Method for monitoring the voltages of battery cells of a vehicle battery, and corresponding control unit

Abstract

The invention relates to a method for monitoring a plurality of voltages (Uk-1, Uk, Uk+1) of a plurality of battery elements (Zk-1, Zk, Zk+1) of a battery (101), wherein one voltage (Uk) of the plurality of voltages (Uk-1, Uk, Uk+1) is unknown, the method comprising the following steps: determining a distribution of at least some of the plurality of voltages (Uk-1, Uk, Uk+1); and estimating the unknown voltage (Uk) based on the distribution. The invention further relates to a corresponding control unit and to a computer program.

Description

[0001] 202401534

[0002] 1

[0003] DESCRIPTION

[0004] Method for voltage monitoring of battery cells of a vehicle battery and corresponding control unit

[0005] TECHNICAL AREA

[0006] The present disclosure relates to methods and control devices for voltage monitoring of battery elements of a vehicle battery, for example battery cells.

[0007] BACKGROUND OF THE INVENTION

[0008] One of the main tasks of a battery management system, for example for a motor vehicle, is to monitor the individual battery cells within a battery pack. This involves using suitable measuring instruments to record and monitor various measurement parameters of the battery cells online, in particular the individual cell voltages and cell temperatures.

[0009] In a battery pack, individual battery cells are connected in series and / or parallel in various configurations to achieve different levels of pack voltage and capacity. Integrated hardware circuits are typically used to monitor cell voltage and temperature, thus providing a measurement-based representation of the cells' condition.

[0010] The cell voltage is recorded via a suitable measuring chain. Because each integrated hardware component has only a finite number of measuring channels, a specific number of battery cells, for example, twelve battery cells grouped into a block, are assigned to each hardware component. A galvanic connection is established from the positive terminal of each individual cell to the hardware component, and another from the negative terminal of the cell to the hardware component. The individual cell voltages are measured via these galvanic connections. Because the measuring chain is safety-critical, 202401534

[0011] 2

[0012] Diagnostic mechanisms are required to detect potentially occurring errors and initiate appropriate responses that bring the system into a safe state.

[0013] One possible diagnostic mechanism is the detection of problems in the measurement chain of the individual cell voltages. A possible fault pattern is the interruption of the galvanic connection between a cell terminal and the voltage measuring instrument, i.e., the integrated hardware component. In addition to interruptions, short circuits to ground or the battery can also occur. Faults of this kind result in the cell voltages of the individual cells being unknown, meaning they can no longer be measured and therefore can no longer be monitored.

[0014] Therefore, if there is a fault in the described cell voltage measurement chain, it can happen that safety precautions are initiated due to the unknown voltage, even though the battery cell in question is actually intact and functioning properly, but this cannot be determined due to the defective voltage measurement.

[0015] SUMMARY AND FORMS OF EXECUTION

[0016] It is therefore an objective of the present disclosure to provide a method and control unit for voltage monitoring of battery elements, in particular battery cells, which can estimate a voltage of a battery element that is unknown due to problems in the measuring chain.

[0017] This problem is solved by a method for monitoring a plurality of voltages of a plurality of battery elements of a battery, in particular a vehicle battery, as well as a control unit and a computer program according to the independent patent claims. Advantageous embodiments and further developments will become apparent from the respective dependent claims, the following description and the drawings. 202401534

[0018] 3

[0019] [1] Thus, according to a first aspect, a method for monitoring a plurality of voltages of a plurality of battery cells of a battery, in particular by means of a voltage monitoring circuit, is provided, wherein one voltage of the plurality of voltages is unknown. The method comprises the following steps: determining a distribution of at least a part of the plurality of voltages; and estimating the unknown voltage based on the distribution. The plurality of voltages can be a plurality of battery cell voltages, i.e., each representing the voltage of one battery cell of the plurality of battery cells. According to one embodiment, the voltages of the part of the plurality of voltages are known.

[0020]

[0014] According to another aspect, a control unit comprising a voltage monitoring circuit is provided which is configured to carry out the previously described method.

[0021] According to another aspect, a battery system is provided which includes the previously described control unit including voltage monitoring circuit, the battery and corresponding measuring connections between the voltage monitoring circuit and the battery.

