Calculation of the battery charge level with reduced load on the central processing unit and reduced memory usage

Abstract

Vehicle which includes the following: a battery pack (14) with cells grouped into several subgroups, and at least one control device programmed to derive the state of charge of the battery pack (14) from an initial state of charge of each of the cells at vehicle activation and from an electrical charge accumulated or consumed by at least one representative cell from each of the subgroups, but by fewer than all cells from each of the subgroups, since vehicle activation, and to charge and discharge the battery pack (14) based on the derived state of charge.

Description

[0001] This disclosure concerns provisions regarding the state of charge of a battery pack.

[0002] Modern hybrid and electric vehicles use battery packs to provide energy for propulsion and to store regenerated energy. Battery packs typically consist of multiple individual battery cells that can be connected in parallel, in series, or in a combination of these configurations. One characteristic of the battery pack is its state of charge (SOC). The battery pack SOC is a measure of the fraction of the total charge remaining in the battery pack and can be viewed analogously to a fuel gauge. The SOC can be important for determining when and how to charge or discharge the battery pack. The SOC can also be important for providing the driver with information about the vehicle's driving range and for operating the vehicle.

[0003] A method for voltage balancing of a battery pack consisting of several battery cells is known from US patent 2011 / 0025258 A1. For this purpose, the charge states of all cells are determined beforehand, and the individual cells are then divided into different groups depending on their charge state in order to distinguish between cells to be charged and those to be discharged. A diagnostic method for monitoring the state of a battery pack is known from German patent DE 10 2006 033 629 A1. For this, the available cells are divided into groups, and a diagnosis is performed only temporarily for a subset of the groups. Methods for determining the state of charge of a battery are also known from German patent DE 10 2010 030 491 A1 and German patent DE 10 2010 039 915 A1. Here, too, the cells are grouped, and only the cells of a limited number of groups are considered for determining the state of charge. Another method is known from DE 10 2005 029 096 A1.

[0004] A vehicle is disclosed that includes a battery pack with cells grouped into subgroups and at least one control unit programmed to charge and discharge the battery pack. Charging and discharging the battery pack is based on a pack state of charge (SOC) determined by the initial state of charge of each cell after vehicle activation and the electrical charge accumulated or consumed by less than all cells of each subgroup since vehicle activation. The cells of each subgroup have a specified set of cell parameter values ​​that fall within a specified range of values, with each subgroup having a different range of values. The cell parameter values ​​may include a specified range of temperatures, initial cell states of charge or capacities, or cell ages.The cells can be grouped into subgroups, so that the cells can be scattered throughout the battery pack.

[0005] A method for controlling a vehicle comprising a battery pack is disclosed. The steps include identifying subgroups of cells with similar parameter values ​​and calculating the electrical charge accumulated or consumed for fewer than all cells in each subgroup. The similar parameter values ​​may represent a specified range of temperatures, initial cell states of charge, or cell ages. According to one possible embodiment, the accumulated or consumed electrical charge for a cell in each subgroup can be calculated. An average electrical charge accumulated or consumed for the pack is calculated based on the electrical charge accumulated or consumed for each subgroup. A cell state of charge for each cell of the pack is calculated based on the average accumulated or consumed charge.The state of charge of the battery pack is calculated based on the state of charge of all cells. A control output for the vehicle is generated based on the state of charge of the battery pack. The step of identifying subgroups can be performed after vehicle activation or during a vehicle driving cycle.

[0006] A method for estimating the state of charge of a battery pack is disclosed, in which the cells of the battery pack are grouped into subgroups. The state of charge of the battery pack is calculated based on an initial state of charge after vehicle activation for each cell and an electrical charge accumulated or consumed since vehicle activation for less than all cells in each of the subgroups. The state of charge of the battery pack is used to output information for vehicle control or display. The grouping into subgroups can be based on temperature, initial state of charge, capacity, and cell age. Fig. Figure 1 is a diagram of a hybrid electric vehicle showing typical drive train and energy storage components. Fig. Figure 2 is a diagram of a possible battery pack arrangement consisting of multiple cells and monitored and controlled by a battery control module. Fig. Figure 3 is a diagram of a battery cell equivalent circuit serving as an example. Fig. Figure 4 is a diagram of a sample cell grouping based on temperature and cell charge level. Fig. Figure 5 is a graph of a possible relationship between an open-circuit voltage (Voc) and a battery state of charge (SOC) for a typical battery cell. Fig. Figure 6 is a flowchart of an algorithm for determining the battery SOC.

