Methods for determining battery health status, methods for predicting battery life, equipment and systems

Abstract

This application provides a method, lifespan prediction method, device, and system for determining the health status of a battery. The method includes: acquiring battery state data during vehicle operation, wherein the battery state data includes at least: measured voltages corresponding to multiple timestamps; determining target electrochemical parameter values ​​for multiple battery cells in the battery that minimize the error between the measured voltage and the simulated voltage of the electrochemical model based on the battery state data, battery nominal data, and an electrochemical model; determining the inconsistency of the target electrochemical parameters of the multiple battery cells based on multiple sets of target electrochemical parameter values ​​acquired through multiple cycles; determining the health status of the multiple battery cells based on the inconsistency of their target electrochemical parameters; and determining that the battery is abnormal if the health status exceeds a first threshold. This allows for early detection of battery anomalies, improving the accuracy of battery anomaly detection.

Description

Technical Field

[0001] This application relates to the field of power battery technology, and in particular to a method for determining the health status of a battery, a method for predicting its lifespan, an apparatus, and a system. Background Technology

[0002] Battery health diagnostics is a key technology for ensuring the safety, reliability, lifespan, and operational efficiency of battery systems. Currently, related technologies typically diagnose battery abnormalities by analyzing the overall battery pack voltage. However, this method is prone to misjudgment and cannot detect battery anomalies in a timely and accurate manner. Summary of the Invention

[0003] This application provides a method for determining the health status of a battery, a method for predicting its lifespan, an apparatus, and a system to improve the accuracy and timeliness of detecting battery anomalies.

[0004] In a first aspect, embodiments of this application provide a method for determining the health status of a battery. The method includes: acquiring battery status data during vehicle operation, wherein the battery status data includes at least: measured voltages corresponding to multiple timestamps; determining target electrochemical parameter values ​​for multiple battery cells in the battery that minimize the error between the measured voltage and the simulated voltage of the electrochemical model, based on the battery status data, battery nominal data, and an electrochemical model; wherein the electrochemical model is a pseudo-two-dimensional model; determining the inconsistency of the target electrochemical parameters of the multiple battery cells based on multiple sets of target electrochemical parameter values ​​acquired through multiple cycles; determining the health status of the multiple battery cells based on the inconsistency of their target electrochemical parameters; determining that the battery is abnormal if the health status exceeds a preset threshold; and determining the battery is abnormal based on the battery status data, battery nominal data, and electrochemical model. To determine the target electrochemical parameter values ​​for multiple battery cells in a battery that minimize the error between the measured voltage and the simulated voltage of the electrochemical model, the following steps are taken for each battery cell: Based on the initial state of charge in the battery nominal data and battery state data, determine the initial values ​​of multiple types of electrochemical parameters for the battery cell; using the measured voltage of the battery cell as a reference, adjust the initial values ​​of multiple types of electrochemical parameters for the battery cell until the error between the measured voltage and the simulated voltage of the electrochemical model is minimized, thus obtaining the identification results of multiple types of electrochemical parameters for the battery cell; determine the identification results of the target electrochemical parameters from the identification results of the multiple types of electrochemical parameters for the battery cell; and determine the target electrochemical parameter value for the battery cell based on the identification results of the target electrochemical parameters and the preset values ​​of the target electrochemical parameters.

[0005] By employing the aforementioned technical means, after acquiring battery state data during vehicle operation, since the battery state data includes at least multiple timestamps corresponding to measured voltages, target electrochemical parameter values ​​for multiple battery cells can be determined based on the battery state data, battery nominal data, and electrochemical model. These target electrochemical parameter values ​​minimize the error between the measured voltage and the simulated voltage of the electrochemical model. This not only improves the convergence speed and rapidly determines the target electrochemical parameter values ​​for multiple battery cells, but also makes the determined target electrochemical parameter values ​​for multiple battery cells more closely reflect actual conditions. Because the target electrochemical parameters can directly and quantitatively characterize the health of specific materials or interface properties within a battery cell, Therefore, by identifying the inconsistencies in the target electrochemical parameters of multiple battery cells obtained through multiple cycles, the actual condition of the battery cells can be reflected at a more microscopic level. Based on these inconsistencies, the health status of multiple battery cells can be determined. If the health status exceeds a preset threshold, an abnormality is identified. Compared to methods that diagnose battery abnormalities by analyzing the overall battery pack voltage, this approach can determine whether a battery is abnormal before the overall battery pack voltage shows significant abnormalities. This allows for timely and accurate detection of battery abnormalities, thereby improving the accuracy and timeliness of battery anomaly detection.

[0006] In some embodiments, acquiring battery status data during vehicle operation includes: receiving a target data packet sent by the vehicle; wherein the target data packet is generated by packaging target data segments collected by the vehicle when a preset trigger condition is met, and the target data segments meet a preset excitation condition; and performing data preprocessing on the target data packet to obtain battery status data during vehicle operation.

[0007] By using the above-mentioned technical means, the target data segments that meet the preset incentive conditions are received, rather than all the historical data. This can not only effectively alleviate network bandwidth pressure, but also reduce the storage costs of cloud devices, thereby making the management of a fleet of thousands of vehicles economically feasible.

[0008] In some embodiments, the target data segment satisfies a preset excitation condition, including: the current rate change in the target data segment exceeds a preset rate threshold.

[0009] By using the above-mentioned technical means, the preset excitation conditions are limited to include the current rate change amplitude in the target data segment exceeding the preset rate threshold. The target data segment obtained in this way includes obvious current excitation changes. This makes it easier and more accurate to determine the target electrochemical parameter values ​​of multiple battery cells in the battery that minimize the error between the measured voltage and the simulated voltage of the electrochemical model, thereby further improving the accuracy of anomaly detection in the battery.

[0010] In some embodiments, the target data packet includes at least one of the following: vehicle identification code, battery identification information, measured voltage array, total current array, temperature array, timestamp, and initial state of charge.

[0011] Through the aforementioned technical means, vehicle identification codes and battery identification information can distinguish between different vehicles and different battery models, allowing the battery health status determination method to be adapted to the corresponding vehicle and the corresponding battery, avoiding cross-battery parameter mismatches, reducing misjudgments, and improving versatility; timestamps can accurately align the voltage, current, and temperature time-series data of each battery cell, reducing the problem of misalignment and distortion of the determined target electrochemical parameter values ​​of the battery cells; and by utilizing the initial state of charge, the initial electrochemical state of each battery cell can be accurately locked, thereby eliminating identification bias caused by unknown initial states.

[0012] In some embodiments, data preprocessing is performed on the target data packet to obtain battery status data during vehicle operation, including: decrypting and decompressing the target data packet to obtain decompressed data; and allocating and aligning the decompressed data based on multiple battery cells to determine the battery status data during vehicle operation.

[0013] By using the above-mentioned technical means to decrypt and decompress the target data packet, the collected target data fragments can be restored, eliminating data distortion, packet loss, and garbled characters caused by transmission encryption and compression. This provides a real data source for subsequently determining the target electrochemical parameter values ​​of multiple battery cells. In addition, by allocating and aligning the decompressed data based on multiple battery cells, the problem of data mismatch between battery cells can be reduced, further improving the accuracy of anomaly detection in the battery.

[0014] In some embodiments, the method further includes: based on multiple timestamps, using the timestamps as variables, repeatedly executing the step of determining the target electrochemical parameter values ​​of multiple battery cells in the battery that minimize the error between the measured voltage corresponding to the timestamp and the simulated voltage of the electrochemical model, according to battery state data, battery nominal data and electrochemical model, so as to obtain multiple sets of target electrochemical parameter values ​​of multiple battery cells obtained in multiple cycles.

[0015] Through the above-mentioned technical means, one timestamp can correspond to a set of electrochemical parameter values ​​for multiple battery cells, and multiple timestamps can correspond to multiple sets of target electrochemical parameter values ​​for multiple battery cells. This can provide rich data for determining the inconsistency of target electrochemical parameters for multiple battery cells, reduce the occurrence of accidental determination of battery anomalies, and further improve the accuracy of battery anomaly detection.

[0016] In some embodiments, the method further includes: obtaining the number of cycles and the estimated health status under the timestamp; establishing a correlation between the timestamp and the target electrochemical parameter value, the number of cycles and the estimated health status of the battery cell; and storing the correlation between the timestamp and the target electrochemical parameter value, the number of cycles and the estimated health status of the battery cell in a preset time series database.

[0017] By using the above-mentioned technical means, and taking the timestamp as a benchmark, a correlation is established between the timestamp and the target electrochemical parameter values, cycle count, and health estimate of a single battery cell. This allows multiple battery cells to obtain their own unique target electrochemical parameter values, cycle count, and health estimate at the same timestamp, making it easier to observe multiple battery cells at the same time and identify abnormal battery cells more quickly.

[0018] In some embodiments, based on multiple timestamps, the step of establishing the correlation between timestamps and the target electrochemical parameter values, cycle counts, and estimated health status values ​​of battery cells is repeatedly executed to obtain multiple sets of correlations between timestamps and the target electrochemical parameter values, cycle counts, and estimated health status values ​​of battery cells; based on the correlations between multiple sets of timestamps and the target electrochemical parameter values, cycle counts, and estimated health status values ​​of battery cells, a preset time series database is constructed.

[0019] Through the above-mentioned technical means, each battery cell can obtain multiple target electrochemical parameter values, multiple cycle counts, and multiple health status estimates. Each target electrochemical parameter value corresponds to a cycle count and a health status estimate, which is beneficial for observing the health status of each battery cell at the same time.

[0020] In some embodiments, determining the inconsistency of target electrochemical parameters of multiple battery cells based on multiple sets of target electrochemical parameter values ​​obtained through multiple cycles includes: obtaining multiple sets of target electrochemical parameter values ​​of multiple battery cells based on a preset time series database; determining the decay trajectory of the target electrochemical parameters of multiple battery cells and the standard deviation of the target electrochemical parameters of multiple battery cells based on the multiple sets of target electrochemical parameter values ​​of multiple battery cells; and determining the inconsistency of target electrochemical parameters of multiple battery cells based on the decay trajectory of the target electrochemical parameters of multiple battery cells and the standard deviation of the target electrochemical parameters of multiple battery cells.

[0021] By employing the aforementioned technical methods, after obtaining multiple sets of target electrochemical parameter values ​​for each battery cell, a decay trajectory corresponding to the target electrochemical parameters of each battery cell can be plotted based on these values. This is essentially a decay trajectory map of each battery cell throughout its entire lifecycle, which is beneficial for analyzing the decay status of each battery cell. Furthermore, the standard deviation of the target electrochemical parameters for each battery cell can be determined based on these multiple sets of values, which helps confirm the internal stability of each battery cell and the possibility of early failures. By analyzing the decay trajectories and standard deviations of the target electrochemical parameters of multiple battery cells, inconsistencies in the target electrochemical parameters of multiple batteries can be identified. Compared to related technologies that use voltage anomalies to characterize battery anomalies, this approach can significantly advance the warning time of battery anomalies. That is, battery anomalies can be detected in the early stages before obvious voltage anomalies appear, which is beneficial for "prevention" and reduces the occurrence of safety accidents such as thermal runaway. Thus, it further improves both the accuracy and timeliness of battery anomaly detection.

[0022] In some embodiments, determining the decay trajectory of the target electrochemical parameters of multiple battery cells and the standard deviation of the target electrochemical parameters of multiple battery cells based on multiple sets of target electrochemical parameter values ​​of multiple battery cells includes: for each battery cell, performing the following steps: plotting a scatter plot of the target electrochemical parameters of the battery cell with the timestamp as the horizontal axis and the multiple sets of target electrochemical parameter values ​​of the battery cell as the vertical axis; smoothing and fitting the scatter plot of the target electrochemical parameters of the battery cell to obtain the decay trajectory of the target electrochemical parameters of the battery cell; and calculating the standard deviation of the multiple sets of target electrochemical parameter values ​​of the battery cell to obtain the standard deviation of the target electrochemical parameters of the battery cell.

[0023] Using the aforementioned technical means, for each battery cell, by using multiple sets of target electrochemical parameter values ​​for each battery cell to plot the decay trajectory of each battery cell, the differences in the overall aging trajectory morphology of each battery cell can be measured. In addition, by calculating the standard deviation of multiple sets of target electrochemical parameter values ​​for each battery cell, the dispersion of the target electrochemical parameter values ​​of each battery cell at the current moment can be monitored.

[0024] In some embodiments, determining the inconsistency of target electrochemical parameters of multiple battery cells based on the decay trajectory of target electrochemical parameters of multiple battery cells and the standard deviation of target electrochemical parameters of multiple battery cells includes: for each battery cell, performing the following steps: determining the target trajectory difference of target electrochemical parameters of battery cells based on the decay trajectory of target electrochemical parameters of battery cells and a preset reference trajectory; determining the target standard deviation of target electrochemical parameters of battery cells based on the standard deviation of target electrochemical parameters of battery cells; and performing a weighted calculation on the target trajectory difference and the target standard deviation to obtain the inconsistency of target electrochemical parameters of battery cells.

[0025] By employing the aforementioned techniques, and based on the decay trajectory of the target electrochemical parameters for each battery cell and a preset reference trajectory, the target trajectory difference of the target electrochemical parameters for each individual battery cell can be determined. This allows for the identification of inconsistencies in the aging rhythm of each individual battery cell. By fusing the target standard deviation and the target trajectory difference of the target electrochemical parameters for each individual battery cell, the inconsistencies in the target electrochemical parameters of each individual battery cell can be determined. This not only allows attention to the dispersion of the target electrochemical parameter values ​​of each individual battery cell at the current moment, but also to the measurement of the differences in the overall aging trajectory pattern of each individual battery cell. This enables the capture of complex inconsistency patterns such as "rapid aging followed by slow aging," further improving the accuracy of anomaly detection for batteries.

[0026] In some embodiments, determining the target trajectory difference of the target electrochemical parameters of a battery cell based on the decay trajectory of the target electrochemical parameters of the battery cell and a preset reference trajectory includes: obtaining multiple trajectory difference values ​​of the target electrochemical parameters of the battery cell based on the distance between the decay trajectory of the target electrochemical parameters of the battery cell and the preset reference trajectory; and averaging the multiple trajectory difference values ​​of the target electrochemical parameters of the battery cell to obtain the target trajectory difference of the target electrochemical parameters of the battery cell.

[0027] Using the aforementioned technical methods, for each battery cell: by calculating the distance between the decay trajectory of the target electrochemical parameters of the battery cell and the preset reference trajectory, multiple trajectory difference values ​​of the battery cell can be obtained. This can eliminate normal aging factors and accurately characterize the abnormal degradation degree and consistency deviation of the battery cell. Then, by averaging the multiple trajectory difference values ​​of the battery cell, the random disturbances of the battery cell can be smoothed, and a system-level trajectory difference degree (i.e., target trajectory difference degree) characterizing the overall deviation of the battery cell from the standard aging degree can be obtained. This achieves the organic unity of battery cell anomaly identification and battery consistency assessment, which can not only improve the sensitivity and robustness of early anomaly detection of the battery, but also provide a reliable basis for battery safety early warning and health management.

[0028] In some embodiments, determining the target standard deviation of the target electrochemical parameters of a battery cell based on the standard deviation of the target electrochemical parameters of the battery cell includes: normalizing the standard deviation of the target electrochemical parameters of the battery cell to obtain the target standard deviation of the target electrochemical parameters of the battery cell.

[0029] By normalizing the standard deviation of the target electrochemical parameters of a single battery cell using the above-mentioned technical means, a dimensionless feature characterizing the relative fluctuation of the target electrochemical parameters can be obtained. This not only eliminates the difference in magnitude but also achieves unified evaluation of multiple features, thereby further improving the accuracy of battery anomaly detection.

[0030] In some embodiments, when the target electrochemical parameters include multiple types, the method further includes: determining that the battery is abnormal when the inconsistency of any type of target electrochemical parameter in any of the multiple battery cells meets a preset condition.

[0031] By employing the aforementioned technical means, when the inconsistency of any type of target electrochemical parameter in any of the multiple battery cells meets preset conditions, it can be determined that the battery is abnormal. This ensures timely detection of battery abnormalities and further improves the accuracy of battery abnormality detection.

[0032] Secondly, embodiments of this application provide a battery lifetime prediction method, which includes: identifying the abnormal battery cell after an abnormality occurs in the battery; matching the target electrochemical parameter values ​​of the abnormal battery cell with a failure boundary library to determine a first degradation mode of the abnormal battery cell; extrapolating the decay trajectory of the target electrochemical parameters of the abnormal battery cell based on the first degradation mode to determine a baseline degradation trajectory of the abnormal battery cell; using a probabilistic prediction model to predict the lifetime of the baseline degradation trajectory of the abnormal battery cell to obtain a probability distribution of the remaining lifetime of the abnormal battery cell; and determining the battery lifetime prediction result based on the probability distribution of the remaining lifetime of the abnormal battery cell.

[0033] Using the aforementioned techniques, a degradation mode that best matches the target electrochemical parameter values ​​of the abnormal battery cell is found from the failure boundary database. Then, based on the first degradation mode, the decay trajectory of the target electrochemical parameters of the abnormal battery cell is extrapolated to determine the baseline degradation trajectory of the abnormal battery cell. Since the degradation mode is matched to the target electrochemical parameters of the abnormal battery cell, the obtained baseline degradation trajectory is a relatively accurate one. However, because the first degradation mode is an ideal mode, the obtained degradation trajectory is also an ideal one. In practical applications, there are many uncertainties. If the abnormal battery cell is analyzed directly based on the baseline degradation trajectory... The lifespan of individual battery cells may not align with actual conditions, leading to inaccurate analysis results. Therefore, it is necessary to incorporate a probabilistic prediction model. This model applies multiple uncertainties consistent with actual operating conditions to the baseline degradation trajectory, yielding a probability distribution of the remaining lifespan of abnormal battery cells. This makes the probability distribution of the remaining lifespan of abnormal battery cells more consistent with actual operating conditions, thereby making the subsequent lifespan prediction results of abnormal battery cells more accurate. This not only significantly improves the practicality and reliability of predictions but also enables asset managers to quantify risks and make more scientific decisions regarding replacement, warranty, residual value assessment, and tiered utilization, thereby improving asset operation efficiency.

[0034] In some embodiments, determining the battery life prediction result based on the probability distribution of the remaining lifespan of the abnormal battery cells includes: determining the remaining lifespan of the abnormal battery cells based on the probability distribution of the remaining lifespan of the abnormal battery cells, and determining the remaining lifespan of the abnormal battery cells as the battery life prediction result.

[0035] By employing the aforementioned technical means, since the overall remaining lifespan of a battery is determined by the remaining lifespan of the abnormal individual battery cells, determining the remaining lifespan of the abnormal individual battery cells as the battery lifespan prediction result can make the battery lifespan prediction result more consistent with actual usage patterns.

[0036] In some embodiments, identifying abnormal battery cells in the battery includes: identifying battery cells in the battery whose trajectory difference value of the target electrochemical parameter exceeds a preset value as abnormal battery cells; and / or identifying battery cells in the battery with the largest standard deviation of the target electrochemical parameter as abnormal battery cells.