[0022]

[0015] According to a further aspect, a computer program is provided which includes instructions that, when executed by a computer, cause it to carry out the method described above. In the context of the present disclosure, a computer is defined, for example, as a device that processes data by means of programmable computational instructions. Computers can be embedded in everyday devices, for example, in the control units of motor vehicles.

[0023] According to another aspect, a storage medium is provided with a computer program, wherein the computer program includes instructions which, when executed by a computer, cause the computer to perform the procedure described above. 202401534

[0024] 4

[0025] In the context of this disclosure, a battery is defined, for example, as a storage device for electrical energy, particularly on an electrochemical basis. In one embodiment, the battery is an accumulator, i.e., a rechargeable battery. The battery can contain several battery cells, which can be connected in series, at least partially, and in particular completely. The battery can contain several battery modules, wherein several battery cells are grouped together to form a battery module. Furthermore, the battery can contain several battery packs, wherein several battery modules are grouped together to form a battery pack. The battery is, for example, a lithium-ion accumulator. The battery can be a traction battery suitable for providing energy for a vehicle propulsion system. The battery can be a high-voltage battery.

[0026] In the context of this disclosure, a battery element is defined, for example, as a part of the battery comprising at least one battery cell. A battery element can be a single battery cell or an arrangement of a plurality of battery cells. The plurality of battery cells can, in particular, be connected in parallel. In the context of this disclosure, a battery cell can comprise an anode and a cathode, which are, in particular, separated by a conductive medium, for example, an electrolyte.

[0027] A battery element can also be or comprise a battery module or an arrangement of multiple battery modules and / or a battery pack or an arrangement of multiple battery packs. A battery module can comprise multiple battery cells, which can be connected in series and / or parallel. Within a battery module, the battery cells can be grouped in such a way that it is modularly replaceable. Similarly, a battery pack can comprise multiple battery modules.

[0028] In the context of the present disclosure, adjacent battery elements are defined, for example, as those battery elements that are directly electrically connected to one another, that is, without other battery cells being interposed. According to one embodiment, adjacent 202401534

[0029] 5

[0030] Battery elements connected in series. For example, two battery cells directly connected in series are adjacent to each other.

[0031] In the context of this disclosure, a measuring connection is defined, for example, as an electrical connection between the voltage monitoring circuit and one of the poles of a battery element, such as a battery cell. The measuring connection can connect the voltage monitoring circuit to opposite poles of adjacent battery elements. Then, one and the same measuring connection can be used to measure the voltages of two adjacent battery elements.

[0032] In the context of this disclosure, a voltage is defined as unknown, for example, if a problem has occurred in the associated measuring chain. Such a problem can lead to either no voltage being measured or an incorrect voltage being measured. For example, a measuring lead may have failed, or there may be a short circuit between the measuring lead and ground or the battery.

[0033] In the context of this disclosure, a distribution is defined, for example, as a frequency distribution, a statistical distribution, or a probability distribution. From the distribution, a frequency or probability can be derived that the unknown voltage takes on a certain value or lies within a certain range of values. The distribution can be discrete or continuous.

[0034] In the context of the present disclosure, a battery management system or BMS is defined, for example, as a component associated with a battery which performs at least one of the following functions:

[0035] Monitoring, control, and protection of the battery. For example, the battery management system can implement voltage diagnostics, especially cell voltage diagnostics, charge and discharge control, state-of-charge detection, temperature control, deep discharge protection, and / or overcharge protection. The battery management system can also be used for balancing or symmetrizing the different cells.

[0036] 6

[0037] be set up, in particular to ensure a more even distribution of electrical charge in different battery cells.

[0038] In multi-cell batteries, the battery management system may be configured to monitor and / or control individual cells. For this purpose, the battery management system may include a voltage monitoring circuit. This circuit may be configured to measure all or some of the voltages using appropriate measuring connections, particularly by direct measurement. Alternatively, the circuit may be configured to measure the voltages indirectly, i.e., to determine them from other measured quantities. The voltage monitoring circuit may be, or include, a battery monitoring IC.

[0039] In the context of this disclosure, estimating a voltage is defined, for example, by specifying a minimum probability that the voltage lies within a predetermined range of values. The range of values ​​may be defined by a lower and / or an upper voltage limit. In other words, it is stated that the voltage lies within the predetermined range of values ​​with a probability at least equal to the minimum probability.

