FR3013459A1 - METHOD OF ESTIMATING THE VALUE OF A CHARACTERISTIC OF AN ELECTROCHEMICAL CELL - Google Patents

Abstract

The invention relates to a method for estimating the value of a characteristic of an electrochemical cell (41) of accumulators being charged or discharged. According to the invention, this method comprises steps of: a) acquisition of a voltage signal (U1) at the terminals of said cell, b) determination of an AC component (U2) of the voltage signal (U1) acquired in step a), and c) evaluating the value of said characteristic as a function of the determined alternative component.

Description

TECHNICAL FIELD TO WHICH THE INVENTION RELATES The present invention generally relates to the management of charging and discharging cycles of a storage battery. It relates more particularly to a method for estimating the value of a characteristic of an electrochemical cell of accumulators during charging or discharging, such as the charge level of this cell, comprising a step a) of acquisition a voltage signal across said cell during charging or discharging. The invention applies particularly advantageously to motor vehicles equipped with an electric motor powered by a storage battery (called traction) comprising a plurality of cells. BACKGROUND ART In a well-known manner, the electric power that can be provided by a storage battery decreases during a discharge cycle. To predict when it will be necessary to charge the battery and to draw the best part of the available electrical power, it seeks to determine different characteristics of the battery. One seeks in particular to determine its level of charge. The battery charge level, which is usually expressed as a percentage, reflects the battery charge level between a minimum charge level where the battery is no longer usable, and a maximum charge level. This level of charge is determined by calculating the charge level of each cell making up the accumulator battery. One method for determining the charge level of a cell is to use two easily measurable physical quantities that are the DC voltage across the cell and the current flowing in the cell. Thus, it is known to evaluate the charge level of a cell being discharged or charged, using a predetermined curve associating a charge level value with each measured DC voltage value. The evaluation of this charge level is then refined using the value of the intensity of the current measured at the input or at the output of the cell. This method unfortunately has a random precision, except to use expensive electronic components to obtain a voltage measurement having a precision of the order of millivolt over a range of a few volts (0 to 5 volts). Even using such components, the accuracy of measurements can still be affected by external phenomena such as temperature variations. Another method described in US6313607 is to excite the cell with a dedicated device generating a non-continuous current, to examine the behavior of the cell in response to this stimulus, and to deduce various characteristics such as its impedance.

The major disadvantage of this method is that its implementation is complex and expensive. OBJECT OF THE INVENTION In order to overcome the aforementioned drawbacks of the state of the art, the present invention proposes to evaluate the value of a characteristic of a cell 15 by examining a physical quantity other than the DC voltage across the terminals of the cell. the cell and the current flowing in the cell. More particularly, the invention proposes an estimation method as defined in the introduction, in which steps are provided: b) determining an AC component of the voltage signal acquired in step a ), and c) evaluating the value of said characteristic as a function of the determined AC component. By "step of determining an AC component of the voltage signal" is meant a step of isolating the AC component of the voltage signal. When charging or discharging a cell, it is known that electrochemical phenomena generate acoustic signals. In Lithium-Ion type cells, it is believed that these acoustic signals are due to insertion and de-insertion phenomena of lithium in graphite. This is for example explained in detail in the document "Evaluation of acoustic emission as a suitable tool for aging characterization of LiAl / LiMn02 cell, N. Kircheva, S. Genies, C. Chabrol, P.-X. Thivel, Electrochimica Acta 88 (2013) 488494 ".