[0037] Through the aforementioned technical means, the trajectory difference of the target parameters of a battery cell can reflect the "deviation" of the cell's trend, and the standard deviation of the target electrochemical parameters can reflect the "fluctuation instability" of the cell. Therefore, by using the trajectory difference and / or the standard deviation of the target electrochemical parameters, the identified abnormal battery cells are more accurate, improving the precision of locating abnormal battery cells within the battery. Whether identifying battery cells whose trajectory difference of the target electrochemical parameters exceeds a preset difference as abnormal battery cells, or identifying the battery cell with the largest standard deviation of the target electrochemical parameters as abnormal battery cells, precise location of abnormal battery cells can be achieved, thereby improving the accuracy of locating abnormal battery cells within the battery.

[0038] In some embodiments, after identifying the abnormal battery cell, the method further includes: matching the target electrochemical parameter value of the abnormal battery cell with a fault knowledge base to determine the root cause of the abnormality of the abnormal battery cell; and sending the root cause of the abnormality of the abnormal battery cell to the user equipment.

[0039] By using the above-mentioned technical means, the target electrochemical parameter values ​​of abnormal battery cells are matched with the fault knowledge base. The fault cause (i.e., the root cause of the abnormality) that matches the changing trend of the target electrochemical parameters of the abnormal battery cell can be matched from the fault knowledge base. Power batteries follow the "weakest link effect". The fault cause of the power battery is the root cause of the abnormal battery cell. Sending the root cause of the abnormal battery cell to the user equipment allows the user to clearly know what caused the power battery failure, which is conducive to targeted maintenance and improves maintenance efficiency.

[0040] Thirdly, embodiments of this application also provide a cloud device, including a receiving circuit and a processing circuit, wherein: the receiving circuit is used to acquire battery state data during vehicle operation, wherein the battery state data includes at least: measured voltages corresponding to multiple timestamps; the processing circuit is used to determine, based on the battery state data, battery nominal data, and an electrochemical model, the target electrochemical parameter values ​​of multiple battery cells in the battery that minimize the error between the measured voltage and the simulated voltage of the electrochemical model; wherein the electrochemical model is a pseudo-two-dimensional model; the processing circuit is further used to determine the inconsistency of the target electrochemical parameters of multiple battery cells based on multiple sets of target electrochemical parameter values ​​acquired through multiple cycles; and to determine the health status of multiple battery cells based on the inconsistency of the target electrochemical parameters of multiple batteries; and to determine the health status of multiple battery cells when the health status is large. Under a preset threshold, an anomaly is determined in the battery. The processing circuit is also used to obtain the target electrochemical parameter value of each battery cell according to the following process: Based on the initial state of charge in the battery nominal data and battery state data, determine the initial values ​​of multiple types of electrochemical parameters of the battery cell; Using the measured voltage of the battery cell as a reference, adjust the initial values ​​of multiple types of electrochemical parameters of the battery cell until the error between the measured voltage of the battery cell and the simulated voltage of the electrochemical model is minimized, thereby obtaining the identification results of multiple types of electrochemical parameters of the battery cell; Determine the identification result of the target electrochemical parameter from the identification results of multiple types of electrochemical parameters of the battery cell; Determine the target electrochemical parameter value of the battery cell according to the identification result of the target electrochemical parameter and the preset value of the target electrochemical parameter.

[0041] By employing the aforementioned technical methods, after acquiring battery state data during vehicle operation, since the battery state data includes at least multiple timestamps corresponding to measured voltages, target electrochemical parameter values ​​for multiple battery cells can be determined based on the battery state data, battery nominal data, and electrochemical model. These target electrochemical parameter values ​​minimize the error between the measured voltage and the simulated voltage of the electrochemical model. This not only improves the convergence speed and rapidly determines the target electrochemical parameter values ​​for multiple battery cells, but also makes the determined target electrochemical parameter values ​​for multiple battery cells more closely reflect actual conditions. This is because target electrochemical parameters can directly and quantitatively characterize the health of specific materials or interface properties within a battery cell. Therefore, by determining the inconsistency of target electrochemical parameters among multiple battery cells obtained through multiple cycles, the actual situation of the battery cells can be reflected at a more microscopic level. Based on the inconsistency of target electrochemical parameters among multiple battery cells, the health status of multiple battery cells can be determined. If the health status is greater than a preset threshold, an abnormality is determined to exist in the battery. Compared with the method of diagnosing whether the battery is abnormal by analyzing the overall voltage of the battery pack, this method can draw conclusions about whether the battery is abnormal before the overall voltage of the battery pack becomes significantly abnormal. It can detect battery abnormalities in a timely and accurate manner, thereby improving the accuracy and timeliness of battery abnormality detection.

[0042] Fourthly, this application also provides a collaborative system, including a vehicle and a cloud device, wherein: the vehicle is used to extract target data segments from historically collected data when the vehicle's state meets preset trigger conditions, and to perform data alignment and packaging on the target data segments to generate target data packets and upload them to the cloud device; the cloud device is used to receive the target data packets sent by the vehicle, perform data preprocessing on the target data packets to obtain battery state data during vehicle operation, wherein the battery state data includes at least: measured voltages corresponding to multiple timestamps; and, based on the battery state data, battery nominal data, and electrochemical model, determining the target electrochemical parameter values ​​of multiple battery cells in the battery that minimize the error between the measured voltage and the simulated voltage of the electrochemical model; wherein the electrochemical model is a pseudo-two-dimensional model; and, based on multiple sets of target electrochemical parameter values ​​of multiple battery cells obtained through multiple cycles, determining the inconsistency of the target electrochemical parameters of multiple battery cells. Based on the inconsistency of target electrochemical parameters of multiple battery cells, the health status of multiple battery cells is determined; if the health status exceeds a preset threshold, an anomaly is determined in the battery; the cloud device is also used to obtain the target electrochemical parameter value of each battery cell according to the following process: based on the initial state of charge in the battery nominal data and battery state data, the initial values ​​of multiple types of electrochemical parameters of the battery cell are determined; using the measured voltage of the battery cell as a reference, the initial values ​​of multiple types of electrochemical parameters of the battery cell are adjusted until the error between the measured voltage of the battery cell and the simulated voltage of the electrochemical model is minimized, thus obtaining the identification results of multiple types of electrochemical parameters of the battery cell; the identification results of the target electrochemical parameters are determined from the identification results of multiple types of electrochemical parameters of the battery cell; and the target electrochemical parameter value of the battery cell is determined according to the identification results of the target electrochemical parameters and the preset values ​​of the target electrochemical parameters.

[0043] Through the aforementioned technical means, vehicles send target data segments to the cloud, rather than all historical data. This effectively alleviates network bandwidth pressure and reduces cloud storage costs, making the management of fleets of thousands of vehicles economically feasible. After acquiring battery status data during vehicle operation, the cloud can determine the target electrochemical parameter values ​​of multiple battery cells that minimize the error between the measured voltage and the simulated voltage of the electrochemical model, based on the battery status data, battery nominal data, and electrochemical model. This not only improves the convergence speed and quickly determines the target electrochemical parameter values ​​of multiple battery cells but also makes the determined target electrochemical parameter values ​​of multiple battery cells more closely reflect actual conditions. Targeted electrochemical parameters can directly and quantitatively characterize the health of specific materials or interfaces within a battery cell. Therefore, by determining the inconsistency of target electrochemical parameters among multiple battery cells obtained from multiple cycles, the actual condition of the battery cells can be reflected at a more microscopic level. Based on the inconsistency of target electrochemical parameters among multiple battery cells, the health status of multiple battery cells can be determined. If the health status exceeds a preset threshold, an abnormality is determined. Compared to diagnosing battery abnormalities by analyzing the overall battery pack voltage, this approach can draw conclusions about whether a battery is abnormal before the overall battery pack voltage becomes significantly abnormal, enabling timely and accurate detection of battery abnormalities and improving the accuracy and timeliness of battery anomaly detection.

[0044] It should be understood that the above general description and the following detailed description are merely exemplary and explanatory, and are not intended to limit the technical solutions of this application. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of the structure of an electrochemical model provided in an embodiment of this application; Figure 2 This is a flowchart illustrating a method for determining the health status of a battery according to an embodiment of this application. Figure 1 ; Figure 3 This is a flowchart illustrating a method for determining the health status of a battery according to an embodiment of this application. Figure 2 ; Figure 4 This is a flowchart illustrating a method for determining the health status of a battery according to an embodiment of this application. Figure 3 ; Figure 5 This is a flowchart illustrating a method for determining the health status of a battery according to an embodiment of this application. Figure 4 ; Figure 6This is a flowchart illustrating a battery life prediction method provided in an embodiment of this application. Figure 1 ; Figure 7 This is a flowchart illustrating a battery life prediction method provided in an embodiment of this application. Figure 2 ; Figure 8 This is a flowchart illustrating a battery life prediction method provided in an embodiment of this application. Figure 3 ; Figure 9 This is a schematic diagram of the structure of a cloud device provided in an embodiment of this application. Figure 1 ; Figure 10 This is a schematic diagram of the structure of a collaborative system provided in an embodiment of this application; Figure 11 This is a flowchart illustrating a method for determining the health status of a battery according to an embodiment of this application. Figure 5 ; Figure 12 This is a flowchart illustrating a method for detecting battery anomalies provided in an embodiment of this application; Figure 13 This is a schematic diagram of the trajectory of the target electrochemical parameter value of a normal battery cell provided in the embodiments of this application; Figure 14 This is a schematic diagram of the trajectory of the target electrochemical parameter value of an abnormal battery cell provided in an embodiment of this application; Figure 15 This is a schematic diagram of the structure of an abnormal battery cell in a battery provided in an embodiment of this application; Figure 16 This is a schematic diagram of the baseline degradation trajectory of an abnormal battery cell provided in an embodiment of this application; Figure 17 This is a schematic diagram of the structure of a cloud device provided in an embodiment of this application. Figure 2 . Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0047] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0048] In the following description, references to "some embodiments," "this embodiment," "this application embodiment," and examples, etc., describe a subset of all possible embodiments. However, it is understood that "some embodiments" may be the same subset or different subset of all possible embodiments and may be combined with each other without conflict.

[0049] The descriptions such as "first," "second," and "third" appearing in the embodiments of this application do not have a specific meaning (such as no order, nor do they indicate a special limitation on the number of devices in the embodiments of this application), but are merely for the purpose of clearly describing the embodiments of this application and do not constitute any limitation on the embodiments of this application.

[0050] Before providing a more detailed description of the embodiments of this application, the nouns and terms that may be involved in the embodiments of this application will be explained. The nouns and terms involved in the embodiments of this application are subject to the following interpretations.

[0051] (1) Electrochemical Health Factor (EHF): A standardized, dimensionless or normalized index derived from electrochemical model parameters, used to directly and quantitatively characterize the health of specific materials or interfaces within a battery. Electrochemical health factors may include, for example, the retention rate of the negative electrode solid-phase diffusion coefficient EHF_Dsn, the charge transfer impedance growth rate EHF_Rct, and the positive and negative electrode attenuation asymmetry index EHF_Balance, etc.

[0052] (2) Standard data package: A segment of continuous operating data with a fixed duration (e.g., 300 seconds) is extracted by the vehicle-side Battery Management System (BMS) according to preset rules (e.g., including a charge-discharge change from low rate to high rate). This operating data includes data such as the voltage, current and temperature of the power battery.

[0053] (3) Two-stage parameter identification algorithm: The cloud device can call the electrochemical model to perform two-stage parameter identification for each cell in the power battery. In the first stage, all parameters (including thermodynamic and kinetic parameters) are identified under mild operating conditions. In the second stage, the thermodynamic parameters are fixed, and the kinetic parameters are identified under different dynamic operating conditions. This algorithm ensures the accuracy of parameter extraction from short-term data.

[0054] For example, the electrochemical model can be a pseudo two-dimensional (P2D) model, a classic electrochemical model in the field of lithium-ion battery simulation, proposed by Doyle, Fuller and Newman in 1993.

[0055] like Figure 1 As shown, the main assumptions of the P2D model include: the active material of the positive electrode 103 includes aluminum current collector 1031, which is composed of spherical particles with the same radius; the active material of the negative electrode 102 includes copper current collector 1021, which is composed of spherical particles with the same radius; the internal reaction of the battery occurs only in the solid and liquid phases; the electrochemical reaction occurs only in the thickness direction of the battery; the influence of the double layer effect is ignored; and the volume fraction of the liquid phase in the battery remains unchanged.

[0056] This model simplifies the complex electrode structure into a one-dimensional geometric model consisting of a positive electrode 103, a separator 104, and a negative electrode 102 through simplified assumptions. These assumptions include: the electrode and separator 104 are porous structures, described by solid volume fraction; the active particles are uniformly spherically distributed, represented by radius r; the electrolyte 101 uniformly fills the pores, described by porosity; and side reactions are ignored, with lithium ions (Li+) only participating in insertion and extraction (i.e., charging and discharging). This model highlights key features and simplifies the modeling process using readily measurable macroscopic parameters.

[0057] The P2D model primarily simulates the migration and diffusion behavior of lithium ions in the electrodes and electrolyte during battery charging and discharging, satisfying the three major conservation conditions: mass conservation, charge conservation, and electrochemical kinetic equilibrium. Its core governing equations include: Fick's second law describing solid-phase lithium ion diffusion, concentrated solution theory equations describing liquid-phase lithium ion diffusion and migration, Ohm's law equations describing the potentials of the solid and liquid phases, and the Butler-Volmer kinetic equations describing electrochemical reactions.

[0058] (4) Parameter decay trajectory: The curve of a certain EHF value (e.g., EHF_Dsn) of a single cell changes with the cumulative equivalent full cycle number or time, reflecting the degradation process of this characteristic parameter.

[0059] (5) Dynamic Time Warping (DTW) Distance: An algorithm for measuring the similarity between two time series that may have different lengths. In the embodiments of this application, it is used to calculate the morphological difference between the decay trajectory curve of each battery cell and the preset decay trajectory curve of the power battery, so that even if the aging "rhythm" of each battery cell is different, it can be effectively compared.

[0060] (6) Failure Boundary Library: A library of rules or models learned from historical data that defines the conditions under which an EHF combination can be considered to have reached the end of life (EOL). For example, when the capacity of a battery cell decays to 80%, it can be considered to have reached the end of life.

[0061] (7) Monte Carlo simulation: a statistical method for obtaining numerical results through repeated random sampling. In the embodiments of this application, it is used to simulate the impact of future use uncertainty on life prediction, thereby generating a probability distribution of Remaining Useful Life (RUL).

[0062] (8) Equivalent full cycle: The number of partial charge-discharge cycles converted into the number of complete charge-discharge cycles, used to unify the aging degree assessment standard (e.g., charging from 20% to 80% is converted into 0.6 equivalent full cycles).

[0063] (9) Instantaneous divergence: The standard deviation of a certain EHF value of all battery cells in the power battery at a certain cycle time, which represents the inconsistency of the values ​​at the current time.

[0064] (10) Target trajectory difference: The average distance between a certain EHF trajectory of each battery cell calculated by DTW distance and the preset attenuation trajectory of a certain EHF of the power battery, which represents the inconsistency of aging rhythm.

[0065] (11) Comprehensive Inconsistency Index (UI): A weighted index that integrates instantaneous divergence and morphological difference. α is the weight coefficient of instantaneous divergence and β is the weight coefficient of target trajectory difference. The comprehensive inconsistency index can fully quantify inconsistency.

[0066] (12) Trajectory extrapolation: Based on a certain EHF trajectory of an existing battery cell, the future trend is fitted using an exponential decay model to predict the time when a certain EHF of the battery cell reaches the failure boundary.

[0067] (13) Distributed parameter solving engine: A parallel computing engine deployed in a cloud cluster, which can simultaneously process parameter identification tasks of multiple vehicles and multiple units, improving computing efficiency.

[0068] (14) Vehicle Identification Number (VIN): Used to uniquely identify a vehicle and associate power battery and battery cell data.

[0069] (15) Smooth spline fitting: A method of fitting scattered data by piecewise polynomials, which can generate smooth decay trajectory curves and eliminate data noise interference.

[0070] (16) Probability distribution: The probability distribution of remaining lifetime (such as normal distribution), which can output the remaining lifetime under different confidence levels (such as 300 cycles remaining under 90% confidence level).

[0071] (17) Cascade utilization: The reuse of retired power batteries (capacity ≤80%) for energy storage, low-speed electric vehicles and other scenarios requires precise sorting based on their health status.

[0072] (18) Prior information: Known information such as the global parameter database mean of the battery model and the initial state of charge (SOC) of the segment is used to initialize the parameter identification process and improve the convergence speed.

[0073] (19) Individual cell equalization: The technique of adjusting the charging and discharging current of each individual cell through equalization circuit to reduce voltage differences can only temporarily eliminate surface inconsistency.

[0074] (20) Asset management: Track, evaluate and optimize the entire life cycle of the battery pack, including maintenance plan formulation, residual value assessment and replacement decision-making.

[0075] (21) Preventive maintenance: Based on early diagnosis and life prediction results, targeted maintenance is carried out before the battery fails, reducing operation and maintenance costs.

[0076] Currently, new energy batteries are being used more and more widely in daily life and industry. They are not only used in energy storage systems for hydropower, thermal power, wind power, and solar power plants, but also extensively in electric vehicles such as electric bicycles, electric motorcycles, and electric cars, as well as in aerospace and other fields. With the continuous expansion of the application areas of power batteries, the market demand is also constantly increasing.

[0077] In this embodiment, the battery can be a single battery cell or a battery pack composed of multiple battery cells. A single battery cell refers to a basic unit (or "cell") capable of converting chemical energy into electrical energy, and can be used to manufacture battery modules or battery packs to supply power to electrical devices. A single battery cell can be a rechargeable battery, which is a battery cell that can be recharged after discharge to reactivate its active materials and continue to be used. A single battery cell can be a lithium-ion battery, sodium-ion battery, sodium-lithium-ion battery, lithium metal battery, sodium metal battery, lithium-sulfur battery, magnesium-ion battery, nickel-metal hydride battery, nickel-cadmium battery, lead-acid battery, etc., and is not limited to any particular type.

[0078] In this embodiment, the battery may also be a single physical module comprising one or more battery cells to provide higher voltage and capacity. When there are multiple battery cells, they can be connected in series, parallel, or mixed via a busbar.

[0079] To facilitate understanding of the technical solutions of the embodiments of this application, the relevant technologies or terms of the embodiments of this application are described below. The following relevant technologies or terms are optional solutions and can be combined with the technical solutions of the embodiments of this application in any way, and all of them fall within the protection scope of the embodiments of this application.

[0080] Battery health diagnostics is a key technology for ensuring the safety, reliability, lifespan, and operational efficiency of battery systems. Currently, related technologies typically diagnose battery abnormalities by analyzing the overall battery pack voltage. However, this method is prone to misjudgment and cannot detect battery anomalies promptly and accurately. This is because if the overall pack voltage shows no significant abnormality, it is assumed that the battery is not malfunctioning. This method is inherently lagging, leading to the inability to detect battery abnormalities in a timely and accurate manner.