[0040] The previously described method and / or control unit can be advantageous because, in the event of problems in the voltage measurement chain for a single battery cell, the voltage of that cell remains unknown. For safety reasons and to increase battery lifespan, manufacturers typically specify voltage limits within which the batteries may be operated, for example, an upper cell voltage limit for charging and a lower cell voltage limit for discharging. Such specifications prevent deep discharge or overcharging of the battery cells. However, if the voltage of at least one battery cell is unknown, and therefore it is uncertain whether this voltage meets the specified requirements, then...202401534

[0041] 7

[0042] Appropriate safety measures must be taken; in the worst case, the battery must be disconnected from the consumers.

[0043] If, however, the unknown voltage is estimated based on a distribution of known voltages, it can be determined, for example, whether the safety requirements are met with at least a given probability, and no or only less severe safety measures need to be initiated. This increases the operational safety as well as the operational availability of the vehicle in the event of a fault in the measurement chain for determining the battery cell voltages. If the distribution is determined from current voltage values, the current state of the battery can also be taken into account, for example, with regard to aging or current draw.

[0044] Power feed-in.

[0045] According to one embodiment, the voltage is unknown because at least one of the following faults is present: a measuring connection to the voltage monitoring circuit is faulty; a short circuit to ground and / or battery has occurred.

[0046] [2] According to one embodiment, the method further comprises: determining at least one of the following statistical quantities relating to the distribution: a mean, a median, a variance, a standard deviation; and estimating the unknown tension taking into account the at least one statistical quantity.

[0047] [3] According to one embodiment, estimating the unknown voltage comprises: estimating whether the unknown voltage lies within a predetermined range of values ​​with at least a predetermined probability. The standard deviation of the distribution can be taken into account. The predetermined range of values ​​can be determined by a lower and / or an upper limit. The limits can be set due to safety requirements, for example to prevent deep discharge.

[0048] 8

[0049] or to prevent overcharging of the battery cells. The limit values ​​can depend, for example, on the cell type and / or the cell chemistry of the battery.

[0050] Such an embodiment can be advantageous when so-called "out of range" (OOR) monitoring of the battery units is required for safety reasons. This allows the voltages of the battery cells to be kept within the regular operating voltage range, for example, a cell voltage range of 1.5 V to 4.5 V. The embodiment described above can also enable OOR monitoring for unknown voltages of battery cells. For this purpose, a range can be determined using the standard deviation, within which the unknown voltage lies with a specified probability. Within a predefined range of values ​​that includes the aforementioned range, the unknown voltage then lies with at least the specified probability.

[0051] According to one embodiment, the predetermined probability is determined by the standard deviation, in particular by a product of the standard deviation and a factor. If, for example, the distribution is at least approximately a normal distribution, then the unknown voltage lies with a probability of about 68 percent within a range given by the mean of the distribution plus / minus one standard deviation, with a probability of about 95 percent within a range given by the mean of the distribution plus / minus two standard deviations, and with a probability of about 99.7 percent within a range given by the mean of the distribution plus / minus three standard deviations.Therefore, starting from a probability specified, for example due to safety requirements, the aforementioned factor can be determined, which then defines the range around the mean of the distribution, which must lie within the specified range of values.

[0052] [4] According to one embodiment, at least some of the multiple voltages and the unknown voltage are assigned to battery elements of the same battery unit, in particular the same battery block. Eine202401534

[0053] 9

[0054] Such an embodiment can be advantageous because the battery elements of a battery unit all have the same or a similar state of aging, for example, if some battery units of the battery have been replaced and others have not.

[0055] Alternatively, all known voltages of the battery cells are considered to determine the distribution. Such an embodiment can be advantageous because the distribution is based on more data compared to the previous embodiment, which considers less, but more reliable, data.

[0056] According to one embodiment, the battery unit is at least one of a battery block, a battery module and a battery pack, each of which combines a predetermined number of battery elements, in particular of adjacent battery elements.

[0057] [5] According to one embodiment, a further voltage of the plurality of voltages is unknown. The unknown voltage and the further unknown voltage can be assigned to adjacent battery cells and / or battery cells that have a common measuring lead that contacts a negative terminal of one battery cell and a positive terminal of the other battery cell. In case of problems with this common measuring lead, the voltages of both adjacent battery cells can be unknown.

[0058] According to one embodiment, the procedure further includes: estimating the remaining unknown voltage based on the distribution.

