WO2014128395A1

WO2014128395A1 - Method and device for estimating the impedance of a motor vehicle battery

Info

Publication number: WO2014128395A1
Application number: PCT/FR2014/050327
Authority: WO
Inventors: Antoine Saint-Marcoux; Thomas Peuchant; Yves LE-VOURCH
Original assignee: Renault S.A.S
Priority date: 2013-02-21
Filing date: 2014-02-17
Publication date: 2014-08-28
Also published as: FR3002325B1; FR3002325A1

Abstract

A method for estimating the impedance of at least one component of a motor vehicle battery during a phase of charging said battery from the mains, comprising: controlling (40) the application, to said component, of a current signal from the mains and defining a harmonic frequency, so as to charge said component, receiving (41) at least two voltage values measured at the terminals of said component at respective times defining a time period less than or equal to half of a period corresponding to said harmonic frequency, and determining (42, 43, 44), from said at least two voltage values and from the applied current signal, an impedance value associated with the harmonic frequency. A device for estimating impedance comprising means for carrying out this method.

Description

METHOD AND DEVICE FOR ESTIMATING IMPEDANCE

The invention relates to the impedance estimation of at least one component of a motor vehicle battery, in particular an electric and / or hybrid vehicle. The invention can find a particularly advantageous application in the detection of electrical circuit fault of the battery.

A battery comprises a set of cells connected together by busbar ("busbar" in English) The cells are grouped into modules of for example 24 cells each. Other bus bars electrically connect two modules together.

In a known manner, a battery management system (BMS) (battery management system) makes it possible to detect electrical circuit faults of the cells themselves, the faults between a cell and a busbar supposed to be connected to this cell. , and / or defects of another component of the battery. In particular, the BMS may make it possible to measure a voltage across each cell element, a cell element comprising a cell and a busbar supposed to be in electrical contact with this cell. The BMS can furthermore make it possible to measure the voltage between two modules electrically connected by a busbar.

To measure the impedance of a motor vehicle battery component, it is preferable to apply a current or voltage varying over time, because there may be residual voltages. During rolling, during a strong current spike, the variation of current during a given period is compared with the corresponding voltage variation across the component. The ratio of these two variations then corresponds to an apparent resistance value.

Nevertheless, there is a need for impedance measurements with better repeatability.

One could consider adding to the motor vehicle, an impedance meter applying excitation current or voltage with a frequency ramp, and to obtain an impedance curve of the component considered. However, such a device is relatively heavy and expensive. US 6,653,817 discloses a system for estimating the impedance of the battery, based exclusively on the power electronics of the vehicle. It describes a phase of excitation of the battery by running over a wide frequency band by its power electronics. However, since the current excitation is spread over this wide frequency band, this method allows a rough estimation of the impedance on this frequency band, but it does not allow an accurate estimation of the impedance at the frequency reference.

There is provided a method for estimating the impedance of at least one component of a motor vehicle battery during a mains charging phase of this battery, comprising:

controlling the application to this component of a current signal coming from the sector and defining a harmonic frequency, so as to charge this component,

receiving at least two values of voltages measured at the terminals of this component at respective instants defining a period of time less than or equal to half of a period corresponding to this harmonic frequency,

determining from said at least two voltage values and the applied current signal an impedance value associated with the harmonic frequency.

Advantageously, the harmonic frequency associated with the impedance value may be the first harmonic frequency, also called fundamental, even if it can be provided to base this method on higher order harmonics. As the amplitude associated with the first harmonic is generally relatively high, the determination of the impedance value is more reliable.

A charger can turn an alternating current (AC) signal into an input from the mains into a charging signal at the input of the battery - the characteristics of which depend on the topology of the charger. At the output of the charger (and therefore, at the input of the battery), it comes to take advantage of relatively large variations in the battery charging current signal (and its voltage response) to make reliable measurements of the impedance. Due to this use of the battery charge signal, the impedance estimate has better repeatability than in the prior art, without adding any particular device of the impedance meter type.