[0007] Fig. Figure 1 shows a typical hybrid electric vehicle. A typical hybrid electric vehicle can include one or more electric motors 4 mechanically connected to a hybrid transmission 3. Additionally, the hybrid transmission 6 is mechanically connected to an internal combustion engine 8. The hybrid transmission 6 is also mechanically connected to a drive shaft 10, which is mechanically connected to the wheels 12. The electric motors 4 can provide propulsion and deceleration capabilities when the internal combustion engine 8 is switched on or off. The electric motors 4 also act as generators and can provide fuel-saving benefits. by recovering energy that would otherwise be lost as heat in the friction braking system. The electric motors 4 can also provide reduced pollutant emissions because the hybrid electric vehicle 2 can be operated in electric mode under certain conditions.

[0008] The battery pack 14 stores energy that can be used by the electric motors 4. A vehicle battery pack 14 typically provides a high DC output voltage. The battery pack 14 is electrically connected to the power electronics module 16. The power electronics module 16 is also electrically connected to the electric motors 4 and provides the capability for bidirectional energy transfer between the battery pack 14 and the electric motors 4. For example, a typical battery pack 14 can provide a DC voltage, while the electric motors 4 require three-phase AC to operate. The power electronics module 16 can convert the DC voltage into the three-phase AC required by the electric motors 4.In regenerative mode, the power electronics module 16 converts the three-phase alternating current from the electric motors 4, which act as generators, into the direct current required by the battery pack 14. The method described here is equally applicable to a purely electric vehicle or any other device that uses a battery pack.

[0009] In addition to providing energy for propulsion, the battery pack 14 can supply energy to other electrical systems of the vehicle. A typical system may include a DC / DC converter module 18, which converts the high DC output voltage of the battery pack 14 into a low DC supply compatible with other vehicle loads. Other high-voltage loads can be connected directly without the use of a DC / DC converter module 18. In a typical vehicle, the low-voltage systems are electrically connected to a 12 V battery 20.

[0010] Battery packs can be made from a variety of chemical formulations. Typical battery pack chemistries include lead-acid, nickel-metal hydride (NiMH), or lithium-ion. Fig. Figure 2 shows a typical battery pack 30 in a simple series configuration of N battery cells 32. However, other battery packs can be composed of any number of individual battery cells connected in series, parallel, or a combination thereof. A typical system may include one or more control devices such as a battery control module (BCM) 36, which monitors and controls the operation of the battery pack 30. The BCM 36 can monitor several battery pack-level characteristics such as the pack current 38, the pack voltage 40, and the pack temperature 42.

[0011] In addition to the battery pack-level properties, there may be battery cell-level properties that need to be measured and monitored. For example, the terminal voltage, current, and temperature of each cell can be measured. A system can use a sensor module 34 to measure the battery cell properties. Depending on its capabilities, the sensor module 34 can measure the properties of one or more of the battery cells 32. The battery pack 30 can hold up to N c Sensor modules 34 are used to measure the properties of all battery cells 32. Each sensor module 34 can transmit the measurements to the BCM 36 for further processing and coordination. The sensor module 34 can transmit signals to the BCM 36 in analog or digital form.

[0012] An important measurement of the battery system can be the state of charge (SOC) of the battery pack. The battery pack SOC indicates how much charge remains in the battery pack. The battery pack SOC can be displayed to inform the driver, similar to a fuel gauge, about the remaining charge in the battery pack. The battery pack SOC can also be used to control the operation of an electric vehicle or hybrid electric vehicle. The battery pack SOC can be calculated using a variety of methods. One possible method is to perform an integration of the battery pack current over time. This is well known in the field as ampere-hour integration. A potential drawback of this method is that the current measurement can be noisy.A possible inaccuracy in the charge level can occur as a result of integrating this noisy signal over time.

[0013] Some modern state-of-charge (SOC) estimation methods use model-based techniques, such as Kalman filtering, to determine a more accurate SOC. A model-based method works by using a model of the battery cell and then predicting the internal states of the battery cell based on some actually measured values. The estimated internal states can include, but are not limited to, voltages, currents, or SOC. A typical approach is to apply a Kalman filter to each cell in the battery pack and then use these cell values ​​to calculate the overall pack characteristics. This requires the control unit to execute a number of Kalman filters equal to the number of cells in the battery pack. The number of cells in a battery pack varies, but a modern vehicle battery pack can consist of 80 or more cells.