Thanks to the presence of these acoustic signals, the applicant has thought to find if the voltage delivered to the terminals of the cell, which is predominantly continuous, had an alternating component. Having succeeded in isolating this alternative component, the Applicant has discovered that this alternative component incorporates exploitable information relating to the cell. Thus, for example, the applicant could determine the level of charge of the cell using this single AC component. Thanks to the invention, it is no longer necessary to accurately measure the DC voltage across the cell to obtain the characteristics of the cell. On the contrary, it is only necessary to measure the AC component of the voltage at the terminals of the cell, which proves to be simple and inexpensive in that the AC voltage considered varies around 0 volts (it is more difficult to measure precisely a voltage that varies in a range of several volts). Thus, the invention makes it possible to obtain precise results which are furthermore insensitive to temperature variations. Other advantageous and non-limiting features of the method according to the invention are the following: said characteristic comprises the state of charge of said cell; in step b), said AC component is determined by filtering the voltage signal acquired in step a) by means of a high-pass filter; in step c), the alternative component determined in step b) is amplified with a gain greater than 50 in absolute value; in step c), said characteristic is evaluated using an output signal which corresponds to said AC component cut in a determined frequency band; said output signal is obtained by filtering said AC component 30 by means of a bandpass filter; in step c), said characteristic is evaluated as a function of the rms value of said output signal; - prior to the evaluation of said characteristic, there is provided a removal operation of the rms values considered aberrant; during said elimination operation, a substantially constant reference signal is determined by cutting off said AC component in a second frequency band distinct from said frequency band, the rms value of said reference signal is calculated, the times are determined; to which the rms value of the reference signal has an amplitude peak, and excluding the rms values of said output signal calculated at said times; and - said characteristic is determined using a predetermined function or a predetermined table, corresponding to each effective value of the output signal a value of the characteristic.

DETAILED DESCRIPTION OF AN EXEMPLARY EMBODIMENT The following description with reference to the accompanying drawings, given as non-limiting examples, will make it clear what the invention consists of and how it can be achieved. In the accompanying drawings: FIG. 1 is a schematic view of a motor vehicle equipped with a storage battery comprising a plurality of cells, and a calculator adapted to implement a method of estimating the state of charge of one of the cells; FIG. 2 is a diagram of the equivalent circuit proposed for one of the cells of FIG. 1 in operation, that is to say in charge or in discharge; FIG. 3 is an electrical diagram of the filters applied to the voltage measured at the terminals of the cell of FIG. 2, making it possible to obtain an output signal; FIG. 4 is a graph illustrating the variations of the rms value of the output signal as a function of time; FIG. 5 is a graph illustrating the variations of the rms value of a reference signal as a function of time; FIG. 6 is a graph illustrating the variations of the corrected effective value of the output signal as a function of time; and FIG. 7 is a graph illustrating the variations of the state of charge as a function of the rms value of the output signal. In Figure 1, there is shown very schematically a motor vehicle 1. This motor vehicle 1 is here an electric vehicle. It then comprises an electric motor 10 designed to drive its driving wheels 20 in rotation. In a variant, it could be a hybrid vehicle comprising an internal combustion engine and an electric motor for driving its driving wheels.

As shown in Figure 1, the motor vehicle 1 comprises a storage battery, called traction battery 40. This traction battery 40 is here exclusively intended to supply the electric motor 10 with current. Alternatively, this traction battery could also be provided for supplying current to various current-consuming electrical appliances, such as the power steering system, the air-conditioning system, etc. This traction battery 40 comprises an outer housing 43 projecting from two positive 44 and negative 45 terminals are emerge connected to the electric motor 10 via an electronic power unit (not shown).

The traction battery 40 also comprises a plurality of cells 41 which are housed in the outer casing 43 and which are here connected in series between the two positive and negative 44 terminals. Alternatively, it could be provided that the cells are connected in pairs. in parallel, and that these pairs of cells are connected in series between the two positive and negative terminals. The number of cells 41 used is determined so that the electric motor 10 can develop sufficient torque and power to propel the motor vehicle 1 for a predetermined time. A traction battery cell usually delivering a voltage of the order of 3 to 5 V, the cell number 41 is then calculated such that the voltage across the traction battery 40 can reach 400 V. At this point it will be noted that the term "cell" means an electrochemical element capable of storing electrical energy when it is supplied with current by an external electrical network, and then subsequently restoring this electrical energy. Here, the cells 41 considered are lithium-ion batteries, but it may be an alternative to any other type of cells (lead, nickel-cadmium, nickel-metal hydride, ...). Classically, each cell 41 comprises an envelope which encloses an electrolyte and from which two electrodes protrude.

In the remainder of this talk, we will only be interested in one of these cells 41, the other cells being treated in the same way. FIG. 2 shows an equivalent diagram of this cell 41 when it supplies the electric motor 10 and delivers a voltage U1 to it.