The current signal may have a bandwidth around the fundamental frequency less than or equal to 5% of this fundamental frequency and advantageously less than or equal to 1%.

The bandwidth can for example be determined by choosing the frequencies around the fundamental frequency, corresponding to the maximum of the peak of the frequency transform, such as the ratio between the maximum amplitude of the peak and the amplitude for these frequency values. or by measuring the difference between these two frequencies.

The applied current signal may be such that the amplitude of the current associated with the fundamental frequency may exceed the highest amplitude by 50% for the other non-zero frequencies and advantageously be at least two times greater than this amplitude. high.

It can be noted that the amplitude associated with the fundamental frequency can be of the same order of magnitude as the DC component, or not - depending on the topology of the charger.

Typically, the fundamental frequency derives directly from the frequency of the sector: for example, these frequencies are equal, or the fundamental frequency is a multiple of the frequency of the sector, for example twice the frequency of the sector.

The applied current signal may have amplitude variations as a function of time greater than 30% of the maximum amplitude of this signal, advantageously greater than 50%, advantageously greater than 80%, and advantageously between 95% and 100%.

This method can further simplify the design of a motor vehicle charger system in that it is possible to remove capacitive elements to smooth a signal with a non-zero DC component obtained from the sector. A charger device may be arranged to rectify the mains signal, for example a 50 Hertz or 60 Hertz signal, without clipping through a capacitance or the like. Alternatively, the charger device may include a diode to pass substantially than the positive current values, the loading being effected from the signal obtained at the output of the diode.

Thus, the battery is charged by means of a signal having relatively large variations and these time variations are used to evaluate the impedance of a component of the battery.

Advantageously and without limitation, it is possible to measure more than two voltage values at the terminals of the component in question and to develop an impedance value associated with the fundamental frequency from this plurality of voltage values.

Advantageously and without limitation, the time interval separating two voltage measurement values may be strictly less than one half-period corresponding to the fundamental frequency. For example, this period of time may vary between a quarter period and a hundredth of a period, for example of the order of one-tenth of a period.

Advantageously and in a nonlimiting manner, the method may comprise a processing step of performing a frequency transform, for example a Fast Fourier transform (FFT), from voltage values. measured, and determining a value of this frequency transform of the voltage signal corresponding to the fundamental frequency.

The method may further comprise a step of deriving the impedance value associated with the fundamental frequency from the voltage value associated with the fundamental frequency value thus determined and from a charge current value associated with this fundamental frequency, for example by performing a ratio of these two values.

Thus, voltage values at the terminals of the battery component are measured with a sufficiently fine sampling pitch to be able to determine the value of a voltage component corresponding to the fundamental frequency.

The invention is in no way limited by the calculation of a frequency transform, nor by the choice of such a fine sampling step. For example, a sampling step that is close to or equal to half of a half-period corresponding to the fundamental frequency could be provided. For example, one could expect to read two values of voltage corresponding to a portion of the continuously varying current signal, for example an upward or downward phase of the current signal. The equivalent impedance could then be determined by performing a ratio between a difference between these two measured voltage values and a difference between two current values corresponding to these two voltage values, for example two values of the current signal taken at the same times. than these two values of the voltage signal.

The invention is therefore in no way limited by the processing performed to deduce the impedance value associated with the fundamental frequency.

Advantageously and in a nonlimiting manner, the component may comprise:

- a cell,

an electrical connection between a cell and a busbar supposed to be connected by welding to this cell,

- a bus bar between two cells,

a cell element, that is to say a cell and a busbar supposed to be in electrical contact with this cell, a busbar supposed to be in electrical contact with two modules of a battery pack,

- a module,

- a complete pack, and / or

- other.

In general, the invention is not limited by the nature of the component. For example, it may be possible to measure the voltage signal across a set of 2, 3, 4 or other cell elements.

Advantageously and in a nonlimiting manner, the method may comprise a step of comparing the estimated impedance value with at least one threshold value, and emitting an alarm signal, for example an electrical circuit fault detection signal into function of the result of the comparison.