[0014] Fig. Figure 3 shows a typical battery cell equivalent circuit model. A battery cell can be considered a voltage source (V oc ) 50 is modeled, to which resistors (52 and 54) and a capacitance 56 are assigned. Due to the battery cell impedance, the terminal voltage V 58 is typically not equal to the open-circuit voltage V oc 50. The open-circuit voltage V oc 50 is not easily measurable because only the terminal voltage 58 of the battery cell is accessible for measurement. Because the voltage V oc Since the resistance (50) is not easily measurable, a model-based method can be used to estimate the value. For any model, the resistance and capacity values ​​must be known or estimated. It should be noted that the battery cell model can depend on the battery chemistry. The exact model chosen for the battery cell is not crucial for the described method.

[0015] Using Kirchhoff's laws for current and voltage and well-known properties of circuit elements, the equivalent circuit of Fig. 3 can be expressed by the following equations: Voc(t)=R∗I(t)+Vc(t)+V(t) C∗dVc(t)dt=I(t)−Vc(t)Rc

[0016] Using the following relationships dSOCdt=ηIQ and dVOCdt=dVOCdSOC∗dSOCdt, can the derivative of V oc can be expressed as follows: dVOC(t)dt=−dVOCdSOC∗ηIQ where η the charge / discharge efficiency is and Q is the cell charge capacity. I is the current flowing into and out of the battery. In this case, the discharge current (the current flowing out of the battery) is considered positive.

[0017] Combining equations (1) to (3) yields the following: [dVOCdtdVCdt]=[000−1c∗RC]∗[VocVc(t)]+[−dVocdSOC∗ηQ 1C]∗I V(t)=[1 −1]∗[Voc(t)Vc(t)]+[−R]∗I

[0018] An observer for equations (4) and (5) can be expressed as follows: [dV^ocdt dV˜cdt]=[000−1C∗Rc]∗ [V^ocV^c(t)]+[−dV^ocdSOC∗ηQ 1C]∗I+L∗(V(t)−V^(t)) V^(t)=[1 −1]∗[V^oc(t) V^c(t)]+[−R]∗I where V(t) is the measured cell connection voltage, V̂(t) is an estimate of the cell terminal voltage V̂ oc . an estimate of the cell open-circuit voltage is, V̂ c an estimate of the voltage across the capacitive element is and L is a gain matrix chosen such that the error dynamics are stable under all conditions.

[0019] A recursive parameter estimation scheme based on a Kalman filter can be used to estimate the parameters (R, R). c, C) of the observer of equations (6) and (7). A discretized form of these parameters can be expressed as a function of the system states as follows: [τs2∗(Voc(k+1)−V(k+1)+Voc(k)−V(k))]=[(V(k+1)−Voc(k+1))τs2∗(I(k+1)+I(k)) I(k+1) ]∗[Rc∗CR+RcR∗Rc∗C]

[0020] The recursive Kalman filter parameter estimation can be achieved by expressing equation (8) in the form Y(k)=ΦT(k)∗Θ(k) This can be achieved. The Kalman filter estimation scheme can then be expressed by the following equations: Θ^(k+1)=Θ^(k)+K(k)∗(Y(k+1)−ΦT(k)∗Θ^(k)) K(k+1)=Q(k+1)∗Φ(k+1) Q(k+1)=P(k)B2I(4T(k+1)∗P(k)∗+(k|1)) P(k+1)=P(k)+R1−P(k)∗Φ(k)∗ΦT(k)∗P(k)R2+(ΦT(k+1)∗P(k)∗Φ(k+1)) where Θ̂(k + 1) is the estimate of the parameters from equation (8), K, Q, and P are calculated as shown, and R1 and R2 are constants. After the parameters have been calculated using the Kalman filter algorithm, the parameters in equations (6) and (7) can be used to obtain an estimate of the state variables. Once V oc Once the value of SOC has been estimated, it can be determined.