[0059] [6] According to one embodiment, estimating the unknown voltage involves: determining a further distribution for the unknown voltage, taking into account an additional condition that causes the further distribution to differ from the distribution. The further unknown voltage can also be distributed according to the further distribution. Such an embodiment can be advantageous, in particular if the further distribution is narrower than the distribution due to the additional condition, i.e., the 202401534

[0060] 10

[0061] Possible voltage values ​​will vary less widely around the mean. A narrower distribution can improve vehicle availability in the event of a fault.

[0062] [7] According to one embodiment, the additional condition is a known sum of the unknown voltage and the other unknown voltage. The sum can, for example, be derived from a known voltage of a battery unit, such as a block voltage or a module voltage, if the battery unit contains the battery elements associated with the unknown voltage and the other unknown voltage. The sum can also be derived from a known voltage of the entire battery. When the known sum is taken into account, the distribution may change compared to when the sum is unknown.

[0063] In particular, the distribution may narrow.

[0064] [8] According to one embodiment, a battery unit of the battery comprises a predetermined number of the plurality of battery elements, including the battery elements associated with the unknown voltage and the further unknown voltage, wherein the known sum of the battery unit voltages is determined. The battery unit can be a battery block, a battery module, or a battery pack. The battery unit voltage can be measured by means of a separate measuring channel of the voltage monitoring circuit.

[0065] According to one embodiment, the two battery elements with the unknown voltage and with the other unknown voltage are adjacent, and the known sum is determined by means of two measuring lines, one of which is assigned to each of the two adjacent battery units.

[0066] For example, the two voltages may be unknown due to a faulty common measuring lead of adjacent battery cells. If the potential of the common measuring lead is still within a measurable range, the known sum can be calculated from the sum of the two measured voltages, even though both voltages are faulty. 202401534

[0067] 11

[0068] [9] According to one embodiment, the estimation comprises: determining whether the unknown voltage is above a predetermined upper limit and / or below a predetermined lower limit, in particular with at least one predetermined probability in each case, and / or determining whether the further unknown voltage is above the predetermined upper limit and / or below the predetermined lower limit, in particular with at least the predetermined probability in each case. The limits may depend on a cell type and / or cell chemistry of the battery. Such an embodiment may be advantageous due to the aforementioned safety-related importance of out-of-range monitoring.

[0069]

[0010] According to one embodiment, the estimation takes into account a mean value of the further distribution, wherein the mean value of the further distribution is determined from the known sum. The mean value of the distribution can, in particular, be determined by or equal to half of the sum. In some cases, considering the mean value alone may be sufficient to investigate, for example, whether at least one of the two voltages lies above the specified upper limit or below the specified lower limit. This is the case, for example, if the mean value lies above the upper limit or below the lower limit.

[0070]

[0011] According to one embodiment, the estimation takes into account a confidence value, in particular a standard deviation, of the further distribution. According to one embodiment, the confidence value correlates with the aforementioned predetermined probability. In the context of the present disclosure, a confidence value of a distribution is defined, for example, as a feature of the distribution that characterizes a spread of the distribution. The confidence value can be a product of the standard deviation and a factor. Because the factor correlates with the aforementioned predetermined probability, a level of certainty can be specified via the factor. However, the confidence value can also depend on the standard deviation in another functional way.

[0071] 12

[0072]

[0012] According to one embodiment, the confidence value of the further distribution differs from a corresponding confidence value of the distribution by a factor, in particular 2. The confidence values ​​can correspond to the standard deviations of the respective distributions, a product of the standard deviation and a further factor, or any function of the standard deviation.

[0073]

[0013] According to one embodiment, the method further comprises: setting an error variable if the mean of the further distribution plus the confidence value is greater than the predetermined upper limit and / or if the mean of the further distribution minus the confidence value is less than the predetermined lower limit. Such an embodiment can be advantageous to ensure that the unknown voltage and / or the further unknown voltage is less than the predetermined upper limit and / or greater than the predetermined lower limit with at least a predetermined probability that correlates with the confidence value.

[0074] According to one embodiment, the method further includes: when the fault variable is set, performing at least one of the following measures: limiting the power consumption of a consumer, disconnecting the consumer from the battery, galvanically isolating the battery, activating a vehicle emergency mode, or displaying an acoustic, visual, or haptic warning signal.

