Method for the predictive identification of ageing reactions in a device battery

Abstract

The invention relates to a method for performing a prognostic diagnosis on a device battery (11) of a technical device (1) in order to identify dendrite formation, in particular lithium plating, having the following steps: - performing (S5) a charging process or a discharging process of the device battery (11) with a predetermined charge change; - measuring and recording (S6) an end voltage time curve of the device battery (11) during a relaxation phase following the charging process; - providing (S7) a simulated end voltage time curve during the relaxation phase by means of a provided relaxation model; - comparing (S8) the measured end voltage time curve with the simulated end voltage time curve; - signaling (S9) a fault of the device battery (11) when a deviation between the measured curve and the simulated curve is determined.

Description

Technical Field

[0001] This invention relates to device batteries, and more particularly to a method for diagnosing lithium-ion device batteries in terms of lithium plating. Furthermore, this invention relates to a method for early identification of lithium plating formation. Background Technology

[0002] In addition to cyclic aging, device batteries undergo other degradation effects that can lead to premature failure or malfunction. Such failures can result in thermal runaway, sudden death, or rapid capacity loss.

[0003] Various methods are known to be available that can identify developing faults in advance. This allows for the implementation of appropriate countermeasures tailored to the specific fault type.

[0004] For lithium-ion batteries, the main problem lies in the lithium coating. The lithium coating manifests as metallic lithium deposited on the negative electrode in the form of needle-like structures or dendrites within the separator, leading to expansion characteristics of the active material and, consequently, of the individual cell.

[0005] If a needle-like structure made of lithium grows through the separator, it can cause a short circuit within the cell and lead to sudden failure. Depending on the stored charge, this short circuit can also cause a sudden temperature rise, which could potentially cause the device's battery to catch fire.

[0006] This effect of lithium coatings is promoted by the rapid charging of lithium-ion batteries, low temperatures, and the intense lithiation of graphite.

[0007] Lithium plating cannot be easily identified based on operating parameters of the device's battery, such as terminal voltage, single-cell voltage, battery current, battery temperature, and state of charge. Even if aging conditions can be obtained using a suitable aging condition model, it cannot provide information about whether lithium plating has formed or to what extent it has formed. Summary of the Invention

[0008] The purpose of this invention is to provide a method for predictively identifying the aging reaction of equipment batteries during maintenance and operation, which can identify in advance the gradual short circuits caused by lithium plating.

[0009] This objective is achieved by the method and a corresponding device for predictively identifying harmful aging reactions caused by lithium plating in the battery during battery maintenance, as described in claim 1.

[0010] Other design options are given in the dependent claims.

[0011] According to the first aspect, a method for performing predictive diagnostics on the battery of a technical device to identify dendrite formation, particularly lithium plating, is provided, comprising the following steps: - To perform the charging or discharging process of the device battery with a predetermined amount of charge change (Ladungshub); - Measure and record the time-varying waveform of the device's battery terminal voltage during the subsequent relaxation phase of the charging process; - The provided relaxation model is used to provide simulated terminal voltage time variation curves during the relaxation phase; - Compare the measured terminal voltage change curve with the simulated terminal voltage change curve; - When a deviation is found between the measured change curve and the simulated change curve, a signal is used to indicate a fault in the device's battery.

[0012] The above method can be used to identify dendrite formation in the form of lithium plating at the anode of the device's battery. To this end, the terminal voltage-time variation curve during the relaxation phase is observed and compared with the corresponding simulated terminal voltage-time variation curve.

[0013] The relaxation phase corresponds to the zero-current phase following the operation of a technical device running on its battery. Thus, a relaxation phase occurs during operation, such as after a charging process (if the technical device does not immediately restart by drawing power from the battery after charging), and after the operation phase, and after each shutdown of the technical device. Specifically, after charging, the terminal voltage decreases over time during the relaxation phase, ideally in a hyperbolic manner.

[0014] It is certain that dendrite formation in lithium-ion batteries, especially lithium plating, causes the terminal voltage change curve during the relaxation phase to deviate from a hyperbolic shape. This hyperbolic deviation is formed by a raised portion of the dendrite formation; the more intense the dendrite formation, the greater the deviation. After charging, the hyperbolic curve is characterized by a leftward bend as a function of the terminal voltage-time change curve. Similarly, after discharging, it is characterized by a rightward bend as a function of the terminal voltage-time change curve. If dendrite formation occurs, the curve's trajectory changes after charging or discharging, causing the curvature to decrease and even become negative, or increase and even become positive, during the terminal voltage change process.

[0015] Furthermore, the deviation between the measured terminal voltage change curve and the simulated terminal voltage change curve can be determined by checking whether the deviation exceeds a predetermined threshold value or whether the measured terminal voltage change curve at least partially exceeds or falls below an upper confidence threshold or a lower confidence threshold, which is provided by the relaxation model based on the simulated terminal voltage change curve.