[0021] The procedure described is also applicable to other model-based formulations. The filter scheme described above by equations (1) to (13) is merely an example of a model-based estimation scheme. This algorithm is applicable to any implementation where internal state computation is performed to calculate the state of charge. The internal state computation can be used directly or indirectly to calculate the state of charge. The selection of internal states can vary, and the procedure can still be applied. Another possibility is to use the state of charge (SOC) as an internal state and directly estimate its value. The procedure described in this case computes internal voltage states and then derives the SOC from these voltage predictions.The model used and the states used may depend on the specific circuit model used, where the one in . Fig. The circuit model shown in section 3 is an example.

[0022] For the typical battery cell under consideration, there is a relationship between the SOC and V oc , so that V oc = f(SOC). Fig. Figure 5 shows a typical curve 96, the V oc as a function of SOC. The relationship between SOC and V oc can be determined by analyzing battery properties or by testing the battery cells. Once the estimated V oc If the -value is known, the estimated battery cell SOC can be calculated using the relationship between SOC and V. ocThe SOC can be determined. This can be implemented as a table lookup or a corresponding equation. The previous description is merely one possible implementation of a model-based scheme for calculating the SOC. The described method is also applicable to other model-based schemes.

[0023] One advantage of the Kalman filter method is its potential for higher accuracy than that achievable with ampere-hour integration. However, if a Kalman filter must be performed for each individual cell, more memory and processing time may be required than the control unit has available. This problem can become more pronounced with a larger number of cells in the battery pack. Depending on the available computing resources, there may not be enough to effectively implement a Kalman filter for every battery cell.

[0024] The computation time problem becomes clearer when considering a typical microcontroller used in automotive applications. To remain cost-effective, microcontrollers are chosen with limited static and dynamic memory sizes. Additionally, clock speed limitations can restrict the microcontroller's processing speed. Other functions, such as diagnostic and communication functions, may also be performed within the microcontroller, further reducing the computation time available for the state-of-charge (SOC) algorithm. The net result is that only limited microcontroller resources are available to perform the SOC calculation.

[0025] Under certain circumstances, one of the goals may be to provide a more efficient computational structure for calculating the battery state of charge (SOC). This can be achieved by reducing the number of cells that need to be processed using a computationally complex algorithm similar to the previously described Kalman filter. The approach attempts to group cells with similar parameters into subgroups and then implement the model-based calculations on a representative cell or cells within that subgroup. Other cells in that subgroup can then use the calculated values ​​from the representative cell(s). This algorithm enables the estimation of an accurate state of charge while keeping the computational load and memory usage within the constraints of the microcontroller.The method can be applied to any configuration of battery cells and then extended to various methods for calculating the battery cell SOC.

[0026] The initial step is to determine the number of representative cells that the control unit can effectively process. The runtime requirements for processing a representative cell can be determined by implementing the desired algorithm and measuring its execution time. Static and dynamic memory requirements can be determined using compiler and linker output files. The basic execution time is expressed as R base referred to as, and the basic requirement for dynamic memory is called M base designated. Fig. Figure 6 shows a flowchart of a high-level possible implementation of the algorithm. The first step in initialization 100 can be an offline determination of the number of cell calculations that can be handled by the control unit.

[0027] The maximum runtime and maximum memory allocated to the state-of-charge algorithm can be a function of the system design constraints. It should be noted that if there are no constraints on memory or processing speed, it may not be necessary to find a computationally efficient state-of-charge estimation scheme. The algorithm would simply be executed at each cell without regard to memory or execution time. The constraints for execution time and dynamic memory can be expressed as R max or M maxThe maximum number of cells that can be processed based on the runtime constraint is N. r = R max / R base The maximum number of cells that can be processed based on the dynamic storage boundary condition is N. m = M max / M base To reach the maximum number of cells (N) g To arrive at a minimum of N that can be addressed within the system boundary conditions, the minimum of N should be determined. r and N m This description takes dynamic memory into account, but static memory requirements can be handled in the same way. In fact, this scheme can accommodate any other constraint regarding the number of cells that can be processed.