For example, it is possible to read in a map an expected impedance value associated with the fundamental frequency, as a function of a measured temperature value and a charge state value or SOC. State Of Charge "), measured. We can compare the impedance value determined by applying the method described above to the impedance value resulting from the mapping, and if these two values differ too much from each other, for example if the absolute value of the difference between these two impedance values is greater than a predetermined threshold, an alarm signal can be emitted to signal an electrical circuit fault detection.

Without limitation, two maps may be provided for the value of the "theoretical" impedance depending on whether the excitation is at 50 Hz or at 60 Hz. By simple signal processing methods, a person skilled in the art can integrate an applicable mapping determination logic, depending on the input excitation of the battery.

Alternatively and advantageously, it is possible to estimate a plurality of impedance values at the fundamental frequency, for a plurality of respective components of the battery, and to compare at least two impedance values derived from this plurality. . For example, it is possible to compare the impedance value obtained for a given component with a threshold value obtained from at least one other impedance value corresponding to at least one other respective component.

If an impedance value corresponding to a given component differs too much from the impedance of the other components, then an alarm signal may be emitted, for example to signal an electrical circuit fault detection. This method improves the robustness of this detection, in that it limits the risk of false detection: in fact, this method amounts to comparing with one another, the responses in cell voltage (or bus bars) undergoing exactly the same excitation. current. An impedance value for a given component (cell or busbar) abnormally different from the others would necessarily be related to the specific conditions (aging, faulty electrical contact) of the component. Moreover, this method makes it possible to avoid drawing up and memorizing a map.

The invention can thus find a particularly advantageous application in the detection of total or partial breakage of electrical contact in motor vehicle batteries, but it is of course not limited to this application. For example, it would be possible to use the estimated impedance value thus determined to estimate a parameter representative of aging, for example using a known method. For example, mapping may be used to associate a set of temperature values and SOC values with a set of impedance values at the fundamental frequency. If the fundamental frequency impedance value determined by the method described above differs too much from the expected impedance value at this fundamental frequency, resulting from the mapping, it is possible to emit an alarm signal to signal that the battery is too old. For example, it is possible to determine from these two impedance values at the fundamental frequency a state of health parameter (SOH) of the battery, for example by taking the ratio the expected value and the measured value.

The impedance values determined by the method described above may be actual values, i.e., resistance values, or alternatively may have a complex component.

Advantageously and in a nonlimiting manner, the method may comprise a step of determining, from at least one estimated impedance value corresponding to a busbar, a latency value or phase shift value. This latency value can be induced by the acquisition chain, that is to say for example a communication network involving processors, of the ASIC type (of the English "Application Specifies Integrated Circuit"), CPU (of Central Processing Unit), communication buses, analog digital converters, and others. Thus, as it is known that a busbar does not as such complex impedance, is based on the delay observed across the busbar to determine a latency value introduced by the acquisition chain. This latency value can then be used to estimate the impedance values of other battery components, including cell elements. For at least one other component of the battery, comprising a cell for example, an impedance value of this component is estimated taking into account this latency value.

There is further provided a computer program product comprising instructions for performing the steps of the described method. above when these instructions are executed by a processor. This computer program can for example be stored on a memory medium, downloaded, or other.

There is further provided a device for estimating the impedance of at least one component of a motor vehicle battery, comprising:

- Mains charge control means for applying to this component a current signal from the sector and defining a harmonic frequency, so as to load this component.

means for receiving at least two values of voltages measured at the terminals of this component at times defining a period of time less than or equal to half a period corresponding to the harmonic frequency, and means for processing in communication with the load means and the receiving means, and arranged to determine from the received voltage values and from the applied current signal an impedance measurement value of the component associated with the harmonic frequency.

This device can for example include or be integrated in a digital signal processing processor, for example a micro processor, a micro controller, a DSP (the "Digital Signal Processor") or other.

This device can for example include or be integrated in a BMS.

