Method for determining the SOH of an accumulator

Abstract

Method for determining the State of Health (SOH) of an accumulator, comprising - Determining the charge level of the battery and charging the battery with a defined charging current; - upon reaching at least a defined charge level of the battery, change the charging current for a time interval ΔT and measure at least one voltage value of the battery at the beginning of the time interval ΔT and at the end of the time interval ΔT as a response signal to the changed charging current; - Calculation of at least one characteristic value from the response signal; - Determining the SOH of the accumulator from the characteristic value using an empirical model for the relationship between the calculated characteristic value and the SOH.

Description

[0001] The present invention relates to a method for determining a State of Health (SOH) of an accumulator, in particular a traction battery of a motor vehicle.

[0002] Rechargeable electrochemical energy storage devices are of great importance, for example, in the field of electromobility. Lithium-ion batteries, in particular, are used as traction batteries for electrically powered vehicles, i.e., purely electric vehicles or vehicles with hybrid drive. A very relevant parameter of a battery is its state of health (SOH), i.e., its remaining capacity relative to its nominal capacity and thus a measure of its aging state. The SOH of a battery is a relevant parameter both for its operation and for service intervals, as well as for further use or recycling.

[0003] The SOH of lithium-ion batteries, for example, can be determined by conducting a full discharge test on initially fully charged batteries or by determining the DC or AC resistance of the lithium-ion battery. Another option is the use of ICA (Incremental Capacity Analysis) or EIS (Electrochemical Impedance Spectroscopy). However, these methods involve significant measurement effort, sometimes under laboratory conditions, and are therefore not suitable for use in ongoing operation.

[0004] It is therefore an object of the present invention to provide a method for determining the SOH of a rechargeable battery, which can be carried out with little technical effort during ongoing operation.

[0005] This problem is solved by the subject matter of the independent patent claim. Advantageous embodiments and further developments are the subject matter of the dependent claims.

[0006] According to one aspect of the invention, a method for determining the SOH of a rechargeable battery is specified, comprising determining a state of charge of the rechargeable battery and charging the rechargeable battery with a specified charging current. When at least one specified state of charge of the rechargeable battery is reached, the charging current is changed for a time interval ΔT, and at least one voltage value of the rechargeable battery is measured at the beginning of the time interval ΔT and at the end of the time interval ΔT as a response signal to the changed charging current. In particular, at least one cell voltage of the rechargeable battery is measured as a voltage value. If the method is carried out for a stack having a plurality of cells, the steps of the method are carried out in particular for each cell.

[0007] Furthermore, the method comprises calculating at least one characteristic value from the response signal and determining the SOH of the accumulator from the characteristic value using an empirical model for the relationship between the calculated characteristic value and the SOH.

[0008] In this method, the battery is charged in a defined manner, i.e., starting from a known charge level, the battery is charged with a specified charging current. Charging is interrupted or the charging current is changed at least once, but typically several times, when specified charge levels of the battery are reached.

[0009] The battery's reaction to the changed charging current is now observed. For this purpose, the voltage value is measured at the beginning of the time interval ΔT. The measurement time is chosen so that the change in the charging current does not yet have an effect, i.e., the measurement is taken immediately at the beginning of the time interval ΔT. Furthermore, the voltage value is measured a second time at the end of the time interval ΔT, when the changed charging current has an effect.

[0010] A characteristic value is calculated from these at least two measured voltage values. In particular, the characteristic value is determined from the measured voltage values ​​and at least one further measured variable recorded in the time interval ΔT. The charging current, for example, is suitable for this purpose. For example, an equivalent resistance value ZTR can be calculated as the characteristic value, which is defined by ZTR = ΔU / ΔI, where ΔU is the difference between the voltage values ​​measured at the beginning and end of the time interval, and ΔI is the change in the charging current in the time interval ΔT.

[0011] From this characteristic value, the SOH of the accumulator is then determined using an empirically derived model.

[0012] As it has been shown, a particularly high accuracy in determining the SOH is achieved when both the equivalent resistance value ZTR and the voltage measured at the beginning of the time interval ΔT are used as characteristic values.

[0013] The method has the advantage of enabling a high-accuracy determination of a battery's SOH during operation. It has been shown that short time intervals of 0.1 second ≤ ΔT ≤ 10 seconds, and in particular 0.1 second ≤ ΔT ≤ 1 second, are sufficient to determine the response signal. It has been shown that the length of the ΔT interval has only a minor impact on the quality of the SOH determination. A relatively short time interval ΔT can therefore be selected to minimize disruption to the charging process. No additional measurement technology is required for the required measurements, as the voltage values, for example, can be measured by the battery management system of the battery.