[0075] According to one embodiment, the method further features: if the error variable is not set: displaying a warning that is indicative of at least one faulty voltage measurement.

[0076] BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Further advantages, beneficial designs, and further developments of the method and the control unit result from the following exemplary embodiments illustrated in conjunction with the figures. 202401534

[0078] 13

[0079] They show:

[0080] Figure 1 shows a battery system with a battery and a voltage monitoring circuit according to an embodiment of the present disclosure;

[0081] Identical, similar, or similarly effective elements are marked with the same reference symbols in the figures. In some figures, individual reference symbols have been omitted for clarity. The figures and the relative sizes of the elements depicted within them are not to be considered to scale. Rather, individual elements may be exaggerated for better representation and / or comprehensibility.

[0082] DETAILED DESCRIPTION OF EXAMPLES OF EXECUTION

[0083] Figure 1 shows a battery system 100 of a motor vehicle, comprising the following elements: (i) a voltage monitoring circuit 102, which may, for example, be part of a battery management system, (ii) a rechargeable battery 101 with at least two battery elements, here battery cells Zk-1, Zk, Zk+1, and (iii) at least two measuring lines Wk-1, Wk, Wk+1, Wk+2, Wm1, Wm2 between the voltage monitoring circuit 102 and the battery 101.

[0084] A first measuring lead Wk connects the voltage monitoring circuit 102 to the positive terminal of battery cell Zk-1 and to the negative terminal of the adjacent battery cell Zk. A second measuring lead Wk+1 connects the voltage monitoring circuit 102 to the positive terminal of battery cell Zk and to the negative terminal of the adjacent battery cell Zk+1. The voltage monitoring circuit 102 has one or more measuring devices, for example, several so-called battery monitoring ICs, with which the cell voltages Uk of cells Zk can be determined using the respective measuring leads Wk and Wk+1 and associated measuring channels.

[0085] 14

[0086] A predetermined number of battery cells Zk, for example twelve adjacent battery cells Zk-1, Zk, Zk+1, are grouped together to form a battery unit, here a battery block Mm. The total voltage Um of the battery block Mm can be determined via measuring connections Wm1, Wm2 to the edge cells of the battery block Mm. The total voltage Um, for example across the aforementioned 12 series cells, can be determined via a separate measuring channel from the cell voltage measurement channels. Thus, in the described example, twelve individual cell voltages, Uk [V], with k = 1 ... 12, are available, as well as a total block voltage Um [V] across all individual cell voltages in the block.

[0087] The voltage monitoring circuit 102 is configured to perform a method for monitoring a plurality of voltages Uk-1, Uk, Uk+1 of a plurality of battery cells Zk-1, Zk, Zk+1 of a battery 101, wherein at least one voltage Uk of the plurality of voltages Uk-1, Uk, Uk+1 is unknown. The method comprises the following steps: (i) determining a statistical distribution of at least some of the plurality of voltages Uk-1, Uk, Uk+1, which are known; and estimating the at least one unknown voltage Uk based on the distribution.

[0088] An important monitoring function for cell voltages Uk-1, Uk, and Uk+1 is the "out of range (OOR)" monitoring. This monitors whether the cell voltages Uk-1, Uk, and Uk+1 are maintained within the normal operating voltage range. The value of this voltage range depends on the cell type and cell chemistry. For simplicity, a cell voltage range of 1.5V to 4.5V is assumed here as an example. If the cell voltage measurement leaves this voltage range, appropriate measures must be taken to prevent deep discharge or overcharging. If a measuring chain to a cell Zk-1, Zk, or Zk+1 fails due to a fault (103), the failed cell voltage Uk-1, Uk, or Uk+1 can no longer be monitored with respect to the OOR thresholds. In Figure 1, the cross represents such a fault 103 in the measuring connection Wk, which leads to a failure of the measuring chain to the two adjacent cells Zk-1 and Zk. Accordingly, from the 202401534

[0089] 15

[0090] Voltage monitoring circuit 102: no or faulty voltages llk-1 , llk of the two cells Zk-1 , Zk measured.

[0091] Due to the circuitry described above in the measuring chain between the individual cells Zk-1, Zk, Zk+1 in a block Mm and the hardware measuring instrument 102 (e.g. integrated hardware module), either only one cell voltage of a block edge cell can fail, or, as shown in Figure 1, the cell voltages llk-1 , llk of two adjacent cells Zk-1 , Zk in the block Mm can fail.