[0016] Therefore, it is possible to monitor the measured terminal voltage change curve during the relaxation phase compared to the simulated change curve. If the deviation from the simulated change curve is determined to exceed a predetermined threshold value during the terminal voltage change process during the relaxation phase, dendrite formation can be identified and signaled accordingly. This allows for operational limitations such as restricting charging and / or discharging currents or replacing the device's battery.

[0017] What can be set is that the relaxation model is used to provide a terminal voltage or terminal voltage-time variation curve related to the duration from the end of the charging or discharging process, based on the aging state, the state of charge after the charging or discharging process, and the average battery temperature during the charging or discharging process.

[0018] Furthermore, the relaxation model can be used to construct a terminal voltage time change curve based on the charging or discharging current during the charging or discharging process, the average terminal voltage during the charging or discharging process, the terminal voltage at the end of the relaxation phase, the charge change during the final charging or discharging, and the final state of charge after the charging or discharging process.

[0019] The pre-defined simulated change curve can be determined based on the provided relaxation model. The relaxation model can be designed as a parametric model, a data-based model, or a hybrid model, and the simulated terminal voltage change curve is determined based on one or more of the following parameters: the current aging state, the charging or discharging current during the charging or discharging process, the average terminal voltage during the charging or discharging process, the terminal voltage at the end of the relaxation phase, the charge change during the last charge or discharge, the average battery temperature during the charging or discharging process, and the final state of charge after the charging or discharging process.

[0020] During maintenance, it is possible to determine the constraints for initial charging or discharging, battery temperature, charge change and / or charging or discharging current, so that the simplified relaxation model only needs to consider the current aging state and other undefined parameters as input parameters.

[0021] Preferably, the relaxation model can be constructed, trained, or parameterized to output a simulated terminal voltage change curve based on the current aging state, the state of charge after charging or discharging, and / or the terminal voltage after charging or discharging, or the average terminal voltage during the charging or discharging process, and the current average battery temperature during the charging or discharging process.

[0022] Alternatively, a device may be provided for performing one of the aforementioned methods. Attached Figure Description

[0023] The implementation method will now be explained in more detail with reference to the accompanying drawings. Wherein: Figure 1 A system for predictive diagnostics of vehicle batteries in technological vehicles is shown to identify dendrite formation, particularly lithium plating. Figure 2 A flowchart illustrating the method for diagnosing vehicle batteries is shown, and Figure 3 The diagram shows the terminal voltage variation curve during the relaxation phase when dendrite formation has occurred in a lithium-ion battery. Detailed Implementation

[0024] The method according to the invention is described below with reference to a vehicle battery, which is a device battery in a motor vehicle. Dendrite formation can be identified based on the voltage change curve during the relaxation phase during maintenance.

[0025] The examples above represent multiple stationary or mobile devices that do not rely on the power grid for energy supply, such as vehicles (electric vehicles, electric bicycles, etc.), systems, machine tools, household appliances, Internet of Things (IoT) devices, etc.

[0026] Figure 1 The image shows vehicle 1 with its battery in operation at diagnostic station 2, which is used to diagnose vehicle batteries in terms of dendrite formation. Figure 1 Vehicle 1 is shown, which is in communication with diagnostic station 2.

[0027] Motor vehicle 1 has a vehicle battery 11, which serves as a rechargeable electrical device battery, an electric drive motor 12, and a control unit 13. The control unit 43 is connected to a communication module 14, which is adapted to transmit data between the respective motor vehicle 1 and the diagnostic station 2.

[0028] For diagnostic purposes, the vehicle 1 is connected to a diagnostic station 2, for example, in a factory, to implement a method for diagnosing the vehicle battery 11 in terms of dendrite formation.

[0029] Diagnostic station 2 has a data processing unit 21 and a database 22 for storing relaxation models, etc., in which the methods described below can be implemented. An algorithm can be implemented in diagnostic station 2 to perform a diagnosis on the vehicle battery 11 regarding the presence of a lithium coating.

[0030] Figure 2 A flowchart is shown to illustrate a method for diagnosing vehicle battery 11 in terms of dendrite formation, particularly lithium plating formation. The method is implemented in diagnostic station 2 and is carried out in software or hardware within the data processing unit 21 of diagnostic station 2.

[0031] The method is explained below using the charging process as an example. The method can also be applied in a similar manner using the discharging process.

[0032] In step S1, diagnostic station 2 is connected to vehicle battery 11.

[0033] Then, in step S2, the current state of charge of the vehicle battery 11 is determined based on the measurement of the opposite-end voltage. This can be achieved, for example, based on the no-load voltage characteristic curve parameterized for the relevant vehicle battery 11.

[0034] In step S3, it is checked whether the state of charge (SOC) is higher than a predetermined SOC threshold, for example, 60%. If this is the case (either option: yes), then in step S4, the vehicle battery 11 is first discharged to, for example, reduce the SOC by a predetermined SOC difference, such as by 10% or 20%. Then, the process jumps back to step S2.