[0014] According to one embodiment, the specified charging current is constant over time, except for the time interval ΔT. Alternatively, however, it can also vary and follow a specified profile.

[0015] During the time interval ΔT, the charging current can be interrupted, meaning no charging current flows during the time interval ΔT. Alternatively, the charging current can be changed during the time interval ΔT by increasing or decreasing the charging current during the time interval ΔT, meaning a higher or lower constant charging current can be selected than outside the time interval ΔT. Alternatively, it is also possible to vary the charging current during the time interval ΔT so that it follows a defined profile.

[0016] According to one embodiment, a plurality of charging states are defined, upon reaching which the charging current is changed for the time interval ΔT and the response signal to the changed charging current is measured.

[0017] Since it is sufficient to select a time interval of just a few seconds, even a large number of interruptions only marginally impacts the battery's charging time. For example, 50 or more time intervals ΔT with varying charging current can be inserted during a battery charging process.

[0018] According to one embodiment, more than two voltage values ​​of the accumulator are measured during the time interval ΔT in order to achieve higher accuracy.

[0019] The measurement of at least one voltage value can be carried out, in particular, by means of a battery management system of the accumulator. The battery management system of the accumulator, for example, of an electrically powered vehicle, already has devices for measuring individual cell voltages.

[0020] The accumulator can be designed, in particular, as a lithium-ion accumulator, for example, a lithium cobalt(III) oxide accumulator. However, the method can also be used with other rechargeable electrochemical storage devices, such as lithium iron phosphate batteries or solid-state batteries.

[0021] The empirical model is created, in particular, by determining a battery's state of charge and charging the battery with a specified charging current. Upon reaching at least one specified battery state of charge, the charging current is changed for a time interval ΔT, and at least one battery voltage value is measured at the beginning of the time interval ΔT and at the end of the time interval ΔT as a response signal to the changed charging current. The steps are repeated for a plurality of charging cycles, with an SOH of the battery being determined for each charging cycle using a laboratory method.

[0022] A laboratory method for determining the SOH is understood here and in the following to mean a method which determines the SOH independently of the method according to the invention in a different way, for example by a complete discharge test or the determination of the direct current resistance or alternating current resistance.

[0023] The model, which can be structured as a neural network or multilinear regression model, for example, provides a relationship between the calculated characteristic value, such as the ZTR value, and the independently determined SOH of the battery. To create the model, data from batteries can be collected over many charge and discharge cycles.

[0024] Embodiments of the invention are described below by way of example with reference to schematic drawings. Fig. 1 shows steps of a method for determining the SOH of a battery according to an embodiment of the invention, Fig. 2 shows the course of the charging current according to an embodiment of the invention, Fig. 3 shows the cell voltage curve in the Fig. 2 shown time interval, Fig. 4 schematically shows input and output variables of a model for determining the SOH of a battery according to an embodiment of the invention, Fig. 5 shows a diagram illustrating the relationship between a characteristic value determined according to an embodiment of the invention, the SOC of a battery and its SOH, Fig. Figure 6 shows a diagram in which cell voltages were plotted at 50% SOC for different SOH of a battery and Fig. Figure 7 shows a diagram in which the equivalent resistance values ​​determined as characteristic values ​​at 50% SOC for different SOH of a battery were plotted.

[0025] Fig. Figure 1 schematically shows the sequence of a method according to one embodiment of the invention for determining the SOH of a rechargeable battery. The method is particularly suitable for determining the SOH during continuous operation of the rechargeable battery, i.e., not only under laboratory conditions.

[0026] In step 101, the current state of charge (SOC) of the battery is determined, and charging of the battery begins. Charging occurs with a specified charging current, which can be constant or vary over time in a predetermined manner.

[0027] During the charging process, a query is made in step 102, for example at regular intervals, as to whether a specified charge level of the accumulator has been reached. For example, specified charge levels can be set at 1%, 2%, 3%, or 70%, so that 70 specified charge levels exist for a charging process, which are reached one after the other during the query in step 102.

[0028] Once such a specified state of charge is reached, the charging current is changed in a specified manner. For example, the charging current can be interrupted for a time interval ΔT.