[0092] The following section deals with emulating cell voltages llk-1, llk, which can no longer be detected and therefore no longer monitored due to a fault 103 occurring in the measurement chain. If a corresponding fault 103 occurs in the vehicle's measurement chain, resulting in the cell voltages llk-1, llk in block Mm and in the battery pack no longer being detected, the system must be brought to a safe state. A safe state is, for example, opening the contactors and thus galvanically isolating the energy source (battery) from the vehicle's power supply. The vehicle can then no longer be operated, resulting in a breakdown. However, the immediate termination of the current driving cycle can lead to an unsafe situation at the vehicle level.Therefore, the aim is to ensure that, through appropriate measures, the vehicle can still complete the current driving cycle, even with limitations such as power restriction, if a fault 103 occurs as described above. However, suitable auxiliary measures must be in place to ensure that the cell voltages llk-1, llk of the affected cells Zk-1, Zk, which are affected by fault 103, can still be monitored with respect to OOR (Out of Range). How can this be achieved if fault 103 prevents direct cell voltage measurement?

[0093] This is where the voltage measurement of the block voltage Um of the corresponding block Mm comes into play. This additional information allows an equation to be formulated that can be used to determine the sum of the failed cell voltages Zk-1, Zk, or, in the case of only one affected edge cell in block Mm, to directly calculate the cell voltage of that edge cell. Kann202401534

[0094] 16

[0095] If the equation cannot be solved using a closed-form formula because it contains two unknown cell voltages, for example, in the case of affected adjacent center cells Zk-1 and Zk of block Mm, a second equation is needed to determine the two missing cell voltages llk-1 and llk. This aspect will now be given more detailed attention.

[0096] A plausible assumption for the second equation is that the two cell voltages llk-1, llk have similar values. After all, there are suitable mechanisms to balance or equalize the potentials of the individual cell voltages llk-1, llk, llk+1 so that they exhibit the same potential as closely as possible. However, since this balancing cannot be perfect and the aging of cells Zk-1, Zk, Zk+1, as well as the current draw depending on the internal resistance of each cell Zk-1, Zk, Zk+1, can affect the actual cell imbalance, the second equation must account for a possible and, ideally, realistic cell imbalance. In this context, information on all measured cell voltages llk-1, llk, llk+1 in the battery pack becomes crucial. For example, such a battery pack could contain 196 individual cells, of which, for instance,Two neighboring cells Zk-1, Zk have failed with regard to cell voltage measurement.

[0097] The 196-2 cells can now be statistically evaluated to arrive at the most valid assumption possible for the second equation. The statistical distribution of the measured 194 cell voltages is now of crucial importance. For example, a mathematical distribution function can be approximated using the 194 cell voltages with respect to the existing variance of the measured cell voltages under defined conditions, e.g., as a function of the current drawn from the battery pack. Using an approximated continuous distribution function, such as a normal distribution, and taking the standard deviation into account, a statement can now be made about the expected value of the minimum and maximum cell voltages of the two failed cell voltages, llk-1 and llk. This is important because, due to functional safety requirements, the failed cell voltages llk-1 and llk must not be underestimated or overestimated under any circumstances.Der202401534.

[0098] 17

[0099] The advantage of considering all measured cell voltages for the second equation is that it eliminates the need for a worst-case assumption across all produced battery packs and the maximum possible aging over their lifetime. This leads to significantly improved vehicle availability in the event of such a fault (103).

[0100] This results in the following system of equations (“center cell affected”):

[0101] I: Um = ZUi + Uk-1 +Uk

[0102] II: Uk-1, k_min = 0.5

[0103] Uk-1 ,k_max = 0,5

[0104]

[0105] With: To == block voltage in V

[0106] Ui == measured individual cell voltages in V in the block; with e.g. i=1..12, excluding k-1, k

[0107] Uk == failed cell voltage in the block in V

[0108] Uk-1 == failed cell voltage in the block in V

[0109] Uk-1,k_min == statistically minimum failed cell voltage in V

[0110] Uk-1,k_max == statistically maximum failed cell voltage in V x 6 == magnitude of standard deviation from normal distribution in V with: xe N The factor x, by which the determined standard deviation is multiplied and which results from the confidence level, determines the safety level; so e.g. x=3 for 99.7% safety level and thus 0.31% probability of failure.