[0035] If the state of charge (SOC) is not higher than a pre-defined SOC threshold (either option: No), then in step S5, the vehicle battery 11 is charged with a predetermined amount of charge in a predetermined manner. Charging is performed at a constant charging current and at a predetermined battery temperature or ambient temperature. The charge change can be, for example, 20% SOC. In step S6, the terminal voltage is measured simultaneously during charging, and the measured terminal voltage change curve is stored in between.

[0036] In step S7, a simulated voltage change curve during the relaxation phase is obtained using the provided relaxation model based on the current state of charge, the aging state of the vehicle battery 11, the battery temperature, and especially the terminal voltage reached at the end of the charging process.

[0037] Therefore, the relaxation model can be constructed to provide, for example, as an output vector, the terminal voltage or, in particular, the terminal voltage time variation curve related to the duration from the end of the self-charging process, based on the aging state and the average battery temperature during the charging process.

[0038] Relaxation models can be constructed, for example, as data-based probabilistic models or Gaussian process models, thus providing confidence intervals as a measure of uncertainty in model estimation for the modeling of terminal voltage variation curves. From this, upper and lower confidence thresholds are derived.

[0039] In step S8, the simulated terminal voltage change curve is compared with the measured terminal voltage change curve, and a deviation is checked. If at a certain point in time the deviation exceeds a pre-given threshold or the deviation is outside the upper and lower confidence thresholds given by the relaxation model (either option: yes), then in step S9, an anomaly in the form of lithium plating is indicated by a signal. This can lead to the replacement of vehicle battery 11 or operational limitation in the form of current limiting of vehicle battery 11. Otherwise (either option: no), a signal indicates that no dendrite formation has occurred in vehicle battery 11.

[0040] Figure 3 Exemplary terminal voltage change curves during the relaxation phase R are schematically shown for a compliant vehicle battery (K1) and a faulty vehicle battery 11 (K2), along with simulated terminal voltage change curves (K3) along with upper and lower confidence thresholds (dashed lines, K4, K5). It can be seen that the terminal voltage change curve for the compliant vehicle battery 11 lies within the upper and lower confidence thresholds, while the terminal voltage change curve for the faulty vehicle battery 11 lies outside the upper confidence threshold region, thereby enabling the identification of dendrite formation.

Claims

1. A method for performing predictive diagnostics on a device battery (11) of a technical device (1) to identify dendrite formation, particularly lithium plating, the method being, in particular, at least partially computer-implemented, comprising the following steps: - Perform the charging or discharging process of the device battery (11) by a predetermined amount of charge change (S5); - Measure and record (S6) the terminal voltage time variation curve of the device battery (11) during the relaxation phase following the charging process; - Using the provided relaxation model, simulated terminal voltage time variation curves are provided (S7) during the relaxation phase; - Compare the measured terminal voltage time variation curve with the simulated terminal voltage time variation curve (S8). - When a deviation is found between the measured change curve and the simulated change curve, a signal is used to indicate (S9) the fault of the device battery (11).

2. The method according to claim 1, wherein, If it is determined that a predetermined amount of charge change during the charging process (S3) causes the final state of charge to exceed a predetermined threshold, then the device battery (11) is discharged (S4) with at least the predetermined amount of charge change, or if it is determined that a predetermined amount of charge change during the discharging process causes the final state of charge to be less than a predetermined threshold, then the device battery (11) is charged with at least the predetermined amount of charge change.

3. The method according to claim 1 or 2, wherein, The relaxation model is constructed to provide a terminal voltage or terminal voltage-time variation curve related to the duration from the end of the charging or discharging process, based on the aging state, the state of charge after the charging or discharging process, and the average battery temperature during the charging or discharging process.

4. The method according to claim 3, wherein, The relaxation model is further constructed to provide a terminal voltage time variation curve based on the charging or discharging current during the charging or discharging process, the average terminal voltage during the charging or discharging process, the terminal voltage at the end of the relaxation phase, the charge change during the last charging or discharging, and the final charge state after the charging or discharging process.

5. The method according to any one of claims 1 to 4, wherein, The relaxation model is constructed as a data-based model, especially a data-based probabilistic model, and particularly a Gaussian process model.

6. The method according to any one of claims 1 to 5, wherein, The deviation between the measured terminal voltage change curve and the simulated terminal voltage change curve is determined by checking whether the deviation exceeds a predetermined threshold value or whether the measured terminal voltage change curve at least partially exceeds or falls below an upper confidence threshold or a lower confidence threshold, which is provided by the relaxation model based on the simulated terminal voltage change curve.

7. An apparatus for performing any one of the methods according to any one of claims 1 to 6.

8. A computer program product comprising instructions that, when implemented by at least one data processing device, cause the data processing device to perform the steps of the method according to any one of claims 1 to 6.

9. A machine-readable storage medium comprising instructions that, when executed by at least one data processing device, cause the data processing device to perform the steps of the method according to any one of claims 1 to 6.

Images