[0081] Based on this, in order to improve the accuracy and timeliness of battery anomaly detection, embodiments of this application provide the following battery health status determination method, battery life prediction method, cloud device, and collaborative system. The battery health status determination method can be applied to a cloud device. First, battery status data during vehicle operation is acquired, including at least measured voltages corresponding to multiple timestamps. Second, based on the battery status data, battery nominal data, and electrochemical model, target electrochemical parameter values ​​for multiple battery cells in the battery are determined to minimize the error between the measured voltage and the simulated voltage of the electrochemical model. Then, based on multiple sets of target electrochemical parameter values ​​obtained from multiple cycles of acquisition for multiple battery cells, the inconsistency of the target electrochemical parameters for multiple battery cells is determined. Next, based on the inconsistency of the target electrochemical parameters for multiple battery cells, the health status of multiple battery cells is determined. If the health status exceeds a preset threshold, an anomaly is determined in the battery.

[0082] Understandably, since target electrochemical parameters can directly and quantitatively characterize the health of specific materials or interface properties within a battery cell, the inconsistency of target electrochemical parameters among multiple battery cells obtained from multiple cycles can reflect the actual situation of the battery cells at a more microscopic level. Based on this inconsistency, the health status of multiple battery cells can be determined. If the health status exceeds a preset threshold, an anomaly is identified. Compared to diagnosing battery anomalies by analyzing the overall battery pack voltage, this approach allows for the conclusion of whether a battery is abnormal before a significant anomaly occurs in the overall battery pack voltage. This enables timely and accurate detection of battery anomalies, thereby improving the accuracy and timeliness of anomaly detection.

[0083] The method for determining the health status of a battery provided in this application will be described in detail below with reference to the accompanying drawings.

[0084] In one embodiment of this application, a method for determining the health status of a battery is provided, applied to a cloud device, such as... Figure 2 As shown, the health diagnosis method for this battery includes the following steps: S101. Obtain battery status data during vehicle operation.

[0085] The battery status data includes at least the measured voltage corresponding to multiple timestamps.

[0086] In some optional embodiments, battery status data during vehicle operation can be obtained through the following steps: S101-1 Receive the target data packet sent by the vehicle.

[0087] It should be noted that the target data packet is generated by packaging the target data fragments collected by the vehicle when the preset trigger conditions are met, and the target data fragments meet the preset excitation conditions.

[0088] For example, the preset triggering conditions may include periodic triggering, event triggering, and exception triggering. When any of the triggering conditions is met, the acquisition and uploading process of the target data segment is initiated. Among them, periodic triggering: every fixed calendar time (such as 7 days) or fixed mileage (such as 5000 kilometers).

[0089] Among them, the event trigger is the detection of the end of a complete driving loop that covers a typical SOC range (such as 30%-70%).

[0090] Among them, abnormal triggering: the BMS local algorithm initially judges that the voltage or temperature behavior of a certain unit is slightly abnormal.

[0091] For example, the target data segment satisfies the preset excitation conditions, including: the current ratio change in the target data segment exceeds the preset ratio threshold, that is, the target data segment must contain obvious current excitation changes.

[0092] For example, the preset multiplier threshold can be 1x (C-rate, C).

[0093] In this context, a significant change in current excitation refers to a current that is not a smooth fluctuation but rather exhibits substantial changes in charge and discharge amplitude. For example, a target data segment showing a current rate change exceeding 1C under typical vehicle driving conditions of acceleration-coasting-braking. During acceleration, the battery discharges with a large current; during coasting, the current is almost zero; and during braking, the battery charges with a large current. This means that the current excitation change is significant under typical driving conditions of acceleration-coasting-braking, and the target device can send this target data segment to the cloud device. Another example is a target data segment showing a charge-discharge change from a low rate to a high rate. Understandably, selecting target data segments containing significant changes in current excitation ensures the "observability" of subsequent electrochemical model identification of electrochemical parameters (i.e., microscopic parameters).

[0094] It should be noted that the target data segment is formed by selecting a preset time period of data from a set of historical operational data. For example, the preset time period ranges from 5 minutes to 10 minutes.

[0095] It should be noted that the battery in this embodiment can be a power battery, an energy storage battery, an auxiliary battery, etc., and this application does not make any special limitation in this regard.

[0096] It is understood that in this embodiment, the preset excitation condition is limited to include the current rate change exceeding the preset rate threshold in the target data segment. The target data segment obtained in this way includes obvious current excitation changes, which makes it easier and more accurate to determine the target electrochemical parameter values ​​of multiple battery cells in the battery that minimize the error between the measured voltage and the simulated voltage of the electrochemical model, thereby further improving the accuracy of anomaly detection in the battery.

[0097] For example, the target data packet includes at least one of the following: vehicle identification number, battery identification information, measured voltage array, total current array, temperature array, timestamp, and starting state of charge (SOC).

[0098] It should be noted that the target data segment sent by the vehicle to the cloud device is a standard data packet. The BMS in the vehicle continuously monitors the vehicle's operating status and stores the collected operating data (such as voltage, current, and temperature). When the BMS detects that the uploaded operating data meets the preset trigger conditions, the vehicle's BMS retrieves historical operating data from the cached historical operating data, starting from the current time and tracing back through the historical operating data within a preset time period to obtain the operating dataset. It then checks whether the operating dataset contains significant changes in current excitation. If so, it performs time alignment, invalid value removal, and lightweight compression on the operating dataset, packaging it into a standard data packet containing information such as the vehicle VIN (Vehicle Identification Number), battery identifier (ID) (i.e., battery identification information), measured voltage array, total current array, temperature array, timestamp, and initial SOC. This standard data packet is then uploaded to the cloud device via the cellular network. This standard data packet is called the target data segment.

[0099] It should also be noted that the measured voltage array in the standard data package refers to the measured voltage of each cell in the battery; the total current array refers to the total current of the battery; the temperature array refers to the temperature of each cell in the battery; and the initial SOC refers to the initial SOC of each cell in the battery.

[0100] Understandably, in this embodiment, the vehicle identification code and battery identification information can distinguish between different vehicles and different battery models, allowing the battery health status determination method to be adapted to the corresponding vehicle and the corresponding battery, avoiding cross-battery parameter mismatch, reducing misjudgments, and improving versatility; the timestamp can accurately align the voltage, current, and temperature time series data of each battery cell, reducing the problem of misalignment and distortion of the determined target electrochemical parameter values ​​of the battery cells; using the initial state of charge, the initial electrochemical state of each battery cell can be accurately locked, the initial lithium concentration distribution and the initial open circuit potential can be determined, and the identification bias caused by the unknown initial state can be eliminated.

[0101] S101-2. Perform data preprocessing on the target data packet to obtain battery status data during vehicle operation.

[0102] In some optional embodiments, the target data packet can be decrypted and decompressed to obtain decompressed data; the decompressed data is then allocated and aligned based on multiple battery cells to determine the battery status data during vehicle operation.

[0103] For example, a distributed parameter calculation engine, such as a Kubernetes cluster-based distributed parameter calculation engine, can be deployed in a cloud device. After receiving the target data packet, the distributed parameter calculation engine preprocesses the target data packet to obtain battery status data during vehicle operation.

[0104] In some embodiments, preprocessing may include validity verification, in addition to decryption, decompression, and alignment.

[0105] The distributed parameter calculation engine distributes the total current array in the target data packet to each battery cell according to the battery topology, and strictly aligns it with the voltage and temperature data of each battery cell according to the timestamp. In other words, each battery cell has corresponding current, voltage, and temperature data at each timestamp of the preset time, which lays the foundation for identifying the electrochemical parameters of each battery cell at each moment of the preset time.

[0106] It is understood that in this embodiment, decrypting and decompressing the target data packet can restore the collected target data fragments, eliminate data distortion, packet loss, and garbled characters caused by transmission encryption and compression, and provide a real data source for subsequently determining the target electrochemical parameter values ​​of multiple battery cells. In addition, the allocation and alignment of the decompressed data based on multiple battery cells can reduce the problem of data mismatch between battery cells and further improve the accuracy of anomaly detection of the battery.

[0107] It is understood that in this embodiment, the target data segment that meets the preset incentive conditions is received, rather than all the historical data. This can not only effectively alleviate network bandwidth pressure, but also reduce the storage cost of cloud devices, thereby making the management of a fleet of thousands of vehicles economical and feasible.

[0108] S102. Based on the battery state data, battery nominal data, and electrochemical model, determine the target electrochemical parameter values ​​of multiple battery cells in the battery that minimize the error between the measured voltage and the simulated voltage of the electrochemical model.

[0109] It is understandable that the full name of the target electrochemical parameter value is the parameter value of the target electrochemical parameter.

[0110] It should be noted that a single battery cell can have many types of electrochemical parameters. If one type of electrochemical parameter is selected as the target electrochemical parameter from among multiple types, then there is one target electrochemical parameter and one type of target electrochemical parameter. If multiple types of electrochemical parameters are selected as the target electrochemical parameter, then the number of target electrochemical parameters is the same as the number of electrochemical parameters in the battery cell, and the number of target electrochemical parameter types is the same as the number of electrochemical parameter types in the battery cell.

[0111] In one example, the battery can be a single battery cell (or "single cell") or a battery module formed by connecting multiple battery cells in series and parallel. This application does not limit the specific battery in this regard.

[0112] In some alternative embodiments, if the battery is a single cell, the target electrochemical parameter value for that single cell is determined; in other alternative embodiments, if the battery comprises multiple cells, the target electrochemical parameter value for each cell may be determined.

[0113] In this embodiment, the target electrochemical parameters of a battery cell can also be referred to as the electrochemical health factors of the battery cell. The types of electrochemical health factors for a battery cell include, but are not limited to, the retention rate of the negative electrode solid-phase diffusion coefficient (EHF_Dsn), the charge transfer impedance growth rate (EHF_Rct), and the positive and negative electrode attenuation asymmetry index (EHF_Balance), etc. The electrochemical health factors selected from the retention rate of the negative electrode solid-phase diffusion coefficient (EHF_Dsn), the charge transfer impedance growth rate (EHF_Rct), and the positive and negative electrode attenuation asymmetry index (EHF_Balance) are called target electrochemical parameters.

[0114] In some optional embodiments, the target electrochemical parameter of the battery cell can be one of the following: the retention rate of the negative electrode solid-phase diffusion coefficient EHF_Dsn, the rate of increase of charge transfer impedance EHF_Rct, and the positive and negative electrode attenuation asymmetry index EHF_Balance, such as the retention rate of the negative electrode solid-phase diffusion coefficient EHF_Dsn; in other optional embodiments, the target electrochemical parameter of the battery cell can be two or all of the following: the retention rate of the negative electrode solid-phase diffusion coefficient EHF_Dsn, the rate of increase of charge transfer impedance EHF_Rct, and the positive and negative electrode attenuation asymmetry index EHF_Balance, and this application does not impose any particular limitation on this.

[0115] It should be noted that when a battery consists of multiple battery cells, the method for obtaining the target electrochemical parameter values ​​of each battery cell is the same. The following uses a single battery cell as an example to illustrate how the target electrochemical parameter values ​​of each battery cell are obtained.

[0116] It is understandable that the target electrochemical parameter values ​​of multiple battery cells can be obtained by summing up the target electrochemical parameter values ​​of each individual battery cell.

[0117] As an optional embodiment, such as Figure 3 As shown, the target electrochemical parameter values ​​for a single battery cell can be determined by following these steps: S102-1. Based on the initial state of charge in the battery nominal data and battery state data, determine the initial values ​​of multiple types of electrochemical parameters of the battery cell.

[0118] To initiate an electrochemical model for a battery cell, using the battery's nominal parameters (i.e., nominal battery data) and the cell's initial state of charge (SOC), provide physically meaningful initial values ​​for all types of electrochemical parameters to be estimated in the cell's electrochemical model.

[0119] Among them, the electrochemical parameters of the electrochemical model of a single battery cell can also be called "microscopic parameters".

[0120] For example, the electrochemical model is a high-fidelity mechanism model, which can be a three-dimensional electrochemical model, a P2D model, a pseudo-one-dimensional model, etc. This application does not make any special limitation on this. The following will take the electrochemical model of the battery cell as an example of the type of P2D model for illustrative purposes.

[0121] For example, all the electrochemical parameters to be estimated include, but are not limited to: negative electrode diffusion coefficient Dsn, positive electrode diffusion coefficient Dsp, negative electrode reaction rate kn, positive electrode reaction rate kp, lumped ohmic resistance Rct, negative electrode initial stoichiometric coefficient θn1, and positive electrode initial stoichiometric coefficient θp0.

[0122] Among them, the initial stoichiometric coefficients θn1 and θp0 of the negative electrode are thermodynamic parameters; the diffusion coefficients Dsn, Dsp, kn, kp, and Rct of the negative electrode and positive electrode are kinetic parameters.

[0123] S102-2. Using the measured voltage of a single battery cell as a reference, adjust the initial values ​​of multiple types of electrochemical parameters of the single battery cell until the error between the measured voltage of the single battery cell and the simulated voltage of the electrochemical model is minimized, thereby obtaining the identification results of multiple types of electrochemical parameters of the single battery cell.

[0124] The optimization process, specifically "adjusting the initial values ​​of multiple types of electrochemical parameters for a single battery cell until the error between the measured voltage and the simulated voltage of the electrochemical model is minimized," essentially involves a two-stage parameter identification process for the battery cell's electrochemical model. The first stage identifies the thermodynamic parameters among the multiple types of electrochemical parameters in the model, and the second stage identifies the kinetic parameters. It's important to note that the identified thermodynamic parameters remain fixed during the identification of the kinetic parameters. The electrochemical parameters of the battery cell's electrochemical model can be adjusted based on the measured voltage of the battery cell to minimize the error between the measured voltage and the simulated voltage. The optimization process employs a highly efficient algorithm, typically converging within seconds.

[0125] For example, the root mean square error (RMSE) can be used to evaluate whether the error between the measured voltage of a single cell and the simulated voltage of the electrochemical model is minimized. For instance, when the RMSE is minimized, the error between the measured voltage of the single cell and the simulated voltage of the electrochemical model is considered to be minimized.

[0126] S102-3. Determine the identification results of the target electrochemical parameter from the identification results of multiple types of electrochemical parameters of the battery cell.

[0127] It is understandable that the identification results of multiple types of electrochemical parameters for a single battery cell include the identification results of thermodynamic parameters and kinetic parameters. Specifically, the identification results of thermodynamic parameters represent the optimized thermodynamic parameters, and the identification results of kinetic parameters represent the optimized kinetic parameters.

[0128] In some optional embodiments, the identification result of one type of electrochemical parameter from multiple types of electrochemical parameters of a battery cell can be used as the identification result of the target electrochemical parameter. For example, the identification result Dsn1 of the negative electrode diffusion coefficient Dsn can be used as the identification result of the target electrochemical parameter, in which case the number of identification results of the target electrochemical parameter is 1. In other optional embodiments, the identification results of a first preset number of electrochemical parameters from multiple types of electrochemical parameters of a battery cell can be used as the identification results of the target electrochemical parameter. For example, it can be the identification results of three types of electrochemical parameters, exemplarily, the identification results Dsn1 of the negative electrode diffusion coefficient Dsn, Dsp1 of the positive electrode diffusion coefficient Dsp, and Rct1 of the lumped ohmic resistance Rct, in which case the number of identification results of the target electrochemical parameter is 3. It should be noted that the embodiments of this application do not impose a special limitation on the number of identification results of the target electrochemical parameter.

[0129] S102-4. Based on the identification results of the target electrochemical parameters and the preset values ​​of the target electrochemical parameters, determine the target electrochemical parameter values ​​of the battery cells.

[0130] As an optional embodiment, the target electrochemical parameter values ​​of a single battery cell can be determined according to the following steps: S102-4A, Obtain preset values ​​for target electrochemical parameters.

[0131] In some optional embodiments, the identification result of one type of electrochemical parameter among the identification results of multiple types of electrochemical parameters of a single cell can be used as the identification result of the target electrochemical parameter. For example, when the identification result Dsn1 of the negative electrode diffusion coefficient Dsn is selected as the identification result of the target electrochemical parameter, the preset value Dsn_fresh of the negative electrode diffusion coefficient Dsn needs to be obtained accordingly.

[0132] In some alternative embodiments, the identification results of a first preset number of electrochemical parameters from the identification results of multiple types of electrochemical parameters of a single cell can be used as the identification results of the target electrochemical parameter. For example, it can be the identification results of three types of electrochemical parameters. For example, when the identification results of the negative electrode diffusion coefficient Dsn (Dsn1), the positive electrode diffusion coefficient Dsp (Dsp1), and the lumped ohmic resistance Rct (Rct1) are selected as the identification results of the target electrochemical parameter, the preset values ​​of the negative electrode diffusion coefficient Dsn (Dsn_fresh), the positive electrode diffusion coefficient Dsp (Dsp_fresh), and the lumped ohmic resistance Rct (Rct_fresh) need to be obtained respectively.

[0133] S102-4B. Based on the identification results of the target electrochemical parameters and the preset values ​​of the target electrochemical parameters, determine the target electrochemical parameter values ​​of the battery cells.

[0134] For example, if the identification results of the target electrochemical parameters are the identification results of the negative electrode diffusion coefficient Dsn1, the positive electrode diffusion coefficient Dsp1, and the lumped ohmic resistance Rct1, then the target electrochemical parameter values ​​of the battery cell can include the value of the negative electrode diffusion capacity retention rate EHF_Dsn, the value of the charge transfer impedance growth rate EHF_Rct, and the value of the positive and negative electrode attenuation asymmetry index EHF_Balance.

[0135] For example, the value of the negative electrode diffusion capacity retention rate EHF_Dsn can be determined according to the following formula (1): (1) For example, the value of the charge transfer impedance growth rate EHF_Rct can be determined according to the following formula (2): (2) For example, the value of the positive and negative electrode attenuation asymmetry index EHF_Balance can be determined according to the following formula (3): (3) Understandably, in this embodiment, initial values ​​are first assigned to multiple types of electrochemical parameters of the battery cell. Then, based on the measured voltage of the battery cell, the initial values ​​of the multiple types of electrochemical parameters of the battery cell are adjusted until the error between the measured voltage of the battery cell and the simulated voltage of the electrochemical model is minimized. This can improve the convergence speed of identification and make the identification results of multiple types of electrochemical parameters of the battery cell more consistent with the actual situation. In other words, the actual values ​​of multiple types of electrochemical parameters of the battery cell are obtained. Finally, the actual value of the target electrochemical parameter of the battery cell is selected from the actual values ​​of multiple types of electrochemical parameters of the battery cell. Based on the actual value of the target electrochemical parameter of the battery cell (i.e., the identification result) and the preset value of the target electrochemical parameter of the battery cell, the target electrochemical parameter value of the battery cell is determined. This can ensure the accuracy of the target electrochemical parameter value of the battery cell and further improve the accuracy of the determination of battery anomalies.

[0136] In some alternative embodiments, such as Figure 4 As shown, after obtaining the target electrochemical parameter values ​​of the battery cell, the battery health status determination method provided in this application embodiment may further include the following steps: S102-5. Obtain the loop count and health status estimate under the timestamp.

[0137] Here, the health status estimate refers to the estimated state of health (SOH).

[0138] For example, the full charge-full discharge capacity method can be used to determine the estimated health status of a single battery cell; the internal resistance-DC impedance method can also be used to determine the estimated health status of a single battery cell, etc., and the embodiments of this application do not particularly limit this.