[0111] The following is a brief derivation of equation II. If the cell voltages Uk-1, Uk, Uk+1 follow a normal distribution with a standard deviation of 6, then the difference Udelta = Uk-1 - Uk between two cell voltages Uk-1 and Uk follows a normal distribution with an expected value of 0 and a standard deviation θUdelta, which results from the standard deviations 6Uk-1 and θUk for the individual cell voltages Uk-1 and Uk:

[0112]

[0113] >

[0114] This allows the maximum and minimum deviation Udelta_max / min for a given confidence level to be estimated as + / - x * 2 * 6. Since the mean is known as 0.5 * (Uk-1 + Uk), the maximum and minimum values ​​for Uk-1 and Uk, respectively, are then given as:

[0115] Uk-1 ,k_max = 0.5 * (Uk-1 + Uk) + Udelta_max / 2 =202401534

[0116] 18

[0117] = 0.5 * (Uk-1 + Uk) + x * 2 * 6 / 2 = 0.5 * (Uk-1 + Uk) + x * 6 H2 Uk-1 ,k_min = 0.5 * (Uk-1 + Uk) + Udelta_min / 2 =

[0118] = 0.5 * (Uk-1 + Uk) - x * 2 * 6 / 2 = 0.5 * (Uk-1 + Uk) - x * 6 H2 By using the known mean 0.5 * (Uk-1 + Uk), the necessary tolerance can be reduced by a factor of 1 / 2 for a given confidence level compared to an estimate without a known mean (see below).

[0119] Alternatively, without the possibility of explicitly measuring the block voltage Mm and therefore without a known mean value, the following equations can be formulated:

[0120] Uk-1 ,k_min = p - x 6

[0121] Uk-1 ,k_max = p + x 6

[0122] With: p == expected cell voltage in V of all measured cell voltages Uk-1,k_min == statistically minimum failed cell voltage in V

[0123] Uk-1,k_max == statistically maximum failed cell voltage in V x 6 == magnitude standard deviation from normal distribution in V with: xe N The expected value p and the standard deviation 6 can be calculated in real time in the control unit. The calculated minimum and maximum reconstructed cell voltage values ​​can be monitored against suitable "OOR voltage thresholds" to continuously prevent impermissible discharge (Uk-1,k_min < 00R_min, e.g., 1.5V) or impermissible overcharging (Uk-1,k_max > 00R_max, e.g., 4.5V) of the two unmeasurable cell voltages Uk-1 and Uk. This enables a vehicle emergency mode, for example, to allow the current driving cycle to be completed normally.

[0124] Alternatively, the calculated standard deviation 6 can not be taken into account in the reconstructed cell voltages Uk-1, Uk, but can be included in the OOR thresholds.

[0125] Furthermore, the statistical analysis of cell voltages can be performed at various levels, for example, at the block level, the module level, or the battery pack level. This allows for appropriate responses to different use cases, such as when a module is replaced in an aged battery pack. The statistical information at these different levels can then be used to respond appropriately to the new situation in the event of a fault. 202401534

[0126] 19

[0127] For example, the statistical distribution for estimating unknown voltages llk-1, llk can be based solely on the cell voltage measurements of a single module or block, not on those of the entire battery pack. While this results in less data, it may be more reliable.

[0128] Another feature is the continuous evaluation of the statistical values ​​of the individual cell voltages Uk-1, Uk, and Uk+1 within a block Mm or the entire battery pack throughout its lifetime. This provides a continuous supply of statistical data that can be used in the event of a failure of one or more measurement parameters. This allows for probabilities-based predictions about the modeled parameters, such as cell voltage, cell temperature, or any other measured value. For example, even without measuring the block voltage, the median or mean value of a block or the battery pack can be used with the standard deviation before a failure occurs to make statistical predictions about the magnitude of the missing measurement.This is not limited to distribution functions such as a normal distribution; other methods, such as a neural network, can also be used. For example, it could be adaptively determined what factor z is necessary for a cell voltage to lie within the interval + / - z * 6 around the mean p of the other cells. In the event of cell voltage failure, a reliable estimate could thus be made without applying an unnecessarily large safety factor.