[0028] Once the number (N gOnce the number of possible subgroups that can be implemented within the system boundary conditions has been determined, a method for dividing the cells into subgroups can be established. One possible approach for determining the subgroups is to analyze cells based on battery cell parameters such as battery physics, battery chemistry, battery age, battery temperature, or state of charge (SOC). Each cell can be parameterized and placed into a subgroup based on its measured parameters. Cells in each subgroup can have a specified set of cell parameter values ​​that fall within a specified range, and the specified ranges can be different for each subgroup. Various measurements or values ​​can be available for each cell, such as temperature data, voltage data, and initial state of charge data.One possible method for grouping cells is to consider the following factors: battery age, battery temperature, and initial state of charge (SOC). The idea is that battery cells with the same age, temperature, and SOC are likely to perform similarly.

[0029] One possible approach to determining the subgroups is to first classify the battery cells into N T to divide subgroups, where N T < N g This subdivision may be based on the temperature during the battery cells' startup. There may be N T Temperature ranges can be defined. This subdivision can be achieved through k-means clustering, where a number of data points are grouped into k(N) based on a well-known statistical algorithm. TThe cells are divided into clusters. Once the cells have been divided according to temperature ranges, each subgroup can then be further subdivided according to the SOC. A number of SOC ranges can now be defined. Again, this can be achieved by k-means clustering of SOC data for each temperature-defined subgroup. If the total number of groups is less than N g If this step is complete, it can be repeated up to N. g Subgroups have been determined. There is no limit to how the grouping or clustering can be performed. A predefined temperature range or state-of-charge range can be selected for the subgroups. Any other statistical, mathematical, or graphical method can be applied to cluster battery cells with similar parameter values ​​into the same subgroups.

[0030] The subgroup into which a battery cell is placed does not depend on its location within the battery pack. The battery cells within a subgroup can be distributed throughout the battery pack. Sorting into subgroups is performed in such a way that cells with similar parameters are placed within a given subgroup. Accordingly, using any cell within a subgroup for calculations on all other cells within the subgroup should yield a reliable result.

[0031] One possible implementation of this method is to determine the subgroups during system startup. Cells can be sorted based on their initial temperature and their initial state of charge (SOC) values ​​calculated during startup. Subgroups can be determined based on a specified range of initial states of charge or a specified range of cell temperatures. For a lithium-ion battery, the cell SOC can be determined during initialization. If the battery pack has been inactive for a certain period, the cell's terminal voltage will be equal to its open-circuit voltage. Fig. Figure 6 shows such an approach, in which the determination of grouping 100 is performed when the system is started up.

[0032] Fig. Figure 4 shows an example grouping for a battery pack containing ten cells (A-J). In this example, the cells are to be grouped into five groups. The first step is to group the cells into three subgroups based on temperature (70, 72, 74, represented by ovals). The temperature grouping produces three subgroups: {A, E, H, I}, {C, G, J}, and {B, D, F}. The next step is to group the cells into three subgroups based on the initial state of charge (SOC) (76, 78, 80, represented by rounded rectangles). The SOC grouping produces three subgroups: {A, I}, {C, D, E, G, H, J}, and {B, F}. The temperature- and SOC-based subgroups are then analyzed for intersections, and these intersections can be selected as the final subgroups. In this example, there are clear boundaries between the temperature and SOC subgroups, which are represented by rectangles.The example grouping in . Fig. Figure 4 shows that cells are grouped as {A, I} 82, {C, G, J} 84, {E, H} 86, {D} 88 and {B, F} 90. The procedure for determining the subgroups may depend on whether it is performed offline or in real time using a software-implemented algorithm.

[0033] Within each subgroup, at least one representative cell must be chosen. For example, in subgroup {C, G, J}, one of the three cells C, G, or J must be chosen as the representative element. Several methods can be used to select the representative element: the SOC closest to the average value, the highest or lowest SOC, or the highest or lowest temperature. The selection of the representative element is robust because the cells have already been grouped and should be similar with respect to these parameters. When selecting the representative cells, it may be preferable to choose cells that have a representative nonlinear property. The selection can also be based on any other criteria without affecting the disclosed fundamental approach. Fig. Figure 6 shows such an approach, in which the determination of the representative cell 110 is performed when the system is started up.

[0034] Subgroup selection can be performed in advance based on a calibration or initialization procedure and stored in non-volatile memory. Subgroups can also be selected during system startup. The algorithm is flexible regarding when subgroups are selected. Subgroup selection can also be performed dynamically at runtime, with the subgroup elements changing over time during the same firing cycle. Dynamic selection is particularly useful if, for example, there is a problem with a cell measurement. In this situation, another representative cell that is functioning correctly can be selected for the calculation. Dynamic selection allows for a more robust, fault-tolerant calculation with the ability to continue despite some measurement errors. There is also no restriction regarding when the subgroups are selected. Fig. Figure 6 shows that the division into subgroups occurs during the initialization of system 108.