[0139] S102-6. Based on timestamps, establish the correlation between timestamps and the target electrochemical parameter values, cycle counts, and estimated health status values ​​of individual battery cells.

[0140] S102-7. Store the correlation between the timestamp and the target electrochemical parameter values, cycle number and health status estimate of the battery cell into a preset time series database.

[0141] It is understood that, in this embodiment, a correlation is established between the timestamp and the target electrochemical parameter values, cycle count, and health estimate of a battery cell, based on the timestamp. This allows multiple battery cells to obtain their own unique target electrochemical parameter values, cycle count, and health estimate at the same timestamp, making it convenient to observe multiple battery cells at the same time and to identify abnormal battery cells more quickly.

[0142] Understandably, the target data package includes multiple timestamps. For a single battery cell, based on these multiple timestamps, the step of establishing the correlation between the timestamps and the target electrochemical parameter values, cycle count, and estimated health status of the battery cell is repeatedly executed. That is, steps S102-5 to S102-7 are repeatedly executed to obtain multiple sets of correlations between the timestamps and the target electrochemical parameter values, cycle count, and estimated health status of the battery cell. Based on the correlations between the multiple sets of timestamps and the target electrochemical parameter values, cycle count, and estimated health status of the battery cell, a preset time-series database is constructed. In other words, the correlations between the multiple sets of timestamps and the target electrochemical parameter values, cycle count, and estimated health status of the battery cell are stored in the preset time-series database. In this way, the target electrochemical parameter values, cycle count, and estimated health status of a battery cell at each timestamp can be obtained.

[0143] Understandably, each battery cell can repeatedly execute steps S102-5 to S102-7 based on multiple timestamps, thereby obtaining the target electrochemical parameter values, cycle count, and health status estimate for each battery cell at each timestamp.

[0144] It is understood that in this embodiment, each battery cell can obtain multiple target electrochemical parameter values, multiple cycle counts, and multiple health status estimates, and each target electrochemical parameter value corresponds to a cycle count and a health status estimate, which is beneficial for observing the health status of each battery cell at the same time.

[0145] S103. Based on the multiple sets of target electrochemical parameter values ​​of multiple battery cells obtained through multiple cycles, determine the inconsistency of the target electrochemical parameters of multiple battery cells.

[0146] In some optional embodiments, multiple sets of target electrochemical parameter values ​​for multiple battery cells obtained through multiple cycles can be obtained by the following method: based on multiple timestamps, with the timestamp as a variable, the steps of determining the target electrochemical parameter values ​​of multiple battery cells in the battery that minimize the error between the measured voltage corresponding to the timestamp and the simulated voltage of the electrochemical model are executed cyclically, so as to obtain multiple sets of target electrochemical parameter values ​​for multiple battery cells obtained through multiple cycles.

[0147] It is understood that in this embodiment, one timestamp can correspond to a set of electrochemical parameter values ​​for multiple battery cells, and multiple timestamps can correspond to multiple sets of target electrochemical parameter values ​​for multiple battery cells. This can provide rich data for subsequently determining the inconsistency of target electrochemical parameters for multiple battery cells, reduce the occurrence of accidental determination of battery anomalies, and further improve the accuracy of battery anomaly detection.

[0148] In some alternative embodiments, the inconsistency of target electrochemical parameters among multiple battery cells can be determined by the following steps: S103-1. Based on a preset time series database, obtain multiple sets of target electrochemical parameter values ​​for multiple battery cells.

[0149] It is understandable that the preset time series database stores multiple sets of target electrochemical parameter values ​​for each battery. Therefore, multiple sets of target electrochemical parameter values ​​for each battery cell can be obtained from the preset time series database, resulting in multiple sets of electrochemical values ​​for multiple battery cells.

[0150] It should be noted that multiple sets of target electrochemical parameter values ​​for a single battery cell refer to a single battery cell having multiple parameter values ​​for one type of target electrochemical parameter.

[0151] S103-2. Based on multiple sets of target electrochemical parameter values ​​for multiple battery cells, determine the decay trajectory of the target electrochemical parameters for multiple battery cells and the standard deviation of the target electrochemical parameters for multiple battery cells.

[0152] It should be noted that since the methods for determining the decay trajectory and standard deviation of the target electrochemical parameters for each battery cell are the same, this explanation will take the method for determining the decay trajectory and standard deviation of the target electrochemical parameters for a single battery cell as an example.

[0153] It is understandable that by summing up the decay trajectory of the target electrochemical parameters of each battery cell, the decay trajectory of the target electrochemical parameters of multiple battery cells can be obtained. By summing up the standard deviation of the target electrochemical parameters of each battery cell, the standard deviation of the target electrochemical parameters of multiple battery cells can be obtained.

[0154] For example, the decay trajectory of the target electrochemical parameters of a single battery cell is obtained as follows: a scatter plot of the target electrochemical parameters of the battery cell is plotted with the timestamp as the horizontal axis and multiple sets of target electrochemical parameter values ​​of the battery cell as the vertical axis; and the scatter plot of the target electrochemical parameters of the battery cell is smoothed and fitted to obtain the decay trajectory of the target electrochemical parameters of the battery cell.

[0155] In some optional embodiments, the decay trajectory of each type of target electrochemical parameter in a battery cell can be plotted. Since the plotting process for the decay trajectory of each type of target electrochemical parameter in a battery cell is the same, this example uses the types of target electrochemical parameters in a battery cell, namely, negative electrode diffusion capacity retention rate EHF_Dsn, charge transfer impedance growth rate EHF_Rct, and positive and negative electrode decay asymmetry index EHF_Balance. Using the above method, the decay trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn, the decay trajectory of the charge transfer impedance growth rate EHF_Rct, and the decay trajectory of the positive and negative electrode decay asymmetry index EHF_Balance can be plotted.

[0156] For example, the standard deviation of the target electrochemical parameter of a single battery cell is obtained as follows: the standard deviation of multiple sets of target electrochemical parameters of the battery cell is calculated to obtain the standard deviation of the target electrochemical parameter of the battery cell.

[0157] In some optional embodiments, the standard deviation of each type of target electrochemical parameter in a battery cell can be determined. Since the process of determining the standard deviation of each type of target electrochemical parameter in a battery cell is the same, this example uses the following: the types of target electrochemical parameters in a battery cell are negative electrode diffusion capacity retention rate (EHF_Dsn), charge transfer impedance growth rate (EHF_Rct), and positive and negative electrode attenuation asymmetry index (EHF_Balance). The standard deviation is calculated by querying the value of the negative electrode diffusion capacity retention rate (EHF_Dsn) at each time point, and then calculating the standard deviation of the negative electrode diffusion capacity retention rate (EHF_Dsn) at each time point. This yields the standard deviation of the negative electrode diffusion capacity retention rate (EHF_Dsn) of the battery cell. F_Dsnδ1 represents: querying the charge transfer impedance growth rate EHF_Rct of a single battery cell at various timestamps, and then calculating the standard deviation of the charge transfer impedance growth rate EHF_Rct at various timestamps to obtain the standard deviation of the charge transfer impedance growth rate EHF_Rct of the single battery cell, denoted as EHF_Rctδ2; querying the positive and negative electrode attenuation asymmetry index EHF_Balance of a single battery cell at various timestamps, and then calculating the standard deviation of the positive and negative electrode attenuation asymmetry index EHF_Balance at various timestamps to obtain the standard deviation of the positive and negative electrode attenuation asymmetry index EHF_Balance of the single battery cell, denoted as EHF_Balanceδ3.

[0158] It is understood that, in this embodiment, for each battery cell, by using multiple sets of target electrochemical parameter values ​​for each battery cell to plot the decay trajectory of each battery cell, the difference in the overall aging trajectory pattern of each battery cell can be measured; in addition, by calculating the standard deviation of multiple sets of target electrochemical parameter values ​​for each battery cell, the dispersion of the target electrochemical parameter values ​​of each battery cell at the current moment can be monitored.

[0159] S103-3. Based on the decay trajectory of the target electrochemical parameters of multiple battery cells and the standard deviation of the target electrochemical parameters of multiple battery cells, determine the inconsistency of the target electrochemical parameters of multiple battery cells.

[0160] It should be noted that since the process for determining the inconsistency of the target electrochemical parameters of each battery cell is the same, this section will use the process for determining the inconsistency of the target electrochemical parameters of a single battery cell as an example for illustration.

[0161] It is understandable that by summing up the inconsistencies in the target electrochemical parameters of each individual battery cell, the inconsistencies in the target electrochemical parameters of multiple battery cells can be obtained.

[0162] Taking a single battery cell as an example, in some optional embodiments, the inconsistency of the target electrochemical parameters of the battery cell can be determined by the following steps: S103-3A. Based on the decay trajectory of the target electrochemical parameters of the battery cell and the preset reference trajectory, determine the target trajectory difference of the target electrochemical parameters of the battery cell.

[0163] In some optional embodiments, the target trajectory difference of the target electrochemical parameters of the battery cell can be determined according to the following steps: S103-3A1: Based on the distance between the decay trajectory of the target electrochemical parameters of the battery cell and the preset reference trajectory, multiple trajectory difference values ​​of the target electrochemical parameters of the battery cell are obtained.

[0164] In this embodiment, the preset reference trajectory refers to the preset reference trajectory of the battery. The preset reference trajectory of the battery is related to the type of target electrochemical parameter of the battery cell. For example, if the target electrochemical parameter of the battery cell is the negative electrode diffusion capacity retention rate EHF_Dsn, then the preset reference trajectory of the battery is the preset reference trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn. In some optional embodiments, the user or R&D personnel can set the preset reference trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn in advance. In other optional embodiments, the decay trajectories of the negative electrode diffusion capacity retention rate EHF_Dsn of each battery cell can be superimposed to obtain the decay trajectory of the battery's negative electrode diffusion capacity retention rate EHF_Dsn. Then, the median or mean of the decay trajectory of the battery's negative electrode diffusion capacity retention rate EHF_Dsn is processed to obtain the preset reference trajectory of the battery's negative electrode diffusion capacity retention rate EHF_Dsn.

[0165] For example, a preset reference trajectory for the negative electrode diffusion capacity retention rate EHF_Dsn of the battery is determined based on the decay trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn of each individual battery cell. In some optional embodiments, the median trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn can be obtained based on the decay trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn of each individual battery cell, and the obtained median trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn can be used as a preset reference trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn of the battery. In other optional embodiments, the average trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn can be obtained based on the decay trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn of each individual battery cell, and the obtained average trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn can be used as a preset reference trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn of the battery, etc. This application does not make any special limitation on this. The following description will take the example of obtaining the median trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn as the preset reference trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn of the battery.

[0166] For example, a preset reference trajectory for the charge transfer impedance growth rate EHF_Rct of the battery is determined based on the decay trajectory of the charge transfer impedance growth rate EHF_Rct of each individual battery cell. In some optional embodiments, the median trajectory of the charge transfer impedance growth rate EHF_Rct can be obtained based on the decay trajectory of the charge transfer impedance growth rate EHF_Rct of each individual battery cell, and the obtained median trajectory of the charge transfer impedance growth rate EHF_Rct is used as the preset reference trajectory of the charge transfer impedance growth rate EHF_Rct of the battery; in other optional embodiments, the average trajectory of the charge transfer impedance growth rate EHF_Rct can be obtained based on the decay trajectory of the charge transfer impedance growth rate EHF_Rct of each individual battery cell, and the obtained average trajectory of the charge transfer impedance growth rate EHF_Rct is used as the preset reference trajectory of the charge transfer impedance growth rate EHF_Rct of the battery, etc. This application does not particularly limit this, and the following explanation will take the example of obtaining the median trajectory of the charge transfer impedance growth rate EHF_Rct as the preset reference trajectory of the charge transfer impedance growth rate EHF_Rct of the battery.

[0167] For example, a preset reference trajectory for the battery's positive and negative electrode attenuation asymmetry index EHF_Balance is determined based on the attenuation trajectory of the positive and negative electrode attenuation asymmetry index EHF_Balance of each individual battery cell. In some optional embodiments, the median trajectory of the positive and negative electrode attenuation asymmetry index EHF_Balance can be obtained based on the parameter trajectory of the positive and negative electrode attenuation asymmetry index EHF_Balance of each individual battery cell, and the obtained median trajectory of the positive and negative electrode attenuation asymmetry index EHF_Balance can be used as the preset reference trajectory of the positive and negative electrode attenuation asymmetry index EHF_Balance of the battery. In other optional embodiments, the average trajectory of the positive and negative electrode attenuation asymmetry index EHF_Balance can be obtained based on the attenuation trajectory of the positive and negative electrode attenuation asymmetry index EHF_Balance of each individual battery cell, and the obtained average trajectory of the positive and negative electrode attenuation asymmetry index EHF_Balance can be used as the preset reference trajectory of the positive and negative electrode attenuation asymmetry index EHF_Balance of the battery, etc. This application does not make any special limitations on this. The following description will take the example of obtaining the median trajectory of the positive and negative electrode attenuation asymmetry index EHF_Balance as the preset reference trajectory of the positive and negative electrode attenuation asymmetry index EHF_Balance of the battery.

[0168] The following details the determination of multiple trajectory differences in the target electrochemical parameters of individual battery cells.

[0169] For example, taking the target electrochemical parameter of a battery cell as the negative electrode diffusion capacity retention rate EHF_Dsn as an example: according to the timestamp, calculate the distance between the decay trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn of the battery cell at each timestamp and the median trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn, and obtain the multiple trajectory difference DTW1 value of the negative electrode diffusion capacity retention rate EHF_Dsn of the battery cell.

[0170] For example, taking the target electrochemical parameter of a battery cell as the charge transfer impedance growth rate EHF_Rct as an example: according to the timestamp, calculate the distance between the decay trajectory of the charge transfer impedance growth rate EHF_Rct of the battery cell at each timestamp and the median trajectory of the charge transfer impedance growth rate EHF_Rct, and obtain the DTW2 value of the multiple trajectory difference of the charge transfer impedance growth rate EHF_Rct of the battery cell.

[0171] For example, taking the target electrochemical parameter of a battery cell as the positive and negative electrode attenuation asymmetry index EHF_Balance as an example: according to the timestamp, calculate the distance between the attenuation trajectory of the positive and negative electrode attenuation asymmetry index EHF_Balance of the battery cell at each timestamp and the median trajectory of the positive and negative electrode attenuation asymmetry index EHF_Balance, and obtain the multiple trajectory difference DTW3 value of the positive and negative electrode attenuation asymmetry index EHF_Balance of the battery cell.

[0172] S103-3A2: Average the multiple trajectory differences of the target electrochemical parameters of the battery cell to obtain the target trajectory difference degree of the target electrochemical parameters of the battery cell.

[0173] For example, the average value of multiple trajectory difference degrees DTW1 values ​​of the negative electrode diffusion capability retention rate EHF_Dsn of the battery cell is processed to obtain the target trajectory difference degree Avg(DTW_Dsn) of the negative electrode diffusion capability retention rate EHF_Dsn of the battery cell.

[0174] For example, the average value of multiple trajectory difference DTW2 values ​​of the charge transfer impedance growth rate EHF_Rct of a battery cell is processed to obtain the target trajectory difference Avg(DTW_Rct) of the charge transfer impedance growth rate EHF_Rct of the battery cell.

[0175] For example, the average value of multiple trajectory difference degrees DTW3 values ​​of the positive and negative electrode attenuation asymmetry index EHF_Balance of a battery cell is processed to obtain the target trajectory difference degree Avg (DTW_Balance) of the positive and negative electrode attenuation asymmetry index EHF_Balance of a battery cell.

[0176] S103-3B, Normalize the standard deviation of the target electrochemical parameters of the battery cell to obtain the target standard deviation δ_normal(i) of the target electrochemical parameters of the battery cell.

[0177] Where i represents the type of target electrochemical parameters for a single battery cell.

[0178] The target electrochemical parameters for individual battery cells continue to be categorized as follows: negative electrode diffusion capacity retention rate (EHF_Dsn), charge transfer impedance growth rate (EHF_Rct), and positive and negative electrode degradation asymmetry index (EHF_Balance). The standard deviation of the negative electrode diffusion capacity retention rate of the battery cell, EHF_Dsnδ1, is normalized to obtain the target standard deviation of the negative electrode diffusion capacity retention rate, δ_normal(EHF_Dsn); the standard deviation of the charge transfer impedance growth rate of the battery cell, EHF_Rctδ2, is normalized to obtain the target standard deviation of the charge transfer impedance growth rate, δ_normal(Rct); and the standard deviation of the positive and negative electrode attenuation asymmetry index of the battery cell, EHF_Balanceδ3, is normalized to obtain the target standard deviation of the positive and negative electrode attenuation asymmetry index, δ_normal(Balance).

[0179] It is understood that, in this embodiment, normalizing the standard deviation of the target electrochemical parameters of the battery cell can obtain dimensionless features characterizing the relative fluctuations of the target electrochemical parameters. This not only eliminates the difference in magnitude but also achieves unified evaluation of multiple features, thereby further improving the accuracy of battery anomaly detection.

[0180] Understandably, in this embodiment, for each battery cell: by calculating the distance between the decay trajectory of the target electrochemical parameters of the battery cell and the preset reference trajectory, multiple trajectory difference values ​​of the battery cell can be obtained, which can eliminate normal aging factors and accurately characterize the abnormal degradation degree and consistency deviation of the battery cell; then, the multiple trajectory difference values ​​of the battery cell are averaged to smooth the random disturbance of the battery cell, and obtain the system-level trajectory difference degree (i.e., target trajectory difference degree) that characterizes the overall deviation of the battery cell from the standard aging degree. This realizes the organic unity of battery cell anomaly identification and battery consistency assessment, which can not only improve the sensitivity and robustness of early anomaly detection of the battery, but also provide a reliable basis for battery safety early warning and health management.

[0181] S103-3C: The target trajectory difference of the target electrochemical parameters of the battery cell and the target standard deviation of the target electrochemical parameters of the battery cell are weighted and calculated to obtain the inconsistency of the target electrochemical parameters of the battery cell.

[0182] In some optional embodiments, taking the case where there is only one target electrochemical parameter, such as the negative electrode diffusion capacity retention rate EHF_Dsn, the target standard deviation δ_normal (EHF_Dsn) of the negative electrode diffusion capacity retention rate EHF_Dsn of the battery cell and the target trajectory difference Avg (DTW_Dsn) of the negative electrode diffusion capacity retention rate EHF_Dsn of the battery cell can be weighted to determine the inconsistency UI1 of the negative electrode diffusion capacity retention rate EHF_Dsn of the battery cell.

[0183] For example, the inconsistency UI1 of the negative electrode diffusion capacity retention rate EHF_Dsn of a single cell can be determined according to the following formula (4): (4) Where α1 and β1 represent weighting coefficients, and their sum is 1; α1 can be adjusted according to the sensitivity of δ_normal (EHF_Dsn); β1 can be adjusted according to the sensitivity of Avg (DTW_Dsn); for example, α1=0.4, β1=0.6.