[0029] In one embodiment, the accumulator is charged with a constant charging current of 0.5 C and the charging current is interrupted for 3 seconds when the specified charging states are reached.

[0030] In step 103, voltage values ​​at the accumulator, ie, in particular cell voltages, are also measured as a response signal to the changed charging current. For this purpose, at least one voltage value is measured at the beginning of the time interval ΔT and one voltage value at the end of the time interval ΔT.

[0031] A characteristic value is calculated from this response signal in step 104. In particular, an equivalent resistance value ZTR is calculated as the characteristic value, which is defined by ZTR = ΔU / ΔI, where ΔU is the difference between the voltage values ​​measured at the beginning of the time interval and at the end of the time interval, and ΔI is the change in the charging current in the time interval ΔT.

[0032] In a step 105, the SOH of the accumulator is determined from the characteristic value or from a plurality of characteristic values ​​calculated for voltage values ​​measured at different SOCs using an empirical model.

[0033] Fig. Figure 2 shows the charging current profile during the method according to the invention. In the illustrated embodiment, a lithium-ion battery is charged with a constant current of 0.5 C. After reaching a specified state of charge, the charging current is switched off at time t1 for a duration of ΔT = t2 - t1. At time t2, the charging process is continued with the same charging current of 0.5 C. The duration of the time interval ΔT is 3 seconds in the example shown, but can be varied and adapted to the prevailing operating conditions.

[0034] Such interruptions of the charging current for the period ΔT can be repeated during the charging process at specified charging states of the accumulator.

[0035] Fig. 3 shows the cell voltage curve during the Fig. 2. The cell voltage is continuously recorded as a response signal to the change in the charging current, i.e., also during the time interval ΔT. The goal is to record the change in the response signal due to the interruption or resumption of the charging process. The cell voltage is typically measured during the entire charging process, but at least within the scope of the method according to the invention during all time intervals ΔT.

[0036] The response signal should meet the requirement of reflecting the changes in the battery's electrochemical process as accurately as possible. Battery management systems can usually detect signals such as voltage, current, and temperature. The response signal that most accurately reflects the change in charging current is the voltage signal. It has a relatively fast response time and can be measured with high accuracy.

[0037] As in Fig. As shown in Figure 3, the voltage value U1 is measured as the last voltage value before the current interruption begins, or before the current interruption begins to take effect, while U2 is measured as the last voltage value during the current interruption. The sampling interval for signal acquisition must be set appropriately, as the accuracy of the acquisition of U2 affects the calculation of the characteristic value, so a shorter sampling interval results in greater accuracy. In the illustrated example, a sampling interval of 100 ms is used. The current and voltage measurements are performed as simultaneously as possible.

[0038] Typically, the cell voltages are sampled by the BMS at fixed time intervals, for example, every 10 ms. If necessary, the sampling rate can be increased during the ΔT time interval and, if necessary, shortly before it. This ensures good resolution during the ΔT time interval, but avoids a generally high bus load caused by a high sampling rate during the entire charging process.

[0039] In the described embodiment, the equivalent resistance value ZTR = (U1 - U2) / (I1 - I2) is used as the characteristic value. The voltage value U1 can serve as a further characteristic value.

[0040] N battery charge states are defined, at which the charging current is changed for the time interval ΔT. For these charge states, n voltage values ​​U1 are determined and n equivalent resistance values ​​ZTR are calculated, which are designated U1.1, U1.2... U1.n or ZTR1, ZTR1... ZTRn. Based on these characteristic values, a regression model is constructed to determine the battery's SOH from the characteristic values.

[0041] The model, its input variables and its output, the SOH of the accumulator, are in Fig. 4. The model can, in particular, be a neural network. It is based on current and voltage measurements performed in advance, i.e., not during operation of the accumulator, but for known SOHs, using the method described in steps 101 to 104. A portion of the determined parameters, for example, 90% of the parameters, can be used to train the model, while the remaining portion can be used to validate the model.

[0042] It turns out that the model becomes more accurate when the measured voltage values ​​are used as parameters instead of the determined equivalent resistance values. However, a significant improvement in accuracy can be achieved for all states of charge when both voltage values ​​and equivalent resistance values ​​are used.

[0043] Fig. Figure 5 shows examples of equivalent resistance values ​​ZTR plotted against the battery's SOC for various battery SOHs. Curve a corresponds to an SOH of more than 97%, curve b to an SOH of 91-94%, curve c to an SOH of 88-97%, curve d to an SOH of 85-88%, curve e to an SOH of 82-85%, and curve f to an SOH of less than 82%.