[0035] Generally, model-based calculations would be performed on only one cell of each subgroup. Depending on available resources, it may be desirable to perform model-based calculations on more than one element of a subgroup. The algorithm supports this modification and can be scaled up or down depending on the control system boundary conditions. Depending on the system boundary conditions, calculations can be performed on one or more representative cells. One possible way to use multiple representative cells is to use the average value of the representative cells.

[0036] One possible technique for implementing this procedure is to determine the required computation rate ΔT. packto determine the battery pack state of charge (SOC). This rate is then determined by the number N. g the subgroups are divided by the sampling period ΔT g to obtain for each subgroup. In each subperiod, the model-based calculations (102 with reference to Fig. 6) for a subgroup. Once all subperiod calculations are complete, the battery pack SOC can be calculated. The battery pack SOC calculation period can be executed in an outer execution loop 104, while the individual battery cell model calculations can be executed in an inner loop 102, which runs more frequently than the outer loop. When the battery cell SOC calculations of the inner loop have been executed, the pack-level battery SOC calculation can be performed. Other implementations are possible and may depend on the hardware and operating system used. Fig. Figure 6 shows a possible subdivision of functions into different high-level tasks for initialization (100), processing of subgroups (102), and processing of packet information (104). Fig. 6. The process group functions 102 can be expressed in intervals of ΔT. g to be executed and process package task 104 in intervals of ΔT pack be carried out.

[0037] Once the representative cells have been selected, battery cell properties can be measured and processed using the model-based algorithm 112. Model values ​​can then be used to calculate the throughput of each cell 114. The throughput can be defined as the amount of charge entering or leaving the battery during a given time interval, or as the amount of electrical charge accumulated or consumed over a given time. The throughput can be calculated using the following formula: Throughput = ∫0Tη I(t)dt where η the charge / discharge efficiency is and I(t) is the current.

[0038] Other ways to calculate the throughput involve model-based approaches. Using equation (14) can lead to accumulated integration errors due to noise in the current measurement. It is assumed that each cell has the same throughput value, although in practice there may be some differences due to cell monitoring hardware, cell balancing hardware, and charging efficiency variations. The method considered in this application relates to a more complex model-based approach, but the method is also applicable to simpler calculations.

[0039] For a model-based approach, the throughput of each representative cell can be calculated. 114 The throughput can be determined from cell charge state data. The open-circuit voltage values ​​(Voc -values) for each representative cell are first determined, possibly using a model-based algorithm 112. Once an estimate of V is obtained oc is available, the SOC can be determined based on the relationship between V oc and SOC are determined (see Fig. 5) The throughput for the representative cell 114 can be increased by Throughput(cell)=(SOC(0)−f−1(Voc(cell)))∗Q(cell) to be determined, whereby cell i one of the representative cells is, f -1 (V oc (Cell i )) based on V oc (see Fig. 5) the representative cell i has a specific SOC value, SOC(0) is the SOC during startup and Q(cell) i ) is the charge capacity of the representative cell. The throughput of group 116 can be calculated as follows: Throughput(Groupi)=(∑1NrcThroughput(Cellj)) / Nrc where N rcThe number of representative cells used in the subgroup.

[0040] It should be noted that equation (16) calculates the average value of the representative cells of the subgroup. In the case where there is only one representative cell per subgroup, equation (16) is not required because it is not necessary to calculate the average of only one value. When the throughput is calculated for more than one representative value, other statistical methods besides averaging can be used. It may be preferable, under different conditions such as the type of loading or unloading, to select the maximum or minimum representative value instead.

[0041] In an implementation where subgroup values ​​are calculated before packet values, there can be computation delays between the time a subgroup is calculated and the time the packet SOC is calculated. These delays can be compensated for by adding an estimated throughput for this period.118 Assume that the packet SOC is calculated every ΔT pack Seconds are calculated. Now assume that a subgroup of all ΔT g = ΔT Pack / N gThe calculation takes place over several seconds. The basic idea is to perform an ampere-hour integration for the time between when the cell throughput is determined and when the battery state of charge (SOC) is to be calculated. Because this is a short period, there is little risk of introducing excessive noise from the current measurement. The throughput for each cell can be compensated for the calculation delays using the following formula: Throughput(Groupj,t(K+1))= Throughput(Groupj,t(K)+j∗ΔTg)+∫l(K)+j∗ΔTyt(K+1)ηI(t)dt where j the group number (1 to N) g ) is.