The mains charge control means may for example comprise or be integrated in a processor controlling the charge of the battery, for example a processor controlling rectifying means of a mains current signal.

The receiving means may for example comprise an input port, an input pin or the like.

The processing means may for example comprise a processor core, a processor or other.

There is further provided a battery system comprising the device described above as well as a charging device for charging the battery from the mains and / or the battery itself. This loading device can be arranged to apply the signal of current from the component sector, and / or the entire battery.

It is further provided a motor vehicle comprising a battery system as described above, for example an electric vehicle, hybrid and / or other.

The invention will be better understood with reference to the figures which illustrate non-limiting embodiments. :

- Figure 1 schematically illustrates a battery portion of a motor vehicle.

FIG. 2 shows an example of a battery system, according to one embodiment of the invention.

FIG. 3A is a graph showing the time course of a mains current signal and a rectified current signal as applied to the battery according to one embodiment of the invention.

FIG. 3B is a graph showing frequency transformations of the signals of FIG. 3A.

FIG. 4 is a flowchart of an exemplary method according to one embodiment of the invention.

With reference to FIG. 1, a battery pack comprises a plurality of modules (not shown), for example eight modules. Each module may for example weigh between 10 and 15 kg and comprise a plurality of cells, for example 24 cells.

Two adjacent modules are electrically connected to each other by a copper busbar (not shown). Within each module, adjacent lithium-ion cells Ci, Ci + i are electrically connected by means of a busbar BBi.

More precisely, the cell Ci comprises a negative electrode ENi and a positive electrode EPi. And the cell Ci + i comprises a negative electrode ENi + i and a positive electrode EPi + i. BBi bus bus is welded firstly to the positive electrode EPi of the cell i and secondly to the negative electrode ENi + i of the cell (i + 1).

A cell element Ci comprises the cell Ci as well as the bus

F -ccll x i is measured across the cell element Ci corresponding to the cell Ci.

This tension can be written as the sum: an empty voltage value;

a cell voltage value equal to the product of the internal resistance of the cell Ci and a current value passing through the cell. This value of internal resistance to the cell is a function of the temperature and the state of charge of the cell. In particular, if the temperature is relatively low, the internal resistance of the cell Ci may have a relatively high value.

- The product of a value of the resistance of the connection between the positive electrode EPi and the bus bar BBi and the current value. This connector resistance value can be relatively high if the solder is damaged; and

the product of the resistance of the busbar BBi and of the current value crossing the busbar BBi. This resistance value of the bus bar BBi is relatively low, for example of the order of 0.1 πιΩ, and is related to the conductivity of the busbar material. In particular, this busbar resistance value varies relatively little with temperature.

It is sought to detect possible electrical circuit faults of the battery, from estimated values of the resistances of the various components of this battery.

With reference to FIG. 2, a battery system 1 comprises the actual battery 10, a charging device, for example a charger 1 1, to recharge the battery 10 from the mains current and a fault detection device 12 .

As is apparent from FIGS. 3A and 3B, the charger 1 1 is arranged to rectify a mains current signal 30 into a charge signal 31. It can be noted that no smoothing is applied, and that it is the signal 31 which is applied to the entire battery 10. This battery 10 is represented here schematically in the form of a succession of elements of cells Ci, C'2,. . . , C'N.

For each cell element, the voltage at the terminals of this cell element is measured and the potential values thus measured are sent to a corresponding component of the device 12, for example an ASIC. These different ASICs are in communication with a CPU processor of the device 12.

The device 12 may be a BMS or other. In addition, measurement elements of the current passing through the battery pack, here represented diagrammatically by the reference 13, allow the CPU to receive a measurement value of the current applied to the battery. These current measuring elements may for example include an ammeter.

With reference to FIG. 3B, if the mains current signal defines a peak at 50 Hertz, the load current signal defines:

a non-zero DC value, ie a positive current makes it possible to recharge the battery,

a peak at a fundamental frequency, here 100 Hz, and

- other harmonics of higher rank.