[0184] In some alternative embodiments, multiple target electrochemical parameters are used as an example. For instance, the target electrochemical parameters may include the negative electrode diffusion capacity retention rate EHF_Dsn, the charge transfer impedance growth rate EHF_Rct, and the positive and negative electrode attenuation asymmetry index EHF_Balance. The inconsistencies of each of the multiple target electrochemical parameters of a single cell are calculated separately.

[0185] Among them, the target standard deviation δ_normal(EHF_Dsn) of the negative electrode diffusion capacity retention rate EHF_Dsn of the battery cell and the target trajectory difference DTW(Dsn) of the negative electrode diffusion capacity retention rate EHF_Dsn of the battery cell are weighted and the inconsistency UI1 of the negative electrode diffusion capacity retention rate EHF_Dsn of the battery cell can be determined by referring to the above formula (4), which will not be repeated here. The other two are illustrated here.

[0186] The inconsistency UI2 of the charge transfer impedance growth rate EHF_Rct of a battery cell is determined by weighting the target standard deviation δ_normal (EHF_Rct) and the target trajectory difference DTW (Rct) of the charge transfer impedance growth rate EHF_Rct of the battery cell.

[0187] For example, the inconsistency UI2 of the charge transfer impedance growth rate EHF_Rct of a single cell can be determined according to the following formula (5): (5) Where α2 and β2 represent weighting coefficients, and their sum is 1; α1 can be adjusted according to the sensitivity of δ_normal (EHF_Rct); β2 can be adjusted according to the sensitivity of Avg (DTW_Rct); for example, α2=0.4, β2=0.6.

[0188] The target standard deviation δ_normal (EHF_Balance) of the positive and negative electrode attenuation asymmetry index EHF_Balance of the battery cell and the target trajectory difference DTW (Banlance) of the positive and negative electrode attenuation asymmetry index EHF_Balance of the battery cell are weighted and calculated to determine the inconsistency UI3 of the positive and negative electrode attenuation asymmetry index EHF_Balance of the battery cell.

[0189] For example, the inconsistency UI3 of the charge transfer impedance growth rate EHF_Rct of a single cell can be determined according to the following formula (6): (6) Where α3 and β3 represent weighting coefficients, and their sum is 1; α3 can be adjusted according to the sensitivity of δ_normal (EHF_Balance); β3 can be adjusted according to the sensitivity of Avg (DTW_Balance); for example, α3=0.4, β3=0.6.

[0190] Understandably, in this embodiment, based on the decay trajectory of the target electrochemical parameters of each battery cell and a preset reference trajectory, the target trajectory difference of the target electrochemical parameters of each battery cell is determined, thereby obtaining the inconsistency in the aging rhythm of each battery cell. By fusing the target standard deviation of the target electrochemical parameters of each battery cell and the target trajectory difference of the target electrochemical parameters of each battery cell to determine the inconsistency of the target electrochemical parameters of each battery cell, it is possible not only to focus on the dispersion of the target electrochemical parameter values ​​of each battery cell at the current moment, but also to measure the differences in the overall aging trajectory pattern of each battery cell, thereby capturing complex inconsistency patterns such as "rapid decay followed by slow decay", further improving the accuracy of anomaly detection for batteries.

[0191] Understandably, in this embodiment, after obtaining multiple sets of target electrochemical parameter values ​​for each battery cell, the decay trajectory corresponding to the target electrochemical parameters of each battery cell can be plotted based on these values. This is essentially a decay trajectory map of each battery cell throughout its entire lifecycle, which is beneficial for analyzing the decay status of each battery cell. Furthermore, the standard deviation of the target electrochemical parameters for each battery cell can be determined based on these values, which helps confirm the internal stability of each battery cell and the possibility of early failures. By analyzing the decay trajectories and standard deviations of the target electrochemical parameters of multiple battery cells, the inconsistencies in the target electrochemical parameters of multiple batteries can be identified. Compared to related technologies that use voltage anomalies to characterize battery anomalies, this approach can significantly advance the warning time of battery anomalies. That is, battery anomalies can be detected early, before obvious voltage anomalies appear, facilitating "prevention" and reducing the occurrence of safety accidents such as thermal runaway. This further improves both the accuracy and timeliness of battery anomaly detection.

[0192] S104. Determine the health status of multiple battery cells based on the inconsistency of their target electrochemical parameters.

[0193] It should be noted that the inconsistency of the target electrochemical parameters can reflect the health status of the battery cells. After obtaining the inconsistency of the target electrochemical parameters of each battery cell, the health status of each battery cell can be determined.

[0194] S105. If the health status is greater than the preset threshold, determine that the battery is abnormal.

[0195] It should be noted that the health status of a single battery cell can refer to the inconsistency of its target electrochemical parameters. The following example illustrates this by determining that a battery is abnormal when the inconsistency of its target electrochemical parameters exceeds a preset threshold.

[0196] In some alternative embodiments, after determining that the battery is abnormal, the abnormal result of the battery can be sent to the user equipment.

[0197] In some optional embodiments, if the type of the target electrochemical parameter is one, and the inconsistency of the target electrochemical parameter of the corresponding battery cell is also one, then if the inconsistency of the target electrochemical parameter of at least one battery cell in the multiple battery cells of the battery is greater than a preset threshold, the battery is determined to be abnormal.

[0198] For example, taking the type of target electrochemical parameter as negative electrode diffusion capacity retention rate EHF_Dsn as an example, when there is an inconsistency UI1 in the negative electrode diffusion capacity retention rate EHF_Dsn of at least one cell in the battery is greater than a preset threshold, the battery is determined to be abnormal.

[0199] In some alternative embodiments, when the target electrochemical parameters include multiple types, an anomaly is determined to exist in the battery if the inconsistency of any type of target electrochemical parameter in any of the multiple battery cells meets a preset condition.

[0200] The condition that the inconsistency of any type of target electrochemical parameter of any battery cell satisfies the preset condition means that the inconsistency of any type of target electrochemical parameter of any battery cell is greater than the corresponding preset threshold.

[0201] For example, when the target electrochemical parameters are negative electrode diffusion capacity retention rate EHF_Dsn, charge transfer impedance growth rate EHF_Rct, and positive and negative electrode decay asymmetry index EHF_Balance, if any one of the inconsistencies in negative electrode diffusion capacity retention rate EHF_Dsn (UI1), charge transfer impedance growth rate EHF_Rct (UI2), and charge transfer impedance growth rate EHF_Rct (UI3) of any one of the multiple battery cells exceeds the corresponding preset threshold, it can be determined that the battery has an abnormality.

[0202] It should be noted that the preset thresholds corresponding to UI1, UI2, and UI3 can be the same or different, and this application does not impose any special restrictions on this.

[0203] It is understood that, in the embodiments of this application, when the inconsistency of any type of target electrochemical parameter of any of the multiple battery cells meets the preset conditions, it is determined that there is an anomaly in the battery, which can ensure timely detection of battery anomalies and further improve the accuracy of battery anomaly detection.

[0204] It is understood that, in this embodiment, after acquiring battery state data during vehicle operation, since the battery state data includes at least multiple timestamps corresponding to measured voltages, the target electrochemical parameter values ​​of multiple battery cells in the battery that minimize the error between the measured voltage and the simulated voltage of the electrochemical model can be determined based on the battery state data, battery nominal data, and electrochemical model. This not only improves the convergence speed and quickly determines the target electrochemical parameter values ​​of multiple battery cells, but also makes the determined target electrochemical parameter values ​​of multiple battery cells more closely reflect actual conditions. This is because the target electrochemical parameters can directly and quantitatively characterize the health of specific materials or interface properties within the battery cell. Therefore, by determining the inconsistency of target electrochemical parameters among multiple battery cells obtained through multiple cycles, the actual situation of the battery cells can be reflected at a more microscopic level. Based on the inconsistency of target electrochemical parameters among multiple battery cells, the health status of multiple battery cells can be determined. If the health status is greater than a preset threshold, the battery is determined to be abnormal. Compared with the method of diagnosing whether the battery is abnormal by analyzing the overall voltage of the battery pack, this method can draw conclusions about whether the battery is abnormal before the overall voltage of the battery pack becomes obviously abnormal. It can detect battery abnormalities in a timely and accurate manner, thereby improving the accuracy and timeliness of battery abnormality detection.

[0205] In another embodiment of this application, a method for predicting battery life is provided, such as... Figure 6 As shown, the battery life prediction method includes the following steps: S301. After a battery malfunctions, identify the abnormal battery cell that is malfunctioning.

[0206] In some embodiments, a battery cell whose trajectory difference value of the target electrochemical parameter in the battery exceeds a preset value is identified as an abnormal battery cell; and / or, the battery cell with the largest standard deviation of the target electrochemical parameter in the battery is identified as the abnormal battery cell.

[0207] It should be noted that the process for determining whether each battery cell is abnormal is the same. Here, we take a single battery cell as an example to illustrate the method for determining whether a battery cell is abnormal.

[0208] In some alternative embodiments, taking a single target electrochemical parameter as an example, such as the negative electrode diffusion capacity retention rate EHF_Dsn: For example, the trajectory difference value of the target electrochemical parameter can be used to determine whether it is an abnormal battery cell: the multiple trajectory difference values ​​DTW1 of the negative electrode diffusion capacity retention rate EHF_Dsn of the battery cell are compared with a preset difference value respectively. When at least one trajectory difference value DTW1 exceeds the preset difference value, it is determined to be an abnormal battery cell.

[0209] For example, the standard deviation of the target electrochemical parameter can be used to determine whether a cell is abnormal: the standard deviation of the negative electrode diffusion capacity retention rate EHF_Dsn of each cell is compared with each other, and the cell with the largest standard deviation of the negative electrode diffusion capacity retention rate EHF_Dsn is identified as an abnormal cell.

[0210] It is understandable that the battery cell corresponding to the trajectory difference degree DTW1 of the negative electrode diffusion capacity retention rate EHF_Dsn exceeding the preset difference degree and the battery cell with the largest standard deviation of the negative electrode diffusion capacity retention rate EHF_Dsn may be the same battery cell or two different battery cells. This application does not make any special limitation in this regard. The following will take the same battery cell as an example for illustrative explanation.

[0211] Understandably, in this embodiment, the trajectory difference of the target parameters of a battery cell can reflect the "deviation" of the battery cell's trend, and the standard deviation of the target electrochemical parameters of a battery cell can reflect the "fluctuation instability" of the battery cell. Therefore, the abnormal battery cells identified based on the trajectory difference and / or the standard deviation of the target electrochemical parameters of the battery cells are more accurate, improving the precision of locating abnormal battery cells in the battery. Whether the battery cell with a trajectory difference exceeding a preset difference is identified as an abnormal battery cell, or the battery cell with the largest standard deviation of the target electrochemical parameters is identified as an abnormal battery cell, accurate location of abnormal battery cells can be achieved, thereby improving the precision of locating abnormal battery cells in the battery.

[0212] S302. Match the target electrochemical parameter values ​​of the abnormal battery cell with the failure boundary library to determine the first degradation mode of the abnormal battery cell.

[0213] This can be achieved by matching the value of a single target electrochemical parameter of the abnormal battery cell with a failure boundary database when the target electrochemical parameter type is singular, thus obtaining a degradation mode that matches the abnormal battery cell. Alternatively, when there are multiple target electrochemical parameters of the abnormal battery cell, the values ​​of all target electrochemical parameters are matched with a failure boundary database to obtain a degradation mode that matches the abnormal battery cell. For example, when the negative electrode diffusion capacity retention rate EHF_Dsn of the abnormal battery cell is <0.5 and the EHF_Rct of the abnormal battery cell is >2.0, it is considered a degradation mode for the abnormal battery cell. This means that when the negative electrode diffusion capacity retention rate EHF_Dsn of the abnormal battery cell is <0.5 and the EHF_Rct of the abnormal battery cell is >2.0, the abnormal battery cell is considered to have reached the end of its life, and its capacity may decay to 80%.

[0214] S303. Based on the first degradation mode, extrapolate the decay trajectory of the target electrochemical parameters of the abnormal battery cell to determine the baseline degradation trajectory of the abnormal battery cell.

[0215] In one optional embodiment, after obtaining the first degradation mode of the abnormal battery cell, a first degradation model is matched for the abnormal battery cell, and the decay trajectory of the target electrochemical parameters of the abnormal battery cell is extrapolated using the first degradation model until it is the same as the first degradation mode of the abnormal battery cell, at which point the extrapolation stops and the baseline degradation trajectory of the abnormal battery cell is obtained.

[0216] In this embodiment, the first degradation model can be an exponential decay degradation model, a power-law degradation model, etc., and this application does not make any special limitation on it.

[0217] For example, when the target electrochemical parameters are the negative electrode diffusion capacity retention rate EHF_Dsn and the charge transfer impedance growth rate EHF_Rct, the first degradation mode of the abnormal battery cell is EHF_Dsn < 0.5 and EHF_Rct > 2.0. Then, the decay trajectory of the negative electrode diffusion capacity retention rate EHF_Dsn and the decay trajectory of the charge transfer impedance growth rate EHF_Rct of the abnormal battery cell are extrapolated using the first degradation model until the negative electrode diffusion capacity retention rate EHF_Dsn < 0.5 and EHF_Rct > 2.0 of the abnormal battery cell, at which point the extrapolation stops, and the baseline degradation trajectory of the abnormal battery cell is obtained.

[0218] S304. Use a probabilistic prediction model to predict the lifetime of the baseline degradation trajectory of the abnormal battery cell and obtain the probability distribution of the remaining lifetime of the abnormal battery cell.

[0219] For example, the probability prediction model can be a Monte Carlo probability prediction model.

[0220] For example, the Monte Carlo probabilistic prediction model predicts the lifetime of a faulty battery cell based on a baseline degradation trajectory following these steps: S304-1, Set future operating conditions, parameter uncertainties of the first degradation model, and random distribution of measurement noise.

[0221] S304-2. Perform a preset number of simulations. In each simulation, extrapolate the baseline degradation trajectory of the abnormal battery cell based on the results of random sampling until the parameter value of the target electrochemical parameter of the abnormal battery cell touches its matching failure boundary (i.e. degradation mode). Record the "remaining number of cycles" under this simulation.

[0222] The result of random sampling represents the result formed by randomly sampling the parameter uncertainty and measurement noise of the future operating conditions, the first degradation model, and respectively.

[0223] S304-3. Summarize the results of "remaining cycle count" obtained from all simulations to obtain the probability distribution of the remaining lifespan of the abnormal battery cell.

[0224] For example, the probability distribution of the remaining lifespan of an abnormal battery cell can be represented in the form of a normal distribution.

[0225] S305. Based on the probability distribution of the remaining service life of abnormal battery cells, determine the battery life prediction result.

[0226] In some optional embodiments, the remaining lifespan of abnormal battery cells is determined based on the probability distribution of the remaining lifespan of the abnormal battery cells, and the remaining lifespan of the abnormal battery cells is determined as the battery life prediction result.

[0227] For example, after obtaining the probability distribution of the remaining lifespan of the abnormal battery cell, the remaining lifespan of the abnormal battery cell can be obtained by outputting according to the preset index.

[0228] For example, the preset indicators may include, but are not limited to, expected RUL, median RUL, and RUL range under different confidence intervals (such as 90%), etc. The expected RUL can be determined as the remaining service life of the abnormal battery cell; the median RUL can also be determined as the remaining service life of the abnormal battery cell; the RUL range under different confidence intervals (such as 90%) can also be determined as the remaining service life of the abnormal battery cell, etc. This application does not make any special limitations on this.

[0229] Subsequently, the remaining lifespan of the identified abnormal battery cells can be used as the battery lifespan prediction result.

[0230] In some optional embodiments, the battery life prediction results may also be sent to the user equipment.

[0231] It is understood that, in this embodiment, since the remaining lifespan of the battery as a whole is determined by the remaining lifespan of the abnormal battery cells, determining the remaining lifespan of the abnormal battery cells as the battery lifespan prediction result can make the battery lifespan prediction result more consistent with actual usage patterns.

[0232] In some alternative embodiments, the battery life prediction method provided in this application may further include: estimating residual value and rating secondary utilization potential based on the battery life prediction results, in order to make warranty, replacement, procurement and financial decisions.

[0233] Understandably, in this embodiment, the degradation mode that best matches the target electrochemical parameter value of the abnormal battery cell is found from the failure boundary library. Then, based on the first degradation mode, the decay trajectory of the target electrochemical parameter of the abnormal battery cell is extrapolated to determine the baseline degradation trajectory of the abnormal battery cell. Since the degradation mode is matched with the target electrochemical parameter of the abnormal battery cell, the obtained baseline degradation trajectory of the abnormal battery cell is a relatively accurate degradation trajectory. However, since the first degradation mode is an ideal mode, the obtained degradation trajectory is also an ideal degradation trajectory. In actual applications, there are many uncertainties. If the analysis is directly based on the baseline degradation trajectory... The lifespan of abnormal battery cells may not match the actual situation, leading to inaccurate analysis results. Therefore, it is necessary to incorporate a probabilistic prediction model. By applying multiple uncertainties consistent with actual operating conditions to the baseline degradation trajectory using this model, a probability distribution of the remaining lifespan of abnormal battery cells can be obtained. This makes the probability distribution of the remaining lifespan of abnormal battery cells more consistent with actual operating conditions, thereby making the subsequent lifespan prediction results of abnormal battery cells more accurate. This not only significantly improves the practicality and reliability of the prediction but also enables asset managers to quantify risks and make more scientific decisions regarding replacement, warranty, residual value assessment, and tiered utilization, thereby improving asset operation efficiency.

[0234] As an alternative implementation method, continue to refer to Figure 7 The battery life extension method provided in this application embodiment further includes the following steps: S306. Match the target electrochemical parameter values ​​of the abnormal battery cell with the fault knowledge base to determine the root cause of the abnormality of the abnormal battery cell.

[0235] For example, taking the target electrochemical parameter of the abnormal monomer as the negative electrode diffusion capacity retention rate EHF_Dsn as an example: the expression of the parameter value of the negative electrode diffusion capacity retention rate EHF_Dsn of the abnormal monomer in the target data segment is matched with the fault knowledge base.

[0236] The manifestation can be an increase or decrease. For example, in the fault knowledge base, "the decrease in EHF_Dsn is mainly associated with the aging of negative electrode graphite", and a diagnostic report is automatically generated, such as: "In the battery pack, the EHF_Dsn decay trajectory of battery cell No. 3 deviates significantly from the group due to the negative electrode diffusion capacity retention rate (i.e., the trajectory difference of the EHF_Dsn of battery cell No. 3 is greater than the preset difference), so No. 3 is determined to be the early inconsistency root cause battery (i.e., abnormal cell), that is, battery cell No. 3 is an abnormal battery cell, and the abnormal root cause is the aging of negative electrode graphite of battery cell No. 3."

[0237] S307. Send the root cause of the abnormal battery cell to the user equipment.

[0238] In some optional embodiments, the root cause of the abnormal battery cell can be directly sent to the user equipment; in other optional embodiments, the root cause of the abnormal battery cell can be sent to the user equipment as the root cause of the battery abnormality. This application does not impose any particular limitation on this.