[0042] The average battery pack throughput of 120 can be considered Average throughput = (∑1NgThroughput(Groupi))Ng (18) can be calculated. The SOC for each cell 122 can then be calculated as SOC(cell,t)=SOC(cell,0)−(Average throughput / Q(cell)) will be calculated.

[0043] Finally, the packet SOC can be determined from the SOC values ​​of the individual cells. There are several methods for determining packet SOC from the SOC values ​​of the individual cells. For higher packet SOC values, the highest cell SOC could be used to prevent overcharging of any of the cells. For lower packet SOC values, the lowest packet SOC could be chosen to prevent complete discharge of any of the cells. In between, an average of the cell SOC values ​​can be chosen as the packet SOC. The described algorithm works with any method for calculating the packet SOC from the individual values. The described equations can be adapted based on the sign convention chosen for the current.

[0044] The control unit can generate an output based on the calculated battery pack state of charge (SOC). This output can be transmitted to other control modules to control vehicle operation. In a hybrid electric vehicle, the battery pack SOC can be used to generate a control output for selecting an appropriate operating mode, such as when the combustion engine should be running. A vehicle control output, such as a combustion engine start request or an electric motor torque request, can also be partially determined based on the battery pack SOC value. In another example, the battery pack SOC output can be used to control the charging and discharging of the battery pack. The control unit can also output the battery pack SOC to a driver display module, informing the driver of the remaining charge in the battery pack.

[0045] The described algorithm can be executed by one or more control units and can be repeated as long as the system is booted. The calculations can be interrupted if the power supply to the control unit is switched off. The system can store values ​​for the next boot cycle in non-volatile memory if necessary.

[0046] The described method provides a computationally efficient way to implement certain complex algorithms with limited computing resources. The method is not necessarily limited to the battery charge level implementation described above. It is also applicable to other areas where a system can consist of similar elements that can be identified and grouped.

[0047] The processes, procedures, or algorithms described herein may be transferable to or implemented on a processing device, a control device, or a computer, which may incorporate any existing programmable electronic control unit or purpose-built electronic control unit. Similarly, the processes, procedures, or algorithms may be stored as data and instructions that can be executed by a control device or computer in any form, including information permanently stored on non-writable storage media such as ROM devices, and information modifiably stored on writable storage media such as floppy disks, magnetic tapes, CDs, RAM devices, and other magnetic and optical media. The processes, procedures, or algorithms may also be implemented in a software-executable object.Alternatively, the processes, procedures or algorithms can be implemented in whole or in part using suitable hardware components such as application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), state machines, control devices or other hardware components or devices, or a combination of hardware, software and firmware components.

Claims

[1] Vehicle comprising the following: a battery pack (14) with cells grouped into several subgroups, and at least one control device programmed to derive the state of charge of the battery pack (14) from an initial state of charge of each of the cells at vehicle activation and from an electrical charge accumulated or consumed by at least one representative cell from each of the subgroups, but by fewer than all cells from each of the subgroups, since vehicle activation, and to charge and discharge the battery pack (14) based on the derived state of charge. [2] Vehicle according to claim 1, wherein the cells are grouped into several subgroups such that at least some of the cells from one or more of the subgroups are distributed over the battery pack (14). [3] Vehicle according to claim 1, wherein all cells of each of the subgroups have a specified group of cell parameter values ​​that fall within a specified range of values, and wherein the specified range of values ​​is different for each of the subgroups. [4] Vehicle according to claim 3, wherein the specified group of cell parameter values ​​has a specified temperature range. [5] Vehicle according to claim 3, wherein the specified group of cell parameter values ​​comprises a specified range of initial charge states or capacities. [6] Vehicle according to claim 3, wherein the specified group of cell parameter values ​​comprises a specified range of cell ages. [7] Vehicle according to claim 1, wherein the at least one control unit is further programmed to group the cells into the several subgroups during vehicle activation.

Images