DE102014216378A1 - Method for the diagnosis of a cell network - Google Patents

Abstract

The invention relates to a method for the diagnosis of a cell network (2), wherein the cell network (2) has at least two battery cells (4) connected in parallel to each other and wherein measured quantities of the cell network (2) comprise a cell composite current intensity and a cell network terminal voltage. The cell network (2) is based on a model whose model sizes are calculated on the basis of the cell network terminal voltage and the cell composite current intensity by solving an optimization task. A computer program and a battery management system (10), which are set up to carry out the method, as well as a battery system and a motor vehicle, wherein a battery is connected to a drive system of the motor vehicle, are also disclosed.

Description

State of the art

The invention relates to a method for the diagnosis of a cell network.

A computer program and a battery management system, which are set up to carry out the method, as well as a battery system and a motor vehicle, wherein a battery is connected to a drive system of the motor vehicle, are also specified.

Batteries typically come in series with battery cells. As far as possible, each individual cell is monitored for capacity and internal resistance. This is necessary to avoid critical or potentially dangerous situations.

WO 2013/016188 shows a method and a system for optimizing parameters of an equivalent circuit of a battery, which has battery cells connected in series. The measured values of voltage sensors, current sensors and temperature sensors are determined as input data. The equivalent circuit of the battery includes charging and discharging resistors and a battery capacity.

US 2013/0300190 shows a method for determining parameters of an equivalent circuit of series-connected batteries, in particular for determining a state of charge of the battery cells. The equivalent circuit of the battery here comprises two RC elements, which describe a self-discharge and a charging resistance of the battery.

If the desired application places corresponding demands on the current level or current dynamics, individual battery cells are switched in parallel. Often, only the sum current is monitored and not the individual cell currents that result for the individual battery cells. In a battery management system, the battery cells are therefore considered in parallel as a battery cell of larger capacity. As long as the battery cells behave similarly, this is justified. If, however, the battery cells differ, for example as a result of varying degrees of aging, the cell currents divide asymmetrically due to the different internal resistances or capacities. This can cause battery cells to operate out of specification, e.g. with elevated temperature or impermissible currents. This can not be detected as a single cell according to the current state of the art due to the substitution consideration. Other sensors, such as Temperature sensors, usually can not be consulted as additional sources of information because they are not mounted on each single cell and provide only a mean temperature per battery module. An elevated temperature or an inadmissible operating state of a single battery cell is therefore not detectable by means of these sensors.

In the publication "Nonlinear Observability and Identity of Single Cells in Battery Packs" by Matthias Rausch, Stefan Streif, Christian Pankiewitz and Rolf Findeisen (Proc. IEEE International Conference on Control Applications, pp. 401-406, Aug 2013) Possibilities of observability and identifiability of parallel and serially connected battery cells are investigated. The authors describe that even for battery cells in parallel connection of combined measurements of current and voltage at a cell network can be concluded in principle on the state of the individual battery cells.

An object of the invention is to avoid the detection and determination of impermissible, i. to allow non-specification operating conditions of parallel connected battery cells.

Disclosure of the invention

According to a first aspect of the invention, a method for the diagnosis of a cell network is proposed, wherein the cell network has at least two battery cells connected in parallel to one another and wherein measured variables of the cell network comprise a cell composite current intensity and a cell network terminal voltage. The cell network is based on a model which has a first ohmic resistance and an RC element connected in series with the first ohmic resistance with a second ohmic resistance and a capacitor for each battery cell present in the cell assembly. Furthermore, the model has an element for the capacity of the individual battery cell. The model has a size of the first ohmic resistance, a size of the second ohmic resistance, a capacitance of the capacitor, an open circuit voltage and a capacity of the battery cell for each battery cell present in the cell assembly as model variables.

wobei der Index i die einzelnen Batteriezellen nummeriert mit i = 1 bis N, wobei N die Anzahl der Batteriezellen ist. The model sizes and measured variables of the cell network are linked as follows:

wherein the index i numbers the individual battery cells with i = 1 to N, where N is the number of battery cells.

