments; and further using the weighted ratio of the several actual charging SOC segments and their corrected charging power as the actual charging capacity; A rated capacity calculation module for collecting the initial battery charging state data and obtaining several initial charging SOC segments and their initial charging power corresponding to the initial battery charging state data; and determining the rated charging capacity using the several initial charging SOC segments and their initial charging power; A battery health assessment module for evaluating the battery health SOH using the ratio of the actual charging capacity to the rated charging capacity.

2. The health assessment system according to claim 1, wherein the actual battery state data includes: Vehicle identification, reporting time, vehicle status, charge and discharge status, total voltage, total current, SOC, temperature; The data preprocessing module specifically includes: A data parsing module for parsing the actual battery state data from unstructured data into structured data; A data conversion module for performing unit conversion and service data conversion on the actual battery state data; A data cleaning module for repairing and / or deleting outliers in the actual battery state data.

3. The health assessment system according to claim 2, when repairing outliers, the data cleaning module is specifically configured to repair the actual battery state data at the intermediate moment to the actual battery state data at the front and back moments when the first time differences between the intermediate moment and the front and back moments are both less than the first difference threshold; When repairing outliers, the data cleaning module is specifically configured to: Delete the first outliers in the actual battery state data that exceed the corresponding value range; Traverse the actual battery state data in turn by using the derivative method to obtain the adjacent change amount between the actual battery state data and the adjacent front and back data; and detect the adjacent change amount by using the change amount threshold to determine the second outliers or the outlier intervals corresponding to the second outliers; Process the second outliers or the outlier intervals by using a preset value as a judgment criterion; and / or divide the actual battery state data according to the interval division criterion to obtain several intervals; determine the outlier intervals from the several intervals; and process the outlier intervals by using a preset value as a judgment criterion.

4. The health assessment system according to claim 1 or 3, the data calculation module includes: A segment splitting module for calculating the third time difference between two adjacent actual battery state data. If the third time difference is greater than or equal to the third difference threshold, the two adjacent actual battery state data are respectively used as segment endpoints for splitting; if the third time difference is less than the third difference threshold, the two adjacent actual battery state data are attributed to the same actual charging SOC segment; A segment patching module for, for each actual charging SOC segment, judging whether the fourth time difference between two adjacent actual battery state data in the actual charging SOC segment is between the fourth difference threshold and the third difference threshold; if so, performing linear interpolation at a preset interval time between the two adjacent actual battery state data, and calculating the SOC value corresponding to the linear interpolation; wherein, the fourth difference threshold is less than the third difference threshold; A feature calculation module for removing the head and tail segments of the interpolated several actual charging SOC segments; and calculating the actual charging power corresponding to each of the removed several actual charging SOC segments.

5. The health assessment system according to claim 1, the model evaluation module includes: A model construction module for collecting the charging capacity sample data generated during charging by all vehicles of the same vehicle model and the same battery pack model as the vehicle in a fixed threshold period, and the corresponding working condition sample data; using the charging capacity sample data and the working condition sample data to perform fitting training on a machine learning algorithm to obtain the capacity correction model.

6. The health assessment system according to claim 1, wherein the rated capacity calculation module is further configured to process the initial charging operating conditions and standard operating conditions corresponding to the several initial charging SOC segments by using the capacity correction model, and obtain the operating condition correction coefficients corresponding to the several initial charging SOC segments respectively under the standard operating condition based on the processing results, and correct the initial charging power corresponding to the several initial charging SOC segments by using the operating condition correction coefficients corresponding to the several initial charging SOC segments respectively; and obtain the rated charging capacity based on the several actual charging SOC segments and their corrected charging power.