The model of the electric circuit used in the present invention consists in considering that the cell 41 is formed of three distinct elements connected in series, of which: a DC voltage source 50 at the terminals of which a voltage UO is measured; an alternating voltage source 51 across which a voltage U2 is measured; and an internal impedance 52. The voltage U0 corresponds to the DC component of the voltage U1, whereas the voltage U2 corresponds to the AC component of this voltage U1. The cell 41 has a level of charge (or "state of charge") noted SOC (acronym for "State Of Charge"), which evolves during its charging or discharge. This SOC load level is expressed as a percentage. It reflects the state of loading of the cell 41 considered, between a minimum charge level where the battery is no longer usable (0%), and a maximum charge level (100%). It is therefore essential, and this is also one of the objects of the present invention, to precisely determine this SOC load level in order to inform the driver of the remaining battery life of the battery 40. 25 The motor vehicle 1 comprises for this purpose a computer 30, which is here shown as being independent of the traction battery 40. This computer is preferably however integrated in the battery. Here, as shown in FIG. 1, the computer 30 comprises a processor (CPU), a random access memory (RAM), a read-only memory (ROM), analog-digital converters (A / D), and various interfaces. entry and exit. Thanks to its input interfaces, the computer 30 is adapted to receive different sensors input signals relating to the traction battery 40. In its random access memory, the computer 30 thus continuously stores: the temperature T of the traction battery 40, measured here by means of a temperature sensor located in the outer casing 43 of the traction battery 40, the intensity I of the current delivered by the traction battery 40, using an ammeter electrically connected between the cells 41 and the negative terminal 45 of the traction battery 40, and an output voltage U4, obtained by filtering the voltage U1 measured at the terminals of the cell 41, by means of a system analog filter 100. Thanks to software stored in its read-only memory, the computer 30 is adapted to determine, as a function of the output voltage U4, the charge level SOC of the cell 41. Finally, thanks to its output interfaces , the callus The culator 30 is adapted to transmit this SOC charge level to the global driving unit of the electric motor 10, so that the latter can display this level of charge in the field of view of the driver of the vehicle. In the remainder of this talk, the architecture of the analog filtering system 100 and the method that this analog filtering system 100 and the software are adapted to implement for determining the SOC charge level will be detailed.

According to a particularly advantageous characteristic of the invention, this analog filtering system 100 and this software are adapted to implement a method of estimating the value of a characteristic of the cell 41 during charging or discharging, which comprises steps of: a) acquisition of the voltage signal U1 at the terminals of the cell 41, b) determination of an alternating component U2 of the voltage signal U1 acquired in step a), and c) evaluation the value of said characteristic as a function of the determined alternative component U2. In the example illustrated here, the characteristic of the cell 41 to be determined is the state of charge SOC of said cell 41. Alternatively, as will be described at the end of this presentation, it will be possible to act of another characteristic. The step a) of acquiring the voltage signal U1 at the terminals of the cell 41 may be made in various ways. It can thus be achieved by means of any voltmeter connected in parallel with the cell 41. It may in particular be carried out using an external sound card on USB port, such as for example the "Sound Blaster X-Fi HD" card. In 24-bit 96 kHz mode, once calibrated in amplitude with a Tektronix AFG3102 signal generator on a 50-ohm through load. This acquisition could also have been performed at a lower sampling frequency, provided that this frequency is in line with the electrochemical technology and with the cell type 41. Thus, for the cell 41 considered here, a sampling frequency 6 kHz would have been sufficient. In step b), the alternating component U2 of the voltage signal U1 is obtained by filtering the voltage signal U1 by means of a high-pass filter 105. As shown in FIG. 3, the analog filtering system 100 comprises for this purpose a capacitor which, connected in series with the cell 41, has a high pass filter function. This capacitor thus makes it possible to isolate the alternating component U2 from the voltage signal U1, by eliminating its DC component U0. The alternating component U2 of the voltage signal U1 oscillates around 0 volts. It is then observed that the measurement of this alternative component U2 is not very sensitive to temperature variations and electromagnetic interference, which makes it possible to obtain an accurate signal at a reduced cost. In step c), an operation is first provided for amplifying the alternating component U2 of the voltage signal U1. Here, as shown in Figure 3, the analog filtering system 100 comprises for this purpose an amplifier 110 which is formed by an operational amplifier mounted in a non-inverter, whose gain is greater than 50 in absolute value. Here, the gain of this filter 110 is chosen equal to -1000. The operational amplifier used for its very low noise and wide bandwidth (several tens of kilohertz) is a Texas Instrument LME49990. Thus, at the output of this filter 110, an amplified signal U3 is obtained. As a variant, an operational amplifier mounted as a non-inverter could have been used. The amplified signal U3 does not provide usable information over its entire bandwidth.