x_1,i(t)

die Spannung des RC-Glieds der i-ten Batteriezelle,

x_2,i(t)

den Ladezustand (SOC, state of charge) der i-ten Batteriezelle,

u_i(t)

den auf die i-te Batteriezelle wirkenden Strom,

C_f,i

die Kapazität des Kondensators des RC-Kreises der i-ten Batteriezelle,

U_OCV,i

die Leerlaufspannung der i-ten Batteriezelle,

R_s,i

die Größe des ersten ohmschen Widerstands der i-ten Batteriezelle und

R_f,i

die Größe des zweiten ohmschen Widerstands der i-ten Batteriezelle, wobei mit dem Punkt in bekannter Weise die Ableitung der betreffenden Größe nach der Zeit gekennzeichnet ist.

It indicates thereby

x _{1, i} (t)

the voltage of the RC element of the ith battery cell,

x _{2, i} (t)

the state of charge (SOC) of the ith battery cell,

u _i (t)

the current acting on the ith battery cell,

C _{f, i}

the capacitance of the capacitor of the RC circuit of the ith battery cell,

U _{OCV, i}

the open circuit voltage of the ith battery cell,

Rs _{, i}

the size of the first ohmic resistance of the i-th battery cell and

R _{f, i}

the size of the second ohmic resistance of the i-th battery cell, wherein the point in a known manner, the derivative of the respective size is characterized by the time.

It still applies τ _{f, i} = R _{f, i} · C _{f, i} , where τ _{f, i} denotes the time constant of the RC element of the ith battery cell. C _{nom, i} is another model parameter and defines the capacity of the ith battery cell. U _{c, i} (t) / y _i (t) defines the cell terminal voltage in the above model.

• a) Anfahren eines Startladezustands oder eines Startspannungswertes,
• b) Warten zur thermischen Homogenisierung der Zellen falls nötig (hängt von den konkreten Randbedingungen ab),
• c) Beginnen einer Messwertaufnahme der Zellverbundklemmenspannung und der Zellverbundstromstärke,
• d) Beaufschlagung des Zellverbundes mit einem definierten Stromprofil,
• e) Beenden der Messwertaufnahme der Zellverbundklemmenspannung und der Zellverbundstromstärke und
• f) Berechnen zumindest einer der Modellgrößen anhand der Messwerte der Zellverbundklemmenspannung und der Zellverbundstromstärke.

The method comprises the steps:

• a) starting a starting state of charge or a starting voltage value,
• b) wait for the thermal homogenization of the cells if necessary (depends on the concrete boundary conditions),
• c) starting a measured value recording of the cell connection terminal voltage and the cell composite current intensity,
• d) charging the cell network with a defined current profile,
• e) termination of the measured value recording of the cell connection terminal voltage and the cell composite current intensity and
• f) calculating at least one of the model quantities based on the measurements of the cell interconnect voltage and the cell interconnection current.

According to the invention, at least one model variable or one model parameter is calculated in step f) by solving an optimization task.

The input variables include only the summation current and the sum voltage of the cell group. In solving the differential equation, a parameter estimation is also performed on the model parameters of interest (e.g., R _{s, i} and R _{f, i} ). A distinction between individual battery cells is advantageously made possible with the method according to the invention, wherein the distinction with respect to the sizes of the ohmic resistors and, where appropriate, the capacities of the individual battery cells can be made.

This makes it possible to detect impermissible operating states and initiate measures to counteract a possible danger situation, for example a signaling to the driver, a provision of information about the improper operating state on a communication channel such as a CAN bus, a transfer of the relevant battery module or the battery cell in an emergency mode (limp home), a shutdown of the system, etc.

unter den Nebenbedingungen

bestimmt. At least one model size or model parameter is preferred in step f) by solving an optimization task of the form

under the constraints

certainly.