7. A method for health assessment of a power battery, the method comprising: Obtaining actual battery state data; When the cumulative charging time period of the power battery reaches a cumulative duration threshold, preprocessing the actual battery state data, the preprocessing including: data parsing, data conversion, and data cleaning; Performing charging segment segmentation on the actual battery state data to obtain several actual charging SOC segments; and searching for the charging operating conditions corresponding to the several actual charging SOC segments from the actual battery state data; Processing the actual charging operating conditions and standard operating conditions corresponding to the several actual charging SOC segments by using the trained capacity correction model, and obtaining the operating condition correction coefficients corresponding to the several actual charging SOC segments respectively under the standard operating condition based on the processing results; correcting the actual charging power corresponding to the several actual charging SOC segments by using the operating condition correction coefficients corresponding to the several actual charging SOC segments respectively; obtaining the actual charging capacity based on the several actual charging SOC segments and their corrected charging power, specifically including: respectively inputting the actual charging operating conditions and standard operating conditions corresponding to the several actual charging SOC segments into the capacity correction model for processing, and respectively obtaining the first predicted charging capacity of the several actual charging SOC segments under their respective actual charging operating conditions and the second predicted charging capacity under the standard operating condition; comparing the second predicted charging capacity with the first predicted charging capacity corresponding to each of the several actual charging SOC segments respectively to obtain the operating condition correction coefficients corresponding to the several actual charging SOC segments respectively; correcting the actual charging power corresponding to the several actual charging SOC segments by using the operating condition correction coefficients corresponding to the several actual charging SOC segments respectively to obtain the corrected charging power corresponding to the several actual charging SOC segments respectively; and further configured to use the weighted ratio of the several actual charging SOC segments and their corrected charging power as the actual charging capacity; calculating the rated charging capacity of the vehicle battery based on the initial charging state data of the battery in the initial stage of the vehicle operation; and performing battery health SOH assessment by using the ratio of the actual charging capacity to the rated charging capacity, specifically including: collecting the initial charging state data of the battery and obtaining several initial charging SOC segments and their initial charging power corresponding to the initial charging state data of the battery; and determining the rated charging capacity by using the several initial charging SOC segments and their initial charging power.

8. The health assessment method according to claim 7, wherein the actual battery state data includes: Vehicle identification, reporting time, vehicle status, charge and discharge status, total voltage, total current, SOC, temperature; The preprocessing of the actual battery status data specifically includes: Parsing the actual battery status data from unstructured data into structured data; Performing unit conversion and service data conversion on the actual battery status data; Repairing and / or deleting abnormal points in the actual battery status data.

9. The health assessment method according to claim 8, wherein the repairing and / or deleting of abnormal points in the actual battery status data specifically includes: When repairing abnormal points, when the first time differences between the intermediate moment and the front and rear moments are both less than the first difference threshold, repairing the actual battery status data at the intermediate moment to the actual battery status data at the front and rear moments; When deleting abnormal points, deleting the first outliers in the actual battery status data that exceed the corresponding value range; Successively traversing the actual battery status data by using the derivative method to obtain the adjacent change amount between the actual battery status data and the adjacent front and rear data; and detecting the adjacent change amount by using a change amount threshold to determine the second outliers or the outlier intervals corresponding to the second outliers; Processing the second outliers or the outlier intervals by using a preset value as a judgment criterion; and / or dividing the actual battery status data according to an interval division criterion to obtain several intervals; determining the outlier intervals from the several intervals; and processing the outlier intervals by using a preset value as a judgment criterion.

10. The health assessment method according to claim 7 or 9, wherein the splitting of the actual battery status data into several actual charging SOC segments includes: Calculating the third time difference between two adjacent actual battery status data. If the third time difference is greater than or equal to the third difference threshold, splitting the two adjacent actual battery status data as segment endpoints respectively; if the third time difference is less than the third difference threshold, attributing the two adjacent actual battery status data to the same actual charging SOC segment; For each actual charging SOC segment, judging whether the fourth time difference between two adjacent actual battery status data in the actual charging SOC segment is between the fourth difference threshold and the third difference threshold; if so, performing linear interpolation at a preset interval time between the two adjacent actual battery status data, and calculating the SOC value corresponding to the linear interpolation; wherein the fourth difference threshold is less than the third difference threshold; Removing the head and tail segments of the interpolated several actual charging SOC segments; and calculating the actual charging power corresponding to each of the removed several actual charging SOC segments.

11. The health assessment method according to claim 7, wherein the capacity correction model is trained by the following method: Collect the charging capacity sample data generated during charging of all vehicles of the same vehicle model and the same battery pack model as the vehicle in a fixed threshold period, as well as the corresponding working condition sample data; use the charging capacity sample data and the working condition sample data to perform fitting training on a machine learning algorithm to obtain the capacity correction model.

12. The health assessment method according to claim 7, wherein calculating the rated charging capacity of the vehicle battery based on the initial charging state data of the battery in the initial stage of the vehicle operation specifically includes: Use the capacity correction model to process the initial charging working conditions and standard working conditions corresponding to the several initial charging SOC segments, and based on the processing results, obtain the working condition correction coefficients corresponding to the several initial charging SOC segments under the standard working conditions respectively. Use the working condition correction coefficients corresponding to the several initial charging SOC segments respectively to correct the initial charging power corresponding to the several initial charging SOC segments; obtain the rated charging capacity based on the several actual charging SOC segments and their corrected charging power.

13. An electronic device, comprising the health assessment system of the power battery according to any one of claims 1-6.

Images