[0239] Understandably, in this embodiment, by matching the target electrochemical parameter values ​​of the abnormal battery cell with the fault knowledge base, the fault cause (i.e., the root cause of the abnormality) that matches the changing trend of the target electrochemical parameters of the abnormal battery cell can be matched from the fault knowledge base. Power batteries follow the "weakest link effect", and the fault cause of the power battery is the root cause of the abnormal battery cell. Sending the root cause of the abnormal battery cell to the user equipment allows the user to clearly know what caused the power battery fault, which is conducive to targeted maintenance and improves maintenance efficiency.

[0240] It is understood that steps S301 to S305 above predict the battery's lifespan under abnormal battery conditions. However, this application also provides a battery lifespan prediction method that predicts the battery's lifespan even before the battery experiences any abnormalities, such as... Figure 8 As shown in the embodiments of this application, another method for predicting battery life includes the following steps: S401. Match the target electrochemical parameter values ​​of each battery cell with the failure boundary library to determine the second degradation mode matched by each battery cell.

[0241] S402. Based on the second degradation mode of each battery cell, extrapolate the decay trajectory of the target electrochemical parameters of each battery cell to determine the baseline degradation trajectory of each battery cell.

[0242] In one optional embodiment, after obtaining the second degradation mode of each battery cell, a second degradation model is matched for each battery cell. For each battery cell, the baseline degradation trajectory of the battery cell is obtained according to the implementation method of the first battery cell as described below: The decay trajectory of the target electrochemical parameters of the first battery cell is extrapolated using the target second degradation model until the parameter value of the target electrochemical parameters of the first battery cell is the same as the target second degradation mode of the first battery cell. Then the extrapolation stops, thus obtaining the baseline degradation trajectory of the first battery cell.

[0243] Among them, the target second degradation model is the second degradation model corresponding to the first battery cell in each second degradation model; the target second degradation mode is the degradation mode of the first battery cell matched from the failure boundary library.

[0244] In this embodiment, the second degradation model can be an exponential decay degradation model, a power-law degradation model, etc., and this application does not make any special limitation on it.

[0245] S403. Use a probabilistic prediction model to predict the baseline degradation trajectory of each battery cell and obtain the probability distribution of the remaining service life of each battery cell.

[0246] Steps S401 to S403 are similar to steps S302 to S304, and will not be described in detail here.

[0247] S404. Based on the probability distribution of the remaining service life of each battery cell, determine the battery life prediction result.

[0248] In some optional embodiments, the remaining lifespan of each battery cell can be determined based on the probability distribution of the remaining lifespan of each battery cell; then, the minimum remaining lifespan of each battery cell is selected as the battery life prediction result, and the battery life prediction result is sent to the user equipment.

[0249] The process of "determining the remaining lifespan of each battery cell based on the probability distribution of the remaining lifespan of each battery cell" is similar to the aforementioned step S305, and will not be repeated here.

[0250] It is understood that, in the embodiments of this application, it is possible to predict the lifespan of a battery even when no abnormal battery cells have appeared in each battery cell, thereby improving the versatility of battery lifespan prediction.

[0251] As an optional embodiment, the battery life prediction method provided in this application embodiment further includes: sending a battery health report to the user equipment.

[0252] It should be noted that the content of the battery health report may vary depending on the user's device role.

[0253] In one example, when the user equipment role is that of an operations and maintenance (O&M) personnel: Inconsistency diagnosis details, location information, and maintenance recommendations (such as "focus on inspecting and testing cell #3") can be sent to the O&M personnel. The O&M personnel can then use cell leveling technology to eliminate inconsistencies or perform preventative maintenance to reduce O&M costs.

[0254] In another example, when the user equipment acts as an asset management platform: the battery pack's health score, RUL probability distribution, residual value estimate, and cascade utilization potential rating can be sent to the asset management platform to facilitate warranty, replacement, procurement, and financial decisions.

[0255] In another example, when the user device acts as a data dashboard, it can provide engineers with in-depth analysis tools such as interactive trajectory maps, historical trends of inconsistency indicators, and performance analysis of predictive models.

[0256] It should be noted that, in some optional embodiments, a virtual battery corresponding to the battery in the vehicle is deployed in the cloud device, and the multiple virtual battery cells included in the virtual battery correspond one-to-one with the physical battery cells included in the battery in the target device.

[0257] Since the virtual battery cell and the physical battery cell are in one-to-one correspondence, it is understandable that the two-stage parameter identification, trajectory extrapolation, and lifespan prediction steps performed by the cloud after receiving the target data of the battery can all be performed on the virtual battery cell, using the virtual battery cell to reflect the physical battery cell.

[0258] In yet another optional embodiment of this application, a cloud device is also provided, such as... Figure 9 As shown, the cloud device 90 includes a receiving circuit 901 and a processing circuit 902.

[0259] The receiving circuit 901 is used to acquire battery status data during vehicle operation. The battery status data includes at least the measured voltage corresponding to multiple timestamps.

[0260] The processing circuit 902 is used to determine the target electrochemical parameter values ​​of multiple cells in the battery that minimize the error between the measured voltage and the simulated voltage of the electrochemical model, based on the battery state data, battery nominal data and electrochemical model.

[0261] Among them, the electrochemical model can be a pseudo-two-dimensional model.

[0262] The processing circuit 902 is also used to determine the inconsistency of the target electrochemical parameters of multiple battery cells based on multiple sets of target electrochemical parameter values ​​obtained from multiple cycles; to determine the health status of multiple battery cells based on the inconsistency of the target electrochemical parameters of multiple batteries; and to determine that the battery is abnormal if the health status is greater than a preset threshold.

[0263] It is understood that, in this embodiment, after acquiring battery state data during vehicle operation, since the battery state data includes at least multiple timestamps corresponding to measured voltages, the target electrochemical parameter values ​​of multiple battery cells in the battery that minimize the error between the measured voltage and the simulated voltage of the electrochemical model can be determined based on the battery state data, battery nominal data, and electrochemical model. This not only improves the convergence speed and quickly determines the target electrochemical parameter values ​​of multiple battery cells, but also makes the determined target electrochemical parameter values ​​of multiple battery cells more closely reflect actual conditions. This is because the target electrochemical parameters can directly and quantitatively characterize the health of specific materials or interface properties within the battery cell. Therefore, by determining the inconsistency of target electrochemical parameters among multiple battery cells obtained through multiple cycles, the actual situation of the battery cells can be reflected at a more microscopic level. Based on the inconsistency of target electrochemical parameters among multiple battery cells, the health status of multiple battery cells can be determined. If the health status is greater than a preset threshold, the battery is determined to be abnormal. Compared with the method of diagnosing whether the battery is abnormal by analyzing the overall voltage of the battery pack, this method can draw conclusions about whether the battery is abnormal before the overall voltage of the battery pack becomes obviously abnormal. It can detect battery abnormalities in a timely and accurate manner, thereby improving the accuracy and timeliness of battery abnormality detection.

[0264] In some optional embodiments, the receiving circuit 901 is further configured to receive a target data packet sent by the vehicle; wherein the target data packet is generated by packaging target data segments collected by the vehicle when a preset trigger condition is met, and the target data segments meet a preset excitation condition; the target data packet is preprocessed to obtain battery status data during vehicle operation.

[0265] In some optional embodiments, the target data segment satisfies a preset excitation condition, including: the current rate change in the target data segment exceeds a preset rate threshold.

[0266] In some optional embodiments, the target data packet includes at least one of the following: vehicle identification code, battery identification information, measured voltage array, total current array, temperature array, timestamp, and initial state of charge.

[0267] In some optional embodiments, the receiving circuit 901 is also used to decrypt and decompress the target data packet to obtain decompressed data; and to allocate and align the decompressed data based on multiple battery cells to determine the battery status data during vehicle operation.

[0268] In some optional embodiments, the processing circuit 902 is further configured to, based on multiple timestamps and using the timestamps as variables, repeatedly execute the step of determining the target electrochemical parameter values ​​of multiple battery cells in the battery that minimize the error between the measured voltage corresponding to the timestamp and the simulated voltage of the electrochemical model, according to battery state data, battery nominal data and electrochemical model, so as to obtain multiple sets of target electrochemical parameter values ​​of multiple battery cells obtained in multiple cycles.

[0269] In some optional embodiments, the processing circuit 902 is further configured to, for each battery cell, obtain the target electrochemical parameter value of the battery cell according to the following steps: determining the initial values ​​of multiple types of electrochemical parameters of the battery cell based on the initial state of charge in the battery nominal data and battery state data; adjusting the initial values ​​of multiple types of electrochemical parameters of the battery cell based on the measured voltage of the battery cell until the error between the measured voltage of the battery cell and the simulated voltage of the electrochemical model is minimized, thereby obtaining the identification results of multiple types of electrochemical parameters of the battery cell; determining the identification result of the target electrochemical parameter from the identification results of multiple types of electrochemical parameters of the battery cell; and determining the target electrochemical parameter value of the battery cell according to the identification result of the target electrochemical parameter and the preset value of the target electrochemical parameter.

[0270] In some optional embodiments, the processing circuit 902 is further configured to obtain the number of cycles and the estimated health status under the timestamp; establish the correlation between the timestamp and the target electrochemical parameter value, number of cycles and estimated health status of the battery cell based on the timestamp; and store the correlation between the timestamp and the target electrochemical parameter value, number of cycles and estimated health status of the battery cell into a preset time series database.

[0271] In some optional embodiments, the processing circuit 902 is further configured to repeatedly execute the step of establishing the correlation between the timestamp and the target electrochemical parameter value, cycle number and health status estimate of the battery cell based on multiple timestamps, so as to obtain multiple sets of correlation between the timestamp and the target electrochemical parameter value, cycle number and health status estimate of the battery cell; and construct a preset time series database based on the correlation between the multiple sets of timestamp and the target electrochemical parameter value, cycle number and health status estimate of the battery cell.

[0272] In some optional embodiments, the processing circuit 902 is further configured to obtain multiple sets of target electrochemical parameter values ​​for multiple battery cells based on a preset time series database; determine the decay trajectory of the target electrochemical parameters of multiple battery cells and the standard deviation of the target electrochemical parameters of multiple battery cells based on the multiple sets of target electrochemical parameter values ​​of multiple battery cells; and determine the inconsistency of the target electrochemical parameters of multiple battery cells based on the decay trajectory of the target electrochemical parameters of multiple battery cells and the standard deviation of the target electrochemical parameters of multiple battery cells.

[0273] In some optional embodiments, the processing circuit 902 is further configured to perform the following steps for each battery cell: plot a scatter plot of the target electrochemical parameters of the battery cell with the timestamp as the horizontal axis and the multiple sets of target electrochemical parameter values ​​of the battery cell as the vertical axis; smooth and fit the scatter plot of the target electrochemical parameters of the battery cell to obtain the decay trajectory of the target electrochemical parameters of the battery cell; and calculate the standard deviation of the multiple sets of target electrochemical parameter values ​​of the battery cell to obtain the standard deviation of the target electrochemical parameters of the battery cell.

[0274] In some optional embodiments, the processing circuit 902 is further configured to perform the following steps for each battery cell: determine the target trajectory difference of the target electrochemical parameters of the battery cell based on the decay trajectory of the target electrochemical parameters of the battery cell and a preset reference trajectory; determine the target standard deviation of the target electrochemical parameters of the battery cell based on the standard deviation of the target electrochemical parameters of the battery cell; and perform a weighted calculation on the target trajectory difference and the target standard deviation to obtain the inconsistency of the target electrochemical parameters of the battery cell.

[0275] In some optional embodiments, the processing circuit 902 is further configured to obtain multiple trajectory difference values ​​of the target electrochemical parameters of the battery cell based on the distance between the decay trajectory of the target electrochemical parameters of the battery cell and a preset reference trajectory; and to perform averaging on the multiple trajectory difference values ​​of the target electrochemical parameters of the battery cell to obtain the target trajectory difference degree of the target electrochemical parameters of the battery cell.

[0276] In some optional embodiments, the processing circuit 902 is further configured to normalize the standard deviation of the target electrochemical parameters of the battery cell to obtain the target standard deviation of the target electrochemical parameters of the battery cell.

[0277] In some optional embodiments, when the target electrochemical parameters include multiple types, the processing circuit 902 is also used to determine that the battery is abnormal if the inconsistency of any type of target electrochemical parameter in any of the multiple battery cells meets a preset condition.

[0278] In some optional embodiments, the cloud device 90 provided in this application embodiment also has the capability to predict battery life.

[0279] For example, the processing circuit 902 is further configured to, after an abnormality is found in the battery, identify the abnormal battery cell in the battery; match the target electrochemical parameter value of the abnormal battery cell with a failure boundary library to determine a first degradation mode of the abnormal battery cell; extrapolate the decay trajectory of the target electrochemical parameter of the abnormal battery cell based on the first degradation mode to determine a baseline degradation trajectory of the abnormal battery cell; use a probability prediction model to predict the lifetime of the baseline degradation trajectory of the abnormal battery cell to obtain a probability distribution of the remaining lifetime of the abnormal battery cell; and determine the battery lifetime prediction result based on the probability distribution of the remaining lifetime of the abnormal battery cell.

[0280] Understandably, in this embodiment, the degradation mode that best matches the target electrochemical parameter value of the abnormal battery cell is found from the failure boundary library. Then, based on the first degradation mode, the decay trajectory of the target electrochemical parameter of the abnormal battery cell is extrapolated to determine the baseline degradation trajectory of the abnormal battery cell. Since the degradation mode is matched with the target electrochemical parameter of the abnormal battery cell, the obtained baseline degradation trajectory of the abnormal battery cell is a relatively accurate degradation trajectory. However, since the first degradation mode is an ideal mode, the obtained degradation trajectory is also an ideal degradation trajectory. In actual applications, there are many uncertainties. If the analysis is directly based on the baseline degradation trajectory... The lifespan of abnormal battery cells may not match the actual situation, leading to inaccurate analysis results. Therefore, it is necessary to incorporate a probabilistic prediction model. By applying multiple uncertainties consistent with actual operating conditions to the baseline degradation trajectory using this model, a probability distribution of the remaining lifespan of abnormal battery cells can be obtained. This makes the probability distribution of the remaining lifespan of abnormal battery cells more consistent with actual operating conditions, thereby making the subsequent lifespan prediction results of abnormal battery cells more accurate. This not only significantly improves the practicality and reliability of the prediction but also enables asset managers to quantify risks and make more scientific decisions regarding replacement, warranty, residual value assessment, and tiered utilization, thereby improving asset operation efficiency.

[0281] In some optional embodiments, the processing circuit 902 is further configured to determine the remaining lifespan of the abnormal battery cell based on the probability distribution of the remaining lifespan of the abnormal battery cell, and to determine the remaining lifespan of the abnormal battery cell as the battery life prediction result.

[0282] In some optional embodiments, the processing circuit 902 is further configured to identify battery cells in the battery whose trajectory difference value of the target electrochemical parameter exceeds a preset value as abnormal battery cells; and / or to identify battery cells in the battery with the largest standard deviation of the target electrochemical parameter as abnormal battery cells.

[0283] In some optional embodiments, after identifying the abnormal battery cell, the processing circuit 902 is further configured to match the target electrochemical parameter value of the abnormal battery cell with the fault knowledge base to determine the root cause of the abnormality of the abnormal battery cell; and send the root cause of the abnormality of the abnormal battery cell to the user equipment.

[0284] In another embodiment of this application, a collaborative system is provided, such as... Figure 10 As shown, the collaborative system 100 includes: vehicle 1001 and cloud device 90.

[0285] Among them: vehicle 1001 is used to extract target data segments from historical data when the vehicle's status meets preset trigger conditions, and to perform data alignment and packaging on the target data segments to generate target data packets and upload them to the cloud device. The cloud device 90 is used to receive target data packets sent by the vehicle, preprocess the target data packets to obtain battery status data during vehicle operation, and the battery status data includes at least: measured voltages corresponding to multiple timestamps; and target electrochemical parameter values ​​of multiple battery cells in the battery that minimize the error between the measured voltage and the simulated voltage of the electrochemical model, based on the battery status data, battery nominal data, and electrochemical model; determining the inconsistency of target electrochemical parameters of multiple battery cells based on multiple sets of target electrochemical parameter values ​​obtained through multiple cycles; determining the health status of multiple battery cells based on the inconsistency of target electrochemical parameters; and determining that the battery is abnormal if the health status is greater than a preset threshold.

[0286] Among them, the electrochemical model can be a pseudo-two-dimensional model.

[0287] As an alternative real-time method, the collaborative system 100 may also include a user device 1002 and a cloud device 90, and may also send the abnormal result of the battery to the user device 1002 after determining that there is an abnormality in the battery.

[0288] Understandably, in this embodiment, the vehicle sends target data segments to the cloud, rather than all historical data. This effectively alleviates network bandwidth pressure and reduces cloud storage costs, making the management of a fleet of thousands of vehicles economically feasible. After acquiring battery status data during vehicle operation, the cloud can determine the target electrochemical parameter values ​​of multiple battery cells that minimize the error between the measured voltage and the simulated voltage of the electrochemical model, based on the battery status data, battery nominal data, and electrochemical model. This not only improves the convergence speed and quickly determines the target electrochemical parameter values ​​of multiple battery cells but also makes the determined target electrochemical parameter values ​​of multiple battery cells more closely reflect actual conditions. Since target electrochemical parameters can directly and quantitatively characterize the health of specific materials or interfaces within a battery cell, the inconsistency of target electrochemical parameters among multiple battery cells obtained from multiple cycles can reflect the actual situation of the battery cells at a more microscopic level. Based on this inconsistency, the health status of multiple battery cells can be determined. If the health status exceeds a preset threshold, an anomaly is identified. Compared to diagnosing battery anomalies by analyzing the overall battery pack voltage, this approach allows for timely and accurate detection of anomalies before a significant voltage abnormality occurs, improving the accuracy and timeliness of anomaly detection.

[0289] The following examples illustrate possible implementation schemes of the battery health status determination method described in one or more of the above embodiments.

[0290] One related technology proposes using historical data from the cloud to train a machine learning model to predict the overall macroscopic health status (i.e., capacity degradation) of a battery pack online. The specific technical solution includes the following steps: First, data acquisition and preprocessing: acquiring battery pack usage data (such as charging data, driving data, etc.) from the vehicle, performing data cleaning, supplementation (e.g., using Lagrange interpolation to fill in missing values), and structured processing (e.g., one-hot encoding of text data). Then feature engineering and dimensionality reduction are performed: ① Feature extraction: Macroscopic statistical features related to battery health status are extracted from preprocessed data, such as cumulative charging time, cumulative driving mileage, charging temperature, average charging power, and average charging rate; ② Feature selection: Pearson correlation coefficients between feature parameters are calculated, and features with correlations higher than a threshold (such as 0.9) are considered redundant and deleted, thus obtaining a simplified feature set; ③ Data augmentation: The Synthetic Minority Over-sampling Technique (SMOTE) algorithm is used to oversample minority class data (such as fast charging data) to solve the data imbalance problem, and the Support Vector Machine (SVM) model is used to label each charge as "fast charging" or "slow charging" to form new supplementary features (such as cumulative fast / slow charging power). Next, SOH calculation and model building are performed: ① SOH calibration: The charging capacity is calculated using the current integral of a specific charging segment (e.g., SOC increasing from 40% to 80%), thereby estimating the actual capacity of the current battery pack and comparing it with the factory rated capacity to obtain the true SOH label for model training; ② Model training: Using the processed feature data (preprocessed data, simplified features, and supplementary features) as input and the calculated SOH as the output label, a Gradient Boosting Decision Tree (GDBT) machine learning model is trained as the battery health status prediction model, and the model parameters are optimized through cross-validation. Finally, cloud deployment and online prediction are performed: The trained GBDT model is deployed on a cloud server using the Flask framework. When a prediction is needed, the system obtains the latest data of the target vehicle online, performs the same feature processing, and calls the model to calculate the battery pack's health status prediction result in real time.