Thus, the frequency transform 30 'of the mains current signal is mainly composed of a peak at 50 Hz.

The frequency transform of the rectified current signal 31 'comprises a DC component, i.e. a peak at 0 Hz, a peak at 100 Hz and harmonics at 200, 300, 400 Hz, and so on.

In this embodiment, the charger 1 1 of Figure 2 is arranged to rectify the mains current. Other variants could of course be provided, provided that it is a positive current that enters the battery and that the charge current signal defines a non-zero discrete frequency, from which the impedance can be determined.

Advantageously, the load is single-phase, but it will be possible to provide a three-phase load provided that the frequency transform of the load current signal has a harmonic component of non-zero battery current.

The impedance value of a given component can then be estimated by applying the following formula:

Z cIompos

in which

- Ur fundamental frequency, here 100 Hz;

is a frequency transform of the complex impedance of component i; component is a frequency transform of a voltage signal measured across the component i;

- I Bat is a frequency transformation of the battery charge signal.

To return to Figure 2, we can provide material filters

("Hardware" in English), for example discrete filters, or software ("software" in English) not shown, and digital conversion means or ADC (of the English "Analog to Digital Converter"), to filter the measurement noise while preserving the frequency component of the current and the cell voltage corresponding to the fundamental frequency. It can for example be low-pass filters having a cutoff frequency sufficiently higher than the fundamental frequency.

With reference to FIG. 4, a method for detecting an electrical circuit fault of a battery may comprise a first step referenced 40 consisting in transmitting a control signal to the charger referenced 1 1 in FIG. 1 so that this charger 1 1 applies a load current signal to the battery 10. This signal of the charging current is obtained by rectification of the mains current. This step of transmitting a control signal can be carried out within the

BMS 12 of Figure 2.

Then, for each cell element Ci, the voltage across this cell element is measured.

n receives a plurality of measured voltage values

during a step 41. In this example, the lapse of time elapsed between two consecutive measurements of voltage V _t (fl) _j

V, (n + 1) is about 1 ms. Such a millisecond sampling step makes it possible to collect about ten values per period of the current signal applied.

During a step 42, it applies to the values received at the step

41 a fast Fourier transform to obtain a frequency transformation.

Then during step 43, a sampling of the charge current signal is applied to a fast Fourier transform, a to obtain a frequency transform signal of the current signal passing through the battery.

Finally, during a step 44, an impedance value corresponding to the fundamental frequency, here 100 Hz insofar as the mains current is at 50 Hz, is determined by taking the ratio between the frequency transform value of the voltage for this fundamental frequency, and the current transform value for this fundamental frequency.

During a step 45, a mapping is read an expected impedance value for a measured battery temperature value and a measured state of charge value.

During a test 46, the values determined in steps 44, 45 are compared with each other. If the absolute value of the difference between these two impedance values is greater than a threshold, then the emission of an alarm signal is commanded in order to signal the detection of a fault.

In this example, a loop, indexed i, is set up in order to browse the different cell elements C'I,. . . , C'n. The initialization step 47 and the incrementing step 48 have been represented. On the other hand, the loop output step, when i has reached the value N of the number of cells, is not represented in this FIG.

In an alternative embodiment, the comparison between the impedance value determined in step 44, and the theoretical value read at step 45 could be performed by calculating a ratio between these two values according to the formula:

SOHP _cel , {ff _fm on _l d _l ),

In which ^ _a is the theoretical impedance value that would have for the fundamental frequency a new cell, resulting from a

mapping. If this term is below a threshold, then the alarm signal is transmitted because it means that the impedance of the cell element has become abnormally high.

According to another embodiment, it is possible to detect a fault in the electrical circuit at the level of the other battery components, for example the bus bars electrically connecting the modules between them. The evolution of the voltage at the terminals of these bus bars is thus measured, and respective impedance values are estimated by frequency analysis, these impedance values corresponding to the fundamental frequency of the signal of the charging current of the battery. The impedance obtained for each of these bus bars is compared, this obtaining being carried out by applying steps similar to steps 41, 42, 43 and 44, to a reference value · This gives the ratio:

If this term SOHPcomposantiffond) falls below a threshold, then a diagnosis is made because it means that the component impedance has become abnormally high.