[0129] The invention is not limited to the exemplary embodiments described therein. Rather, the invention encompasses every new feature as well as every combination of features, which in particular includes every combination of features in the exemplary embodiments and claims. 202401534

[0130] 20

[0131] REFERENCE MARK

[0132] 100 battery system

[0133] 101 Battery

[0134] 102 Voltage monitoring circuit

[0135] 103 errors

[0136] Zk Battery element / battery cell

[0137] Wk-i intact measuring lead to the negative terminal of the battery cell Zk-i Wk defective measuring lead to the negative terminal of the battery cell Zk Wk+i intact measuring lead to the negative terminal of the battery cell Zk+i M m Battery unit / battery block

[0138] W mi Measuring lead to the positive terminal of the battery unit M m

[0139] Wm2 measuring lead to the negative terminal of the battery unit M m

[0140] Uk+i known voltage of the battery cell Zk+i

[0141] Uk unknown voltage of battery cell Zk

[0142] Uk-i further unknown voltage of the battery cell Zk-i

[0143] To voltage of the battery unit M m

Claims

202401534 21 PATENT CLAIMS 1. Method for monitoring a plurality of voltages (llk-1 , llk, llk+1) of a plurality of battery elements (Zk-1 , Zk, Zk+1) of a battery (101), wherein one voltage (llk) of the plurality of voltages (Uk-1 , llk, llk+1) is unknown, the method comprising the following steps: Determining a distribution of at least a part of the majority of voltages (Uk-1, llk, llk+1); and Estimating the unknown voltage (llk) based on the distribution.

2. Method according to the preceding claim, further comprising: Determine at least one of the following distribution-related statistical measures: a mean, a median, a variance, a standard deviation; and Estimating the unknown voltage (llk) taking into account at least one statistical quantity.

3. Method according to the preceding claim, wherein the estimation of the unknown voltage (llk) comprises: estimating whether the unknown voltage (llk) lies within a predetermined range of values ​​with at least a predetermined probability, taking into account the standard deviation of the distribution.

4. Method according to one of the preceding claims, wherein at least a part of the plurality of voltages (Uk-1 , Uk, Uk+1 ) and the unknown voltage (Uk) are assigned to battery elements (Zk-1, Zk, Zk+1) of the same battery unit (Mm), in particular of the same battery block.

5. Method according to one of the preceding claims, wherein a further voltage (Uk-1) of the plurality of voltages (Uk-1, Uk, Uk+1) is unknown.

6. Method according to the preceding claim, wherein the estimation of the unknown voltage (Uk) comprises: Determining a further distribution 202401534 22 for the unknown voltage (llk), taking into account an additional condition which results in a different distribution from the distribution.

7. Method according to the preceding claim, wherein the additional condition is a known sum of the unknown voltage (llk) and the further unknown voltage (llk-1).

8. Method according to the preceding claim, wherein a battery unit (Mm) of the battery (101) comprises a predetermined number of the plurality of battery elements (Zk-1 , Zk, Zk+1) including the battery elements (Zk-1, Zk) belonging to the unknown voltage (llk) and the further unknown voltage (llk-1), wherein the known sum is determined from a battery unit voltage (Um) of the battery unit (Mm).

9. Method according to any one of claims 5 to 8, wherein the estimation comprises: determining whether the unknown voltage (Uk) and / or the further unknown voltage (Uk-1) is above a predetermined upper limit and / or below a predetermined lower limit.

10. Method according to the preceding claim, wherein the estimation takes into account a mean value of the further distribution, wherein the mean value of the further distribution is determined from the known sum.

11. Method according to the preceding claim, wherein the estimation takes into account a confidence value, in particular a standard deviation, of the further distribution.

12. Method according to the preceding claim, wherein the confidence value of the further distribution differs from a corresponding confidence value of the distribution by a factor, in particular 2.

13. Method according to claim 11 or 12, the method further comprising: setting an error variable if the mean of the further distribution plus the 202401534 23 The confidence value is greater than the specified upper limit and / or if the mean of the further distribution minus the confidence value is less than the specified lower limit.

14. Control unit comprising a voltage monitoring circuit which is configured to perform a method according to any one of claims 1 to 13.

15. Computer program comprising instructions which, when executed by a computer, cause the computer to perform the method according to any one of claims 1 to 13.

Images