Here, J _{cost is} a quality criterion which evaluates deviations of the measured cell network terminal voltage from a voltage simulated by means of the above-introduced model.

stellen dabei in bekannter Weise vektorwertige Formulierungen der oben genannten Verknüpfungen der Modellgrößen und Messgrößen miteinander dar, wobei x ˆ(0) = x ˆ₀ die Anfangsbedingungen abbildet und die Optimierungsparameter p ˆ die interessierenden Modellparameter aus dem oben eingeführten Modell (beispielsweise R_s,i und R_f,i) kennzeichnet.

represent in a known manner vector-valued formulations of the above-mentioned linkages of the model sizes and variables with each other, wherein x (0) = x ₀ the initial conditions and the optimization parameters p characterizes the model parameters of interest from the model introduced above (for example, R _{s, i} and R _{f, i} ).

The constraints c form general conditions, in particular the known Kirchhoff laws (node rule and mesh rule), but also additional knowledge about the equivalent circuit parameters, for example, known or meaningful upper / lower limits.

In step a), a desired state of charge or alternatively a voltage level is approached. For this purpose, for example, the battery cells of the cell assembly are discharged or charged with a defined current. The discharging can take place, for example, via a defined load resistor, which is connected to the cell network. The approach of a state of charge or of a voltage level can be used to deliberately target high or low voltage ranges if the impedance behavior or the equivalent circuit parameters are to be determined within a certain range and / or if it is suspected that the differences in this state of charge range or voltage range are of particular relevance.

As initial conditions for the introduced model, the values for the state of charge resulting from the approached voltage level or from the approached state of charge are used. In the event that a voltage level was approached, the state of charge is determined based on the known inverse U _OCV characteristic, which represents a non-linear relationship between open circuit voltage and state of charge.

According to a particularly preferred embodiment, after step a) a defined resting time is maintained, for example up to one hour. This allows a more accurate determination of the state of charge than if immediately followed by step c). By granting the rest period is also achieved in parallel with the battery cells that equalizing currents flow at unequal states of charge until the voltages in the RC elements are equal to 0 and the states of charge have become equal. This defines clear initial conditions that are relevant for solving the optimization task.

Before the battery cells are connected in parallel with a current profile, the measured value recording of the cell connection terminal voltage and the cell composite current intensity is started. The current profile is preferably such that the battery cells are excited as much as possible. For this purpose, profiles with high currents are best suited. However, it should be noted that the cell temperatures do not increase too much to keep a possible temperature influence low. This can be adjusted for example by simulations and tests in advance.

According to a preferred embodiment, the measured value recording in step e) is terminated several 100 s after the end of the application of the current profile in step d). After switching off the current flow in different battery cell balancing currents. Depending on cell size and differences in battery cells, they may still be visible in the voltage curve for several 100 s after the power has been switched off. Therefore, this time range is preferably recorded with.

According to a preferred embodiment, further parameter estimates implemented in the battery management system are used in step f), in which, for example, the parallel-connected battery cells are regarded as a single battery cell. This knowledge is used in optimization.

In one embodiment, the process is performed repeatedly. Particularly preferably, the method is carried out at regular intervals, wherein the choice of distances can be estimated by prior knowledge.

According to one embodiment, during the measured value recording in step c), the measured values of the cell network terminal voltage and the cell composite current intensity are fed to a filter device. If necessary, the robustness of the estimation algorithm can be increased by preprocessing the data.

To specifically cover fault cases in a parallel connection of more than two battery cells, for example in three battery cells, one of which is suspected to be high impedance, two or more battery cells can be combined into a spare cell. Since these two or more battery cells are assumed to be alike, the replacement parameters can be calculated using the well-known nodal and mesh rule of Kirchhoff's laws. The capacity of the spare cell is a multiple of the combined single cell capacities. Analogously, other error cases, eg. B. two high-impedance and two low-impedance battery cells are mapped in a parallel circuit of four battery cells. This can further increase the robustness of the method. High impedance means in this context that the internal resistance of these battery cells is noticeably higher than that of the other cells in the composite and low impedance, that the internal resistance of these battery cells is noticeably lower than that of the high-resistance cells in the composite.

According to the invention, a computer program is also proposed according to which one of the methods described herein is performed when the computer program is executed on a programmable computer device. The computer program may be, for example, a module of a battery management system. The computer program can be stored on a machine-readable storage medium, for example on a permanent or rewritable storage medium or in association with a computer device, for example on a portable storage such as a CD-ROM, Blu-ray Disc, DVD, a USB stick or a memory card , Additionally or alternatively, the computer program may be provided for download on a computing device, such as on a server or a cloud server, for example via a data network, such as the Internet, or a communication link, such as a telephone line or a wireless link.