In another embodiment, the electric circuit fault detection can be performed by comparing the different impedance values obtained in step 44 for the different cell elements. For example, we can calculate an average value over the set of N average cell elements {z _c ^l _omposant (f _fond \ i = \ ..N}, and, for each of the components, calculate a ratio between this average value and the impedance value determined in step 44, according to the formula:

SOHP c ¹ omposant

If SOHP term _component ff _ond) is below a threshold, then it sends an alarm message, because it means that the impedance of the component i became abnormally high compared to other components. This can potentially be related to an electrical circuit fault.

To return to FIG. 2, in the BMS 12, the chain of acquisition of the voltage at the terminals of the components of the cell, busbar, etc. type. is different from the acquisition chain, the current, here referenced 13. In order to ensure that these acquisition chains are properly phased together, it will be possible to measure the phase difference between a voltage value read across a busbar and the value of the load current.

Thus, if the voltages are not perfectly synchronized, the busbar value is not purely

real. If there is a phase shift of φ between the current voltage measurement value, while the resistance of the bus bar cell is R, then we would get busbar ^. It is thus possible to find not only the resistance value R, but also the phase shift value φ. The knowledge of this phase shift value can be used to correct the phase difference between the voltage measurements and those of the current for the other components of the battery, in particular the cell elements.

Claims

claims

A method for estimating the impedance of at least one component of a motor vehicle battery during a mains charging phase of this battery, comprising:

controlling (40) the application to said component of a current signal coming from the sector and defining a harmonic frequency, so as to charge said component,

- receiving (41) at least two measured voltage values across said component at respective times defining a period of time less than or equal to half a period corresponding to said harmonic frequency, and

- determining (42, 43, 44) from said at least two voltage values and the applied current signal an impedance value associated with the harmonic frequency.

2. The method of claim 1, wherein the current signal from the sector and applied to the component is obtained by straightening the mains current signal.

3. Method according to one of claims 1 to 2, wherein the lapse of time varies between a quarter of the period corresponding to said harmonic frequency and one hundredth of said period.

4. Method according to one of claims 1 to 3, further comprising:

comparing (46) the estimated impedance value with at least one threshold value, and

- emit an electric circuit fault detection signal according to the result of the comparison.

5. The method as claimed in claim 4, wherein said threshold value is derived from a map arranged to associate an impedance value with a temperature value and a charge state value.

The method according to one of claims 4 or 5, wherein a plurality of impedance values are estimated for a plurality of respective components of the battery,

characterized in that

for said at least one component, said threshold value is obtained from at least one other impedance value corresponding to at least one other respective component - such as the average value of the impedance calculated on all the components equivalent in the pack.

7. Method according to one of claims 1 to 6, wherein the battery comprises a plurality of modules electrically connected to each other by bus bars, each module comprising a number of cells, characterized in that

an impedance value is estimated for a component comprising a busbar electrically connecting two modules together,

determining from said estimated impedance value a phase shift value,

for at least one other component comprising a cell, an impedance value of said component is estimated by taking into account said phase shift value.

8. An impedance estimation device (12) of at least one component (C'I, ..., C'N) of a motor vehicle battery (10), comprising:

- Mains charge control means for applying to said component a current signal coming from the sector and defining a harmonic frequency, in order to charge said component,

means for receiving at least two voltage values measured at the terminals of said component at times defining a period of time less than or equal to half of a period corresponding to the harmonic frequency, and

processing means in communication with the charging means and the receiving means, and arranged to determine from the received voltage values and from the applied current signal an impedance measurement value of the component associated with the frequency harmonic.

9. Battery system (1) comprising the device (12) according to claim 8, and a charging device (1 1) for charging a battery (10) from the mains and / or the battery itself.

10. Motor vehicle comprising a battery system (1) according to claim 9.