[0291] The overall idea of ​​this solution is to use data mining techniques to find macroscopic features related to battery capacity degradation statistics from massive amounts of operational data, and use these features to train a black-box machine learning model to achieve rapid online cloud-based estimation of the overall SOH of the battery pack.

[0292] However, the aforementioned technologies have the following problems: Problem 1: The prediction results lack interpretability and the ability to diagnose the root causes of aging, and cannot guide precise maintenance.

[0293] This solution employs a purely data-driven "black box" model (GDBT). The model only outputs a macroscopic SOH value (e.g., 85%), but it cannot reveal why the battery is aging, which cell ages first, or what internal mechanisms (e.g., decreased negative electrode diffusion capacity, lithium inventory loss, etc.) lead to aging. When predicting a decrease in SOH, maintenance personnel only know that "the battery is failing," but not "where it's failing" or "why it's failing," making it impossible to pinpoint the specific faulty cell or aging type. This results in blind maintenance decisions (e.g., only being able to replace the entire battery pack), leading to high costs.

[0294] Problem 2: Insensitive to early inconsistencies and latent declines, with a serious lag in early warning.

[0295] This approach predicts the average capacity of the entire battery pack. In the early stages of battery pack use, even if the internal resistance of individual cells increases or the diffusion coefficient decreases (initiating inconsistencies), the model cannot detect this as long as it does not significantly affect the overall charging capacity (calculated by current integration over a specific SOC range). It can only capture relatively late-stage aging already reflected in the capacity, but cannot detect the leading, microscopic parameter changes that cause capacity degradation. Therefore, the early warning is delayed, and the optimal window for preventative maintenance is missed.

[0296] Problem 3: The model heavily relies on data from specific charging segments, limiting its generalization ability and robustness.

[0297] Its core SOH label (the "true value" for model training) relies on selecting a standard constant current charging segment with a SOC between 40% and 80% and an increase of at least 10% for calculation. In real-world complex vehicle usage scenarios (such as frequent shallow charging and discharging, dynamic discharging, and non-standard charging conditions), such ideal segments may be scarce or nonexistent, leading to missing or distorted model input features and prediction failure. Essentially, the model memorizes and fits historical charging patterns, exhibiting poor adaptability to unseen new operating conditions or new battery models.

[0298] Question 4: Unable to provide probabilistic and quantitative predictions of remaining useful life (RUL).

[0299] This solution only predicts the current state of health (SOH), which falls under the category of "state assessment" rather than "lifespan prediction." It lacks a dynamic degradation trajectory model between the state of health and remaining lifespan, thus failing to answer the critical asset management question of "how much longer can the battery last?" Even if the SOH is known, the lack of analysis on degradation rates and trajectory patterns prevents a probabilistic, confidence-interval-based quantitative prediction of the future end of the battery's lifespan, making it difficult to support sophisticated management decisions such as battery replacement, residual value assessment, and secondary utilization.

[0300] The solutions provided by related technologies avoid the internal electrochemical and physical mechanisms of batteries and only look for statistical correlations from external big data. This results in (1) uninterpretable models, (2) insensitivity to early microscopic signals, (3) reliance on ideal data conditions, and (4) inability to perform forward-looking lifetime extrapolation.

[0301] To address the issues of delayed diagnosis, unclear root causes, and inaccurate predictions of battery pack inconsistencies in related technologies, this embodiment aims to solve not only how to non-invasively extract microscopic electrochemical parameters reflecting the aging mechanism of each cell's internal materials from fragmented data of daily battery operation; but also how to utilize the long-term decay trajectory of these parameters to achieve inconsistency diagnosis that is more advanced than voltage differences, and to accurately pinpoint specific cells and aging types; and how to make high-precision, probabilistic predictions of the remaining battery pack life based on the decay trajectory of cell parameters and statistical models, providing quantitative decision support for asset management.

[0302] Based on the technical problems to be solved in this implementation, this embodiment proposes a four-layer cloud-based analysis framework (i.e., a method for determining battery health status) consisting of "microscopic parameter extraction - trajectory map construction - difference quantification - lifetime mapping". Through a distributed parameter calculation engine deployed in the cloud, standardized operational data fragments (i.e., target data fragments) uploaded from the vehicle are processed in batches. This enables non-invasive extraction of microscopic electrochemical parameters reflecting the aging mechanism of materials within each cell, calculating a series of electrochemical health factors (i.e., target electrochemical parameters) for each battery cell, and plotting its full life-cycle decay trajectory map. This allows for early inconsistency diagnosis based on trajectory differences and probabilistic lifetime prediction based on trajectory extrapolation. An intelligent upload mechanism for "standardized data fragments" is designed on the vehicle side (i.e., the vehicle itself): a set of triggering logic and data filtering rules are designed to ensure that the uploaded fragment data (short-term, variable rate) meets the excitation requirements for electrochemical parameter identification while significantly reducing data communication and storage load. A comprehensive inconsistency quantification method integrating "instantaneous divergence" (i.e., target standard deviation) and "dynamic time warping (DTW) distance" (i.e., target trajectory difference) not only focuses on the dispersion of EHF values ​​of each individual cell at the current moment (instantaneous divergence), but also measures the difference in the overall aging trajectory pattern of each cell through DTW distance, thereby capturing complex inconsistency patterns such as "rapid aging followed by slow aging", resulting in more accurate diagnosis. A remaining lifetime probability prediction model based on "failure boundary library" and "Monte Carlo simulation" establishes a mapping relationship between target electrochemical parameters and lifespan termination (failure boundary) using massive historical data. Combining the extrapolation trend of the current individual cell trajectory and multiple sources of uncertainty (measurement noise, model error, future operating conditions), it outputs the probability distribution of remaining lifetime (RUL) through Monte Carlo simulation, rather than a single value, significantly improving the practicality and reliability of the prediction.

[0303] The method for determining the health status of a battery provided in this embodiment will be described in detail below with reference to the accompanying drawings.

[0304] It should be noted that the battery health status determination method provided in this embodiment includes not only the battery health status determination method, but also the battery life prediction method.

[0305] The implementation of this solution is divided into three levels: vehicle-side (i.e., vehicles), cloud-side (i.e., cloud devices), and user-side (i.e., user devices). The specific steps are as follows: Step S100: The vehicle uploads a standard data packet to the cloud.

[0306] like Figure 11 As shown, the specific steps include S201 to S206.

[0307] S201, Vehicle-side battery management system monitoring.

[0308] The vehicle-side BMS continuously monitors the vehicle's operating status.

[0309] S202. Are the triggering conditions met?

[0310] If the triggering condition is met, proceed to step S203; if the triggering condition is not met, proceed to step S201.

[0311] The triggering conditions can include periodic triggering, event triggering, and exception triggering. When any of the triggering conditions is met, the data collection and uploading process for this data segment will begin. Among them, periodic triggering: every fixed calendar time (such as 7 days) or fixed mileage (such as 5000 kilometers).

[0312] Among them, the event trigger is the detection of the end of a complete driving loop that covers a typical SOC range (such as 30%-70%).

[0313] Among them, abnormal triggering: the BMS local algorithm initially judges that the voltage or temperature behavior of a certain unit is slightly abnormal.

[0314] S203, Backtrack and filter data segments.

[0315] Upon triggering, the BMS retrieves a continuous data segment of 5-10 minutes from the cached historical data. This segment must contain a significant change in current excitation (e.g., an acceleration-coasting-braking process with a current rate change exceeding 1C) to ensure the data's "observability" for parameter identification.

[0316] S204, Vehicle-side data preprocessing.

[0317] Subsequently, BMS performs time alignment, invalid value removal, and lightweight compression on the data segment.

[0318] S205, Pack into a standard data package.

[0319] Pack it into a standard data package containing information such as vehicle VIN, battery pack ID (i.e., power battery ID), single cell voltage array, total current array, temperature array, timestamp, and starting SOC.

[0320] S206. Uploaded to the cloud server (i.e., cloud device) via cellular network.

[0321] Step S200: Cloud-based distributed parameter calculation and construction of EHF database (i.e., time series database of target electrochemical parameter values).

[0322] Continue as Figure 11 As shown, the specific steps include S207 to S211.

[0323] S207, Receive data packets in the cloud.

[0324] A distributed parameter calculation engine (such as one based on a Kubernetes cluster) is located in the cloud. After receiving the standard data packet (i.e., the target data packet), the engine executes the following process: S208, Cloud-based data preprocessing.

[0325] Each standard data packet is decrypted, decompressed, and validated. Current data is allocated to each virtual cell according to the battery pack topology and strictly aligned with the cell voltage and temperature data.

[0326] S209, Initiate a parallel parameter identification task for each individual unit.

[0327] For each cell in the battery pack corresponding to the standard data package, an independent parameter identification task is started. Each task calls a two-stage parameter identification algorithm: (1) Initialization: Using the prior information of the battery model (i.e., the nominal parameter library and the initial SOC of the segment), reasonable initial values ​​with physical meaning are provided for all parameters to be estimated (such as the negative electrode diffusion coefficient Dsn, the positive electrode diffusion coefficient Dsp, the negative electrode reaction rate kn, the positive electrode reaction rate kp, the lumped ohmic internal resistance Rct, the negative electrode initial stoichiometric coefficient θn1, and the positive electrode initial stoichiometric coefficient θp0). (2) Optimization solution: Based on the measured voltage curve of the cell in this segment, the electrochemical model parameters are adjusted to minimize the error (such as RMSE) between the model simulation voltage and the measured voltage. The optimization process uses an efficient algorithm and usually converges within a few seconds.

[0328] S210. Calculate the EHF series values.

[0329] For example, calculating the EHF series values ​​is equivalent to calculating the parameter values ​​of the target electrochemical parameters (i.e., the aforementioned "target electrochemical parameter values").

[0330] Extract key parameters from the identification results and calculate the EHF value of this monomer for each type in this data fragment using the following formula: (1) Calculate the negative electrode diffusion capacity retention rate EHF_Dsn: EHF_Dsn=Dsn1 / Dsn_fresh; (2) Calculate the charge transfer impedance growth rate EHF_Rct: EHF_Rct = Rct1 / Rct_fresh; (3) Calculate the positive and negative electrode attenuation asymmetry index EHF_Balance: EHF_Balance=(Dsn1 / Dsp1) / (Dsn_fresh / Dsp_fresh).

[0331] Wherein, Dsn1 represents the optimized value of the negative electrode diffusion coefficient Dsn of this battery model; Rct1 represents the optimized value of the lumped ohmic internal resistance Rct of this battery model; Dsp1 represents the optimized value of the positive electrode diffusion coefficient Dsp of this battery model; *_fresh represents the factory calibration value or early learning value (i.e., preset value) of this battery model in its brand new state.

[0332] S211, Store in EHF time series database.

[0333] The obtained EHF values ​​of each type are associated with the corresponding number of cycles, SOH estimate, timestamp, etc., and written into the cloud EHF time series database.

[0334] Step S300: Construction of parameter decay trajectory map and quantitative diagnosis of inconsistency.

[0335] Continue as Figure 11 As shown, the specific steps include S212 to S217.

[0336] S212, Periodic / On-Demand Aggregation Analysis.

[0337] Cloud devices perform aggregated analysis on the EHF time-series database periodically (e.g., daily) or as needed.

[0338] S213. Draw the attenuation trajectory diagram.

[0339] For a given battery pack, query the time-series data of all its individual cells at various historical time points for a specific EHF type. Plot a scatter plot for each cell with "cumulative equivalent full cycle count" on the horizontal axis and the parameter value of the specific EHF type on the vertical axis. Use smooth spline fitting to generate a continuous and clear "parameter decay trajectory curve." The trajectories of all cells can be overlaid to form the "decay trajectory map" of the battery pack.

[0340] It is understandable that "the trajectories of all individual cells can be overlaid and displayed" here means that the trajectories of all individual cells are plotted on a single graph. For example, if there are 20 individual cells in a battery pack, the trajectories of a certain EHF type of these 20 individual cells are plotted on a single graph. This graph contains 20 trajectories of a certain type of EHF, forming a "degradation trajectory map" for a certain EHF type of the battery pack. The following explanation will use EHF_Dsn as an example to illustrate this further.

[0341] S214, Inconsistency Quantification.

[0342] (1) Calculate the instantaneous divergence (i.e., standard deviation) of each battery cell: Calculate the standard deviation σ of the parameter value of a certain EHF type (e.g., EHF_Dsn) of all cells in the battery pack in the uploaded standard data packet. A large σ value indicates that the values ​​of each cell on this health indicator are large at the current moment.

[0343] (2) Calculate the trajectory morphology difference of each battery cell (i.e., trajectory difference): Select the median trajectory of each EHF_Dsn trajectory in the battery pack as a reference, and calculate the DTW distance between the EHF_Dsn trajectory of each cell and the median trajectory. A large DTW distance indicates that the aging rate pattern of the cell (e.g., fast at first and slow at later) is significantly different from the group pattern.

[0344] (3) Calculate the comprehensive inconsistency index UI: Taking EHF type EHF_Dsn as an example, the comprehensive inconsistency index UI is the first comprehensive inconsistency index UI_index1. Wherein, UI_index1=α1*δ_normal(EHF_Dsn)+β1*Avg(DTW_Dsn).

[0345] Among them, the weights α1 and β1 can be adjusted according to the diagnostic sensitivity. For example, α1=0.4 and β1=0.6; δ_normal is the normalized standard deviation (i.e., the target standard deviation); Avg is the result of meanization (i.e., the target trajectory difference).

[0346] S215. Does the inconsistency index exceed the threshold?

[0347] If the inconsistency index is determined to exceed the threshold, proceed to step S217; if the inconsistency index is determined not to exceed the threshold, proceed to step S216.

[0348] For example, an inconsistency alert is triggered when UI_index1 exceeds the threshold.

[0349] S216. Continue monitoring.

[0350] S217, Root Cause Analysis and Report Generation.

[0351] The cloud server analyzes which battery cell's EHF_Dsn parameter value standard deviation σ contributes the most and which cell has an abnormally high degree of trajectory morphology difference. Combined with the knowledge base (such as "EHF_Dsn decline is mainly associated with negative electrode graphite aging"), a diagnostic report is automatically generated, for example: "In battery pack #001, cell No. 3 is identified as the early inconsistency root cause cell because its negative electrode diffusion capacity (EHF_Dsn) decay trajectory deviates significantly from the group (DTW distance exceeds the standard).

[0352] Step S400: Trajectory-based probability prediction of remaining lifetime.

[0353] Continue as Figure 11 As shown, the specific steps include S218 to S222.

[0354] S218, Failure Boundary Matching.

[0355] The cloud server maintains a failure boundary library learned from massive amounts of retired battery data (i.e., battery state data). For the current battery pack, the latest EHF type of each individual cell is combined and then matched with the failure boundaries to find the most similar degradation mode.

[0356] S219, Individual trajectory extrapolation.

[0357] For each individual, especially the diagnosed abnormal individual, a suitable degradation model (such as exponential decay or power law model) is used to fit and extrapolate the trajectory of a certain EHF type (such as EHF_Dsn).

[0358] S220, Monte Carlo probability prediction.

[0359] Monte Carlo probability prediction is also known as Monte Carlo simulation.

[0360] (1) Set the future operating conditions, model parameter uncertainty, and random distribution of measurement noise.

[0361] (2) Perform thousands of simulations. In each simulation, the trajectory is extrapolated according to the random sampling conditions until a parameter value of a certain EHF type of the unit touches its matching failure boundary. Record the "remaining number of cycles" under this simulation.

[0362] S221. Generate the probability distribution of remaining useful life.

[0363] Summarize all simulation results to form a probability distribution (e.g., normal distribution) of the remaining battery life (RUL). Output key metrics: expected RUL, median RUL, and RUL ranges at different confidence intervals (e.g., 90%).

[0364] S222, Package-level lifetime determination.

[0365] The overall lifespan of a battery pack (i.e., a battery) is determined by the earliest failing cell. The cloud server clearly identifies the "bottleneck cell" that limits lifespan and its predicted failure time.

[0366] Step S500: Multi-dimensional result output and decision support.

[0367] Continue as Figure 11 As shown, the specific steps include S223.

[0368] S223, Multi-dimensional result output.

[0369] The cloud platform pushes the above analysis results to users with different roles: (1) Operation and maintenance personnel: receive alarm work orders, including details of inconsistency diagnosis, location information and maintenance suggestions (such as "focus on testing and inspecting unit 3").

[0370] (2) Asset management platform: Obtain battery pack health scores, RUL probability distributions, residual value estimates and cascade utilization potential ratings for making warranty, replacement, procurement and financial decisions.

[0371] (3) Data dashboard: Provides engineers with in-depth analysis tools such as interactive trajectory maps, historical trends of inconsistency indicators, and performance analysis of predictive models.

[0372] In this embodiment, the early warning time for battery pack inconsistency diagnosis can be significantly advanced (compared to voltage consistency methods). For example, abnormal cells can be detected through EHF trajectory separation in the early stages of cycling before significant voltage differences appear. Simultaneously, the mean absolute error (MAE) of remaining lifetime prediction is superior to traditional prediction methods based on capacity or empirical models. Furthermore, by uploading optimized 5-10 minute data snippets instead of full-process high-frequency data, the monthly data upload volume per vehicle can be reduced by 60%-80%, effectively alleviating network bandwidth pressure and lowering cloud storage costs, making the management of a fleet of thousands of vehicles economically feasible. Moreover, the comprehensive inconsistency index improves the detection rate of early and latent inconsistencies and can pinpoint the root cause of aging (such as negative-electrode-dominated or positive-electrode-dominated degradation), providing a direct basis for precise maintenance. Additionally, the provided remaining lifetime probability distribution (such as "300±25 remaining cycles at 90% confidence level") allows asset managers to quantify risks and make more scientific decisions regarding replacement, warranty, residual value assessment, and tiered utilization, improving asset operation efficiency.

[0373] The following is a specific example using a batch of ternary lithium battery vehicles managed by a car-sharing company. Figure 12 As shown, the collaborative system mainly includes a data upload module A2, a cloud-based calculation module A1, a trajectory analysis module A3, a lifespan prediction module A4, and a decision support module A5.

[0374] The data upload module A2 can execute steps S201 to S205 as described above, and then interact with the cloud device through step S206. For example, after a vehicle (VIN: LSVAU123456) completes a daily operating cycle, the BMS automatically uploads a data segment (approximately 450KB in size) containing acceleration, cruising, and deceleration data on urban roads from the past 10 minutes to the cloud.