[0044] The diagram in Fig. Figure 5 shows that the equivalent resistance value is approximately linearly correlated with the SOH value, and therefore, a change in the SOH value can be expressed by the change in the characteristic value ZTR. Similarly, the voltage value U1 is correlated with the SOH, so it can also be used as an input value for the regression model.

[0045] This is also evident in the Fig. 6 and Fig. 7, which show the trend of the voltage values ​​at 50% SOC for different SOHs and the trend of the equivalent resistance values ​​ZTR at 50% SOC for different SOHs. It can be seen that knowledge of U1 and ZTR for a single battery charge level allows for far-reaching conclusions about the SOH. For greater accuracy, however, these parameters can be determined for a variety of charge levels. For this purpose, the described regression model is used, as described in Fig. 4. Both the voltage values ​​and the equivalent resistance values ​​at different SOHs can be used individually or together as input.

[0046] The method can be used to determine the SOH of a rechargeable battery during operation. The method is particularly suitable for traction batteries in electric vehicles. The software for controlling the method is stored in a vehicle computer, for example, in a central control unit. The required current and voltage measurements are performed, for example, by the battery management system.

[0047] Parts of the process can also be performed by a cloud application. For example, the model used can be stored on a cloud computer that is accessed. The calculation of the characteristic value can also be performed locally or in the cloud. Cloud solutions reduce the load on the vehicle's computer system and facilitate updates. List of reference symbols: 101 steps 102 steps 103 steps 104 steps 105 steps

Claims

[1] Method for determining the State of Health (SOH) of a battery, comprising - Determining the charge level of the accumulator and charging the accumulator with a specified charging current; - when at least one specified charge state of the accumulator is reached, changing the charging current for a time interval ΔT and measuring at least one voltage value of the accumulator at the beginning of the time interval ΔT and at the end of the time interval ΔT as a response signal to the changed charging current; - Calculation of at least one characteristic value from the response signal; - Determining the SOH of the accumulator from the characteristic value using an empirical model for the relationship between the calculated characteristic value and the SOH. [2] Method according to claim 1, wherein the fixed charging current is constant over time except for the time interval ΔT. [3] Method according to claim 1 or 2, wherein the charging current is interrupted in the time interval ΔT. [4] Method according to claim 1 or 2, wherein the charging current is increased or reduced in the time interval ΔT. [5] Method according to claim 1, wherein the charging current varies both within and outside the time interval ΔT and follows a fixed profile. [6] Method according to one of claims 1 to 5, wherein the characteristic value is determined from the measured voltage values ​​and at least one further measured variable recorded in the time interval ΔT. [7] Method according to claim 6, wherein an equivalent resistance value ZTR is calculated as the characteristic value, which is defined by ZTR = ΔU / ΔI, where ΔU is the difference between the voltage values ​​measured at the beginning of the time interval and at the end of the time interval and ΔI is the change in the charging current in the time interval ΔT. [8] Method according to claim 7, wherein both the equivalent resistance value ZTR and at least one measured voltage value are used as the characteristic value. [9] Method according to one of claims 1 to 8, wherein a plurality of charging states are defined, upon reaching which the charging current is changed for the time interval ΔT and the response signal to the changed charging current is measured. [10] Method according to one of claims 1 to 9, wherein more than two voltage values ​​of the accumulator are measured during the time interval ΔT. [11] Method according to one of claims 1 to 10, wherein the measuring of the at least one voltage value is carried out by means of a battery management system of the accumulator. [12] Method according to one of claims 1 to 11, wherein the duration of the time interval ΔT is between 0.1 second and 10 seconds. [13] Method according to one of claims 1 to 12, wherein the accumulator is designed as a Li-ion accumulator. [14] Method according to one of claims 1 to 13, wherein the empirical model is created by - Determining the charge state of a battery and charging the battery with a specified charging current; - when at least one specified charge state of the accumulator is reached, changing the charging current for a time interval ΔT and measuring at least one voltage value of the accumulator at the beginning of the time interval ΔT and at the end of the time interval ΔT as a response signal to the changed charging current; - Repeating the two aforementioned steps for a plurality of charging cycles, whereby for each charging cycle an SOH of the accumulator is determined by means of a laboratory method. [15] A computer program product comprising instructions which, when the program is executed by a computer, cause the computer to carry out the method according to any one of claims 1 to 14.

Images