According to the invention, a battery management system (BMS) is also provided for carrying out one of the methods described above, wherein the battery management system has units for detecting the cell interconnection voltage and the cell interconnection current strength. Furthermore, the battery management system comprises units for carrying out a measured value recording, wherein the units for implementing the measured value recording are set up to approach a starting state of charge or a starting voltage value and to carry out a measurement recording of the cell network voltage and the cell composite current intensity, and units for calculating at least one of the model variables the measured values of the cell connection terminal voltage and the cell composite current intensity by solving the optimization task.

The battery management system is preferably designed and / or set up to carry out the methods described herein. Accordingly, features described in the context of the method correspondingly apply to the battery management system and, conversely, the features described within the scope of the battery management system correspondingly for the methods.

The units of the battery management system are to be understood as functional units that are not necessarily physically separated from each other. Thus, multiple units of the battery management system may be implemented in a single physical unit, such as when multiple functions are implemented in software on a controller. The units of the battery management system can also be implemented in hardware components, in particular by sensor units, memory units, application specific integrated circuits (ASIC, Application Specific Circuit) or microcontroller.

According to the invention, a battery system is also provided with a battery which comprises a plurality of battery cells and such a battery management system. The battery may in particular be a lithium-ion battery or a nickel-metal hydride battery and be connectable to a drive system of a motor vehicle.

The terms "battery" and "battery unit" are used in the present description adapted to the usual language for accumulator or Akkumulatoreinheit. The battery includes one or more battery units, which may be a battery cell, a battery module, a module string or a battery pack. In the battery, the battery cells are preferably combined spatially and circuitally connected to each other, for example, connected in series or parallel modules. Several modules can form so-called battery direct converters (BDC, Battery Direct Converter) and several battery direct converters form a battery direct inverter (BDI, Battery Direct Inverter).

According to the invention, a motor vehicle with such a battery system is also provided, wherein the battery of the battery system is connected to a drive system of the motor vehicle. The motor vehicle may be configured as a pure electric vehicle and exclusively comprise an electric drive system. Alternatively, the motor vehicle may be configured as a hybrid vehicle comprising an electric drive system and an internal combustion engine. In some variants it can be provided that the battery of the hybrid vehicle can be charged internally via a generator with excess energy of the internal combustion engine. Externally rechargeable hybrid vehicles (PHEV) also provide the option of charging the battery via the external power grid. In motor vehicles designed in this way, the driving cycle comprises a driving operation and / or a charging operation as operating phases in which operating parameters are detected.

The method described can, for example, also be carried out on a central server to which the motor vehicles are connected, for example, via a wireless connection and to which the recorded data is transmitted.

Advantages of the invention

The essence of the invention relates to a method which makes it possible to detect dissimilar lines in parallel and to obtain a quantitative description of the inequality via a determination of cell parameters.

In contrast to previously used models, which view parallel-connected battery cells as just a single battery cell with a correspondingly larger capacity, the method uses a model in which a distinction between the individual cells is possible.

It is advantageous that can be dispensed with additional current, voltage or temperature sensors. Thus, the method can be used in battery management systems used today. This allows a cost-effective implementation without changes to existing systems or complex installation of additional sensors.

Due to the fact that a more detailed model is used than hitherto customary, the parameters R _s and R _{f of} each individual cell can be obtained as a result of the optimization, for example. The knowledge of these parameter values makes it possible to estimate the unevenness of the individual cells and a resulting dissimilar current distribution, for example via a simulation using these values.

The method can also be used to detect non-uniform capacitance values by extending the parameter set to be determined by the respective parameters C _{nom, i} .

Brief description of the drawings

Embodiments of the invention are illustrated in the drawings and explained in more detail in the following description. Show it

1 two parallel connected battery cells with uneven current distribution and

2 an equivalent circuit of battery cells in parallel with total current measurement.