[0375] The cloud-based calculation module A1 can execute steps S207 to S213, which will not be elaborated here. For example, the cloud engine completes parameter identification of the 18 cells in the vehicle's battery pack within 2 minutes. Taking the calculation of the negative electrode diffusion retention rate EHF_Dsn as an example: the calculation shows that, except for the negative electrode diffusion retention rate EHF_Dsn of cell #8, the negative electrode diffusion retention rate EHF_Dsn of the other cells in the vehicle's battery pack averages around 0.91 when the cycle period is 150 times. Figure 13 As shown, the negative electrode diffusion retention rate EHF_Dsn of monomer #8 is around 0.67 when the cycle period is 150 times. Figure 14As shown. Since the negative electrode diffusion retention rate EHF_Dsn of cell #8 is abnormal compared to other cells, cell #8 is determined to be an abnormal cell (i.e., an abnormal battery cell). Figure 15 As shown.

[0376] The trajectory analysis module A3 can execute steps S214 to S219 as described above, which will not be repeated here. For example, the cloud server retrieves historical data for the packet and plots the EHF_Dsn trajectory of cell #8. It is found that the EHF_Dsn trajectory of cell #8 has been consistently lower than the average trajectory within the packet since the 100th cycle. For example, its current DTW distance is 0.18 (threshold 0.15), with a significant contribution from the instantaneous standard deviation. The overall inconsistency index UI = 0.13 (threshold 0.10), thus determining that there is "moderate inconsistency," the root cause being "the negative electrode diffusion capacity of cell #8 decays too quickly."

[0377] The lifetime prediction module A4 can execute steps S220 to S222 as described above, which will not be repeated here. For example, the EHF_Dsn trajectory of cell #8 is extrapolated, such as... Figure 16 As shown, and combined with data on similar degradation patterns in the failure boundary library, Monte Carlo simulation predicts that it has a 90% probability of reaching the end-of-life condition within the next 150-220 cycles. Since it is the "weakest link," this prediction represents the overall RUL range of the battery pack.

[0378] The decision support module A5 can execute step S223 as described above, which will not be repeated here. For example, the cloud server sends a work order to the operation and maintenance center: "Warning: Inconsistency of battery pack LSVAU123456 in vehicle. It is recommended that cell #8 be given priority inspection and capacity calibration on the next on-site visit." At the same time, the asset management system updates the expected remaining service life and residual value of the battery pack.

[0379] It should be noted that although the steps of the method in this application are described in a specific order in the accompanying drawings, this does not require or imply that the steps must be performed in that specific order, or that all the steps shown must be performed to achieve the desired result. Additional or alternative steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps; or steps from different embodiments may be combined into a new technical solution.

[0380] The descriptions of the above device embodiments are similar to those of the above method embodiments, and have similar beneficial effects. For technical details not disclosed in the device embodiments of this application, please refer to the descriptions of the method embodiments of this application for understanding.

[0381] It should be noted that the module division in the embodiments of this application is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods. Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, exist as separate physical units, or have two or more units integrated into one unit. The integrated units can be implemented in hardware, as software functional units, or a combination of software and hardware.

[0382] It should be noted that, in the embodiments of this application, if the above-described methods are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, or the parts that contribute to related technologies, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an electronic device to execute all or part of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), magnetic disks, or optical disks. Thus, the embodiments of this application are not limited to any specific hardware and software combination.

[0383] This application provides a cloud device, such as... Figure 17 As shown, the cloud device 90 includes a memory 90-1 and a processor 90-2. The memory 90-1 stores a computer program that can run on the processor 90-2. When the processor 90-2 executes the program, it implements the steps in the method provided in the above embodiments.

[0384] It should be noted that the memory 90-1 is configured to store instructions and applications executable by the processor 90-2, and can also cache data to be processed or already processed in the various modules of the processor 90-2 and the cloud device 90 (e.g., image data, audio data, voice communication data and video communication data), which can be implemented through flash memory or random access memory (RAM).

[0385] This application also provides a computer-readable storage medium for storing computer programs.

[0386] Optionally, the computer-readable storage medium can be applied to the electronic device in the embodiments of this application, and the computer program causes the processor or electronic device to perform the various methods of the embodiments of this application, which will not be described in detail here for the sake of brevity.

[0387] This application also provides a computer program product, including computer program instructions.

[0388] Optionally, the computer program product can be applied to the cloud device in the embodiments of this application, and the computer program instructions cause the processor or cloud device to execute the various methods of the embodiments of this application. For the sake of brevity, they will not be described in detail here.

[0389] This application also provides a computer program.

[0390] Optionally, the computer program can be applied to the cloud device in the embodiments of this application. When the computer program runs on the processor or cloud device, it causes the processor or cloud device to execute the various methods of the embodiments of this application. For the sake of brevity, it will not be described in detail here.

[0391] It should be noted that the descriptions of the cloud devices, storage media, computer program products, and computer program embodiments above are similar to the descriptions of the method embodiments above, and have similar beneficial effects. For technical details not disclosed in the cloud devices, storage media, computer program products, and computer program embodiments of this application, please refer to the descriptions of the method embodiments of this application for understanding.

[0392] It should be understood that the phrases "one embodiment," "an embodiment," or "some embodiments" mentioned throughout the specification mean that a specific feature, structure, or characteristic related to an embodiment is included in at least one embodiment of this application. Therefore, "in one embodiment," "in one embodiment," or "in some embodiments" appearing throughout the specification do not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. It should be understood that in the various embodiments of this application, the sequence numbers of the above-described processes do not imply a sequential order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application. The sequence numbers of the above-described embodiments are merely for descriptive purposes and do not represent the superiority or inferiority of the embodiments. The descriptions of the various embodiments above tend to emphasize the differences between the various embodiments; their similarities or commonalities can be referred to mutually, and for the sake of brevity, they will not be repeated here.

[0393] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that there can be three kinds of relationships. For example, object A and / or object B can represent three situations: object A exists alone, object A and object B exist simultaneously, and object B exists alone.

[0394] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0395] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The embodiments described above are merely illustrative. For example, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple modules or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or modules can be electrical, mechanical, or other forms.

[0396] The modules described above as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules. They may be located in one place or distributed across multiple network units. Some or all of the modules may be selected to achieve the purpose of this embodiment according to actual needs.

[0397] In addition, each functional module in the various embodiments of this application can be integrated into one processing unit, or each module can be a separate unit, or two or more modules can be integrated into one unit; the integrated modules can be implemented in hardware or in the form of hardware plus software functional units.

[0398] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as mobile storage devices, read-only memory (ROM), magnetic disks, or optical disks.

[0399] Alternatively, if the integrated units described above are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, or the parts that contribute to related technologies, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a cloud device to execute all or part of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROMs, magnetic disks, or optical disks.

[0400] The methods disclosed in the several method embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments.

[0401] The features disclosed in the several product embodiments provided in this application can be arbitrarily combined without conflict to obtain new product embodiments.

[0402] The features disclosed in the several method or device embodiments provided in this application can be arbitrarily combined without conflict to obtain new method or device embodiments.

[0403] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, and improvements made within the spirit and scope of this application are included within the scope of protection of this application.

Claims

1. A method of determining a state of health of a battery, the method comprising: The method for determining health status includes: Acquire battery status data during vehicle operation, wherein the battery status data includes at least: measured voltages corresponding to multiple timestamps; Based on the battery state data, battery nominal data, and electrochemical model, target electrochemical parameter values ​​for multiple battery cells in the battery are determined to minimize the error between the measured voltage and the simulated voltage of the electrochemical model; wherein, the electrochemical model is a pseudo-two-dimensional model; Based on the multiple sets of target electrochemical parameter values ​​of the multiple battery cells obtained through multiple cycles, the inconsistency of the target electrochemical parameters of the multiple battery cells is determined. The health status of the multiple battery cells is determined based on the inconsistency of their target electrochemical parameters. If the health status is greater than a preset threshold, it is determined that the battery is abnormal; The step of determining target electrochemical parameter values ​​for multiple battery cells in the battery, based on the battery state data, battery nominal data, and electrochemical model, to minimize the error between the measured voltage and the simulated voltage of the electrochemical model, includes: For each battery cell, the target electrochemical parameter values ​​are obtained according to the following steps: Based on the initial state of charge in the battery nominal data and the battery state data, the initial values ​​of multiple types of electrochemical parameters of the battery cell are determined; Using the measured voltage of the battery cell as a reference, the initial values ​​of multiple types of electrochemical parameters of the battery cell are adjusted until the error between the measured voltage of the battery cell and the simulated voltage of the electrochemical model is minimized, thereby obtaining the identification results of multiple types of electrochemical parameters of the battery cell. The identification result of the target electrochemical parameter is determined from the identification results of multiple types of electrochemical parameters of the battery cell; Based on the identification results of the target electrochemical parameters and the preset values ​​of the target electrochemical parameters, the target electrochemical parameter values ​​of the battery cell are determined.

2. The method of claim 1, wherein, The acquisition of battery status data during vehicle operation includes: Receive the target data packet sent by the vehicle; wherein the target data packet is generated by packaging the target data fragments collected by the vehicle when a preset trigger condition is met, and the target data fragments meet the preset excitation condition; The target data packet is preprocessed to obtain battery status data during vehicle operation.

3. The method of claim 2, wherein, The target data segment satisfies preset excitation conditions, including: the current ratio change in the target data segment exceeds a preset ratio threshold.

4. The method according to claim 2, characterized in that, The target data packet includes at least one of the following: vehicle identification code, battery identification information, measured voltage array, total current array, temperature array, timestamp, and initial state of charge.

5. The method according to claim 4, characterized in that, The step of preprocessing the target data packet to obtain battery status data during vehicle operation includes: The target data packet is decrypted and decompressed to obtain the decompressed data; The decompressed data is allocated and aligned based on multiple individual battery cells of the battery to determine the battery status data during vehicle operation.

6. The method according to claim 1, characterized in that, The method further includes: Based on multiple timestamps, and using the timestamps as variables, the step of determining the target electrochemical parameter values ​​of multiple battery cells in the battery that minimize the error between the measured voltage corresponding to the timestamp and the simulated voltage of the electrochemical model is executed cyclically, so as to obtain multiple sets of target electrochemical parameter values ​​of the multiple battery cells obtained in multiple cycles.

7. The method according to claim 1, characterized in that, The method further includes: Obtain the cycle count and health status estimate under the timestamp; Based on the timestamp, establish the correlation between the timestamp and the target electrochemical parameter value of the battery cell, the number of cycles, and the estimated health status value; The correlation between the timestamp and the target electrochemical parameter value of the battery cell, the number of cycles, and the estimated health status is stored in a preset time series database.

8. The method according to claim 7, characterized in that, The method further includes: Based on multiple timestamps, the step of establishing the correlation between the timestamp and the target electrochemical parameter value of the battery cell, the number of cycles, and the estimated health status is repeatedly executed to obtain multiple sets of correlations between the timestamp and the target electrochemical parameter value of the battery cell, the number of cycles, and the estimated health status. The preset time-series database is constructed based on the correlation between multiple sets of timestamps and the target electrochemical parameter values ​​of the battery cells, the number of cycles, and the estimated health status.

9. The method according to claim 8, characterized in that, The step of determining the inconsistency of the target electrochemical parameters of the multiple battery cells based on multiple sets of target electrochemical parameter values ​​obtained through multiple cycles includes: Based on the preset time series database, multiple sets of target electrochemical parameter values ​​for the multiple battery cells are obtained; Based on multiple sets of target electrochemical parameter values ​​of the multiple battery cells, determine the decay trajectory of the target electrochemical parameters of the multiple battery cells and the standard deviation of the target electrochemical parameters of the multiple battery cells; The inconsistency of the target electrochemical parameters of the multiple battery cells is determined based on the decay trajectory of the target electrochemical parameters of the multiple battery cells and the standard deviation of the target electrochemical parameters of the multiple battery cells.

10. The method according to claim 9, characterized in that, The step of determining the decay trajectory of the target electrochemical parameters of the multiple battery cells and the standard deviation of the target electrochemical parameters of the multiple battery cells based on multiple sets of target electrochemical parameter values ​​includes: For each individual battery cell, the following steps are performed: Using the timestamp as the horizontal axis and the multiple sets of target electrochemical parameter values ​​of the battery cell as the vertical axis, a scatter plot of the target electrochemical parameters of the battery cell is plotted; and the scatter plot of the target electrochemical parameters of the battery cell is smoothed and fitted to obtain the decay trajectory of the target electrochemical parameters of the battery cell. The standard deviation of the target electrochemical parameters of the battery cell is calculated by performing standard deviation calculation on multiple sets of target electrochemical parameter values ​​of the battery cell.

11. The method according to claim 10, characterized in that, The step of determining the inconsistency of the target electrochemical parameters of the multiple battery cells based on the decay trajectory of the target electrochemical parameters of the multiple battery cells and the standard deviation of the target electrochemical parameters of the multiple battery cells includes: For each individual battery cell, the following steps are performed: The target trajectory difference of the target electrochemical parameters of the battery cell is determined based on the decay trajectory of the target electrochemical parameters and the preset reference trajectory. The target standard deviation of the target electrochemical parameter of the battery cell is determined based on the standard deviation of the target electrochemical parameter of the battery cell. The inconsistency of the target electrochemical parameters of the battery cell is obtained by weighting the target trajectory difference degree and the target standard deviation.

12. The method according to claim 11, characterized in that, The step of determining the target trajectory difference of the target electrochemical parameters of the battery cell based on the decay trajectory of the target electrochemical parameters of the battery cell and a preset reference trajectory includes: Based on the distance between the decay trajectory of the target electrochemical parameter of the battery cell and the preset reference trajectory, multiple trajectory difference values ​​of the target electrochemical parameter of the battery cell are obtained. The target trajectory difference degree of the target electrochemical parameter of the battery cell is obtained by averaging multiple trajectory difference values ​​of the target electrochemical parameter of the battery cell.

13. The method according to claim 11, characterized in that, Determining the target standard deviation of the target electrochemical parameters of the battery cell based on the standard deviation of the target electrochemical parameters of the battery cell includes: The standard deviation of the target electrochemical parameters of the battery cell is normalized to obtain the target standard deviation of the target electrochemical parameters of the battery cell.

14. The method according to any one of claims 1 to 13, characterized in that, When the target electrochemical parameters include multiple types, the method further includes: If the inconsistency of any type of target electrochemical parameter in any of the plurality of battery cells meets a preset condition, the battery is determined to be abnormal.

15. A method for predicting battery life, characterized in that, The lifetime prediction method includes: After the battery is found to be abnormal, the abnormal battery cell in the battery is identified. The target electrochemical parameter values ​​of the abnormal battery cell are matched with the failure boundary library to determine the first degradation mode of the abnormal battery cell. Based on the first degradation mode, the decay trajectory of the target electrochemical parameters of the abnormal battery cell is extrapolated to determine the baseline degradation trajectory of the abnormal battery cell. The baseline degradation trajectory of the abnormal battery cell is predicted using a probabilistic prediction model to obtain the probability distribution of the remaining service life of the abnormal battery cell. Based on the probability distribution of the remaining service life of the abnormal battery cells, the life prediction result of the battery is determined.

16. The method according to claim 15, characterized in that, The determination of the battery life prediction result based on the probability distribution of the remaining lifespan of the abnormal battery cells includes: Based on the probability distribution of the remaining service life of the abnormal battery cells, the remaining service life of the abnormal battery cells is determined, and the remaining service life of the abnormal battery cells is determined as the life prediction result of the battery.

17. The method according to claim 15, characterized in that, The process of identifying the abnormal battery cell in the battery includes: Battery cells whose trajectory differences in target electrochemical parameters exceed preset values ​​are identified as abnormal battery cells; and / or, The cell with the largest standard deviation of the target electrochemical parameter in the battery is identified as the abnormal cell.

18. The method according to any one of claims 15 to 17, characterized in that, After identifying the abnormal battery cell, the method further includes: The target electrochemical parameter values ​​of the abnormal battery cell are matched with the fault knowledge base to determine the root cause of the abnormality of the abnormal battery cell. The root cause of the abnormality of the abnormal battery cell is sent to the user equipment.

19. A cloud device, characterized in that, Includes receiving circuitry and processing circuitry, wherein: The receiving circuit is used to acquire battery status data during vehicle operation, wherein the battery status data includes at least: measured voltages corresponding to multiple timestamps; The processing circuit is used to determine, based on the battery state data, battery nominal data, and electrochemical model, the target electrochemical parameter values ​​of multiple battery cells in the battery that minimize the error between the measured voltage and the simulated voltage of the electrochemical model; wherein, the electrochemical model is a pseudo-two-dimensional model; The processing circuit is further configured to determine the inconsistency of the target electrochemical parameters of the multiple battery cells based on multiple sets of target electrochemical parameter values ​​obtained through multiple cycles; determine the health status of the multiple battery cells based on the inconsistency of the target electrochemical parameters; and determine that the battery is abnormal if the health status is greater than a preset threshold. The processing circuit is also used to obtain the target electrochemical parameter values ​​of each battery cell according to the following process: Based on the nominal battery data and the initial state of charge data, initial values ​​of multiple types of electrochemical parameters for the battery cell are determined. Using the measured voltage of the battery cell as a reference, the initial values ​​of the multiple types of electrochemical parameters for the battery cell are adjusted until the error between the measured voltage of the battery cell and the simulated voltage of the electrochemical model is minimized, thus obtaining the identification results of the multiple types of electrochemical parameters for the battery cell. From the identification results of the multiple types of electrochemical parameters for the battery cell, the identification result of the target electrochemical parameter is determined. Based on the identification result of the target electrochemical parameter and the preset value of the target electrochemical parameter, the target electrochemical parameter value for the battery cell is determined.

20. A collaborative system, characterized in that, This includes vehicles and cloud devices, among which: The vehicle is used to extract target data segments from historical data when the vehicle's state meets preset trigger conditions, and to perform data alignment and packaging on the target data segments to generate target data packets and upload them to the cloud device. The cloud device is used to receive the target data packet sent by the vehicle, preprocess the target data packet to obtain battery status data during vehicle operation, wherein the battery status data includes at least: measured voltages corresponding to multiple timestamps; and target electrochemical parameter values ​​of multiple battery cells in the battery that minimize the error between the measured voltage and the simulated voltage of the electrochemical model, based on the battery status data, battery nominal data, and electrochemical model; wherein the electrochemical model is a pseudo-two-dimensional model; determining the inconsistency of the target electrochemical parameters of the multiple battery cells based on multiple sets of target electrochemical parameter values ​​obtained through multiple cycles; determining the health status of the multiple battery cells based on the inconsistency of the target electrochemical parameters; and determining that the battery is abnormal if the health status is greater than a preset threshold. The cloud device is also used to obtain the target electrochemical parameter values ​​for each battery cell according to the following process: Based on the nominal battery data and the initial state of charge data, initial values ​​of multiple types of electrochemical parameters for the battery cell are determined. Using the measured voltage of the battery cell as a reference, the initial values ​​of the multiple types of electrochemical parameters for the battery cell are adjusted until the error between the measured voltage of the battery cell and the simulated voltage of the electrochemical model is minimized, thus obtaining the identification results of the multiple types of electrochemical parameters for the battery cell. From the identification results of the multiple types of electrochemical parameters for the battery cell, the identification result of the target electrochemical parameter is determined. Based on the identification result of the target electrochemical parameter and the preset value of the target electrochemical parameter, the target electrochemical parameter value for the battery cell is determined.

Images