1 shows a cell network 2 by way of example two battery cells 4 which have aged differently. The battery cells 4 For example, lithium ion batteries have a voltage range of 2.8V to 4.2V.

The battery cells 4 have different internal resistances. In the parallel connection of the two battery cells 4 as they are in 1 is shown, resulting from the fact that by the individual battery cells 4 different currents flow. Through the first battery cell 4 a current of 150 A flows through the second battery cell 4 flows a current of 50 A. Overall, flows through the cell network 2 a current of 200 A. In 1 is also a unit 6 for detecting a cell interconnection voltage and a unit 8th for detecting a composite cell current which is part of a battery management system 10 are.

The battery management system 10 also includes a controller 12 which shows the measurements of the unit 6 for detecting the cell interconnection voltage and the unit 8th receives and processes to detect cell composite current. The control unit 12 can also be referred to as a battery control unit (BCU) and implements functions for controlling and monitoring a battery in which the cell network 2 is used.

The control unit 12 includes a unit 14 for carrying out a measured value recording and a unit 16 for calculating model sizes based on the measured values. The unit 14 for carrying out the measured value recording is set up to start a starting state of charge or a starting voltage value at which a measured value recording of the cell connection terminal voltage and the cell composite current is to take place. In practice, the unit is 14 for carrying out the measured value recording, for example, to a controllable load 18 connected, such as a lot of balancing resistors. The adjustable load 18 is parallel to the connected battery cells 4 installed and normally open, ie it has an infinite resistance. When starting a desired state of charge, z. B. from 95% to 90% then becomes a finite resistance of the variable load 18 set.

The unit 14 for performing the measured value recording is also set up to begin a measured value recording of the cell network terminal voltage and the cell composite current intensity after maintaining a defined rest period and the cell network 2 to apply a defined current profile. Besides, the unit is 14 for carrying out the measured value recording, ending the measured value recording of the cell connection terminal voltage and the cell composite current intensity after a further defined rest period after the end of the application with the current profile and the recorded measured values of the unit 16 to calculate the model sizes.

The unit 16 For calculating the model quantities, it is arranged to solve the differential equations described herein using the optimization task described herein, which will also be described below by way of example with reference to FIG 2 is explained in more detail.

2 shows a model 20 with respect to 1 described cell composite 2 with the two battery cells 4 ,

The model 20 points to everyone in the cell group 2 existing battery cell 4 a first ohmic resistance 22 and one in series with the first ohmic resistor 22 switched RC element 24 with a second ohmic resistance 26 and a capacitor 28 on.

The model 20 points to everyone in the cell group 2 existing battery cell 4 as model quantities, a quantity R _{s, 1,...,} R _{s, N of} the first ohmic resistance 22 , a quantity R _{f, 1, ...,} R _{f, N of} the second ohmic resistance 26 on, a capacitance C _{f, 1, ...,} C _{f, N of} the capacitor 28 , a current intensity I _1,..., I _N and a terminal voltage U _{OCV, 1...,} U _{OCV, N} , where the individual battery _{cells have} the respective index 4 in the considered cell network 2 be designated.

Depending on the number of cells N in the parallel circuit, a corresponding number of equivalent circuit equations is used to model the interconnection.

Exemplary for two battery cells 4 the following equations result:

On the basis of Kirchhoff's laws, the rule of knots and the rule of mesh, the two battery cells are exemplary 4 the algebraic relations u (t) = u ₁ (t) + u ₂ (t), 0 = U _{c, 1} (t) - U _{c, 2} (t).

Systems with more than two battery cells 4 are generalized accordingly.

The invention is not limited to the embodiments described herein and the aspects highlighted therein. Rather, within the scope given by the claims a variety of modifications are possible, which are within the scope of expert action.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• WO 2013/016188 [0004]
• US 2013/0300190 [0005]

Cited non-patent literature

• "Nonlinear Observability and Identity of Single Cells in Battery Packs" by Matthias Rausch, Stefan Streif, Christian Pankiewitz, and Rolf Findeisen (Proc. IEEE International Conference on Control Applications, pp. 401-406, Aug 2013) [0007]

Claims (9)

Method for the diagnosis of a cell network ( 2 ), whereby the cell network ( 2 ) at least two parallel connected battery cells ( 4 ) and wherein measured variables of the cell network ( 2 ) comprise a cell composite current strength and a cell interconnect voltage, wherein the cell composite ( 2 ) a model ( 20 ), which is assigned to everyone in the cell group ( 2 ) existing battery cell ( 4 ) a first ohmic resistance ( 22 ), one capacitance and one in series with the first ohmic resistance ( 22 ) switched RC element ( 24 ) with a second ohmic resistance ( 26 ) and a capacitor ( 28 ), the model ( 20 ) to everyone in the cell group ( 2 ) existing battery cell ( 4 ) as model sizes a size of the first ohmic resistance ( 22 ), a magnitude of the second ohmic resistance ( 26 ), a capacitance of the capacitor ( 28 ), an open circuit voltage and a capacity of the battery cell, wherein the model sizes and measured variables of the cell network ( 2 ) are linked together as follows:

where the index i the individual battery cells ( 4 numbered with i = 1 to N, where N is the number of battery cells ( 4 ), where x _{1, i} (t) is the voltage of the RC element ( 24 ) of the i-th battery cell, x _{2, i} (t) the state of charge of the ith battery cell, u _i (t) of the current acting on the i-th battery cell, U _{c, i} (t) / y _i (t) the modeled cell terminal voltage, U _{OCV, i} the open circuit voltage of the ith battery cell, C _{nom, i} the capacitance of the ith battery cell, C _{f, i} the capacitance of the capacitor ( 28 ) of the RC circuit of the ith battery cell, R _{s, i} the size of the first ohmic resistance ( 22 ) of the i-th battery cell and R _{f, i} the size of the second ohmic resistance ( 26 ) of the i-th battery cell, and wherein τ _{f, i} = R _{f, i} · C _{f, i} , where τ _{f, i} denotes the time constant of the RC element of the ith battery cell, with the steps: a) starting of a starting state of charge or a starting voltage value, b) waiting for the thermal homogenization of the battery cells ( 4 ) if necessary, c) starting a measured value recording of the cell network terminal voltage and the cell composite current intensity, d) loading the cell network ( 2 f) calculating at least one of the model quantities on the basis of the measured values of the cell interconnection voltage and the cell interconnected current strength, characterized in that at least one model size is calculated in step f) by solving an optimization task , Method according to Claim 1, characterized in that the optimization task has the following form:

under the constraints

where J _{cost is} a quality criterion, and

represent vector-valued formulations of the linkages of the model quantities and measured variables, wherein x (0) = x ₀ the initial conditions and images p is the parameter vector to be optimized. A method according to claim 1 or 2, characterized in that after step a) a defined rest time is maintained. Method according to one of the preceding claims, characterized in that the measured value recording in step e) several 100 s after the end of the application of the current profile in step d) is terminated. Method according to one of the preceding claims, characterized in that in the measured value recording in step c), the measured values of the cell network terminal voltage and the cell composite current intensity are fed to a filter device. A computer program for performing any of the methods of any one of claims 1 to 5 when the computer program is executed on a programmable computing device. Battery management system ( 10 ) for carrying out one of the methods according to one of claims 1 to 5, wherein the battery management system ( 10 ) one unity ( 6 ) for detecting the cell interconnection voltage and a unit ( 8th ) for detecting the cell composite current intensity, and the battery management system ( 10 ) continue units ( 14 ) for carrying out a measured value recording, wherein the units ( 14 ) are set up to perform the measured value recording, to approach a starting state of charge or a starting voltage value and to carry out a measurement recording of the cell connection terminal voltage and the cell composite current intensity, and units ( 16 ) for calculating at least one of the model quantities on the basis of the measured values of the cell connection terminal voltage and the cell composite current intensity by solving the optimization task. Battery with a battery management system ( 10 ) according to claim 7, wherein the battery is connectable to a drive system of a motor vehicle. Motor vehicle with a battery according to claim 8, wherein the battery is connected to a drive system of the motor vehicle.

Images