There is thus at least a first frequency band AF1 in which it is possible to extract information relating to the state of charge SOC of the cell 41, and a second frequency band AF2 in which no such information is available. is exploitable.

Here, the first frequency band AF1 is between 100Hz and 1100Hz and the second frequency band AF2 is between 2000Hz and 3000Hz. Then, the amplified signal U3 is cut to isolate the first frequency band AF1 and thus obtain the output signal U4.

For this purpose, the analog filtering system 100 comprises a bandpass filter of order 2 or higher. Here, as shown in FIG. 3, this band-pass filter is obtained by serially mounting two operational amplifiers, including a first operational amplifier 120 mounted on a low-pass second-order and a second operational amplifier 130 mounted on a high-pass. Alternatively, one and / or the other of these two operational amplifiers could have an order greater than 2 to give the output signal U4 a better accuracy. In another variant, this filtering operation could have been carried out not with the aid of the analog filtering system, but rather with the aid of the computer 30, in digital form (by means of a band-pass filter or by a Fourier transform integrated in the calculator). In this variant, it will then be possible to first filter the amplified signal U3, in order to clip it to extract all the absurd values.

Here, the output signal U4 is in analog form and is digitized for processing by the computer 30. The computer 30 then uses the output signal U4 by calculating at a determined time interval (called period T) an effective value RMS (acronym English of "Root Mean Square").

Physically, this rms value corresponds to the value of the DC voltage which would cause the same power dissipation as the output signal U4. It is calculated by means of the following formula: ## EQU1 ## This effective value RMS is calculated at each period T. This period T is preferably chosen less than or equal to one minute, for example equal to at 6 seconds. The evolution of this RMS value as a function of time is shown in detail in FIG. 4, for a cell 41 of mark A123 of 2.3 Ah connected to a resistor of 4.7 ohms. For a better readability of this curve, the period T was chosen equal to one minute. As is clear from FIG. 4, the evolution of this RMS rms value is noisy in that it has outliers. Here, the RMS rms value is therefore not used gross to calculate the SOC charge level of the cell 41. On the contrary, it is preferentially provided for a preliminary operation to eliminate these outliers. For this, the evolution curve of the RMS value 15 of the output signal U4 (FIG. 4) is compared with a curve of evolution of the RMSref rms value of a reference signal U5 (FIG. 4). This reference signal U5 is derived from the analog filtering system 100, which makes it possible to be certain that it has undergone the same disturbances as the output signal U4. In practice, this reference signal U5 is obtained by cutting the amplified signal U3 in the second frequency band AF2, by means of a bandpass filter identical to that used to obtain the output signal U4. This reference signal U5 is then digitized and processed by the computer 30 which determines the effective value RMSref. It can be seen that the evolution of the effective value RMSref of the reference signal U5 is substantially constant, but that it nevertheless includes outliers which appear at the same times t1, t2, t3, t4, t5, t6, t7 as the outliers of the RMS rms value of the output signal U4. Since the evolution of the RMSref rms value is supposed to be constant, it is easier to identify the amplitude peaks and thus the times t1, t2, t3, t4, t5, t6, t7 above.

To identify these times, the computer 30, after calculating the effective value RMSref of the reference signal U5, calculates the difference between this effective value RMSref and the average of the ten RMSref effective values calculated above. If this difference is greater than a predetermined threshold, the computer 30 deduces that the effective value RMSref considered is aberrant and it is the same for the last calculated RMS value. To identify these instants t1, t2, t3, t4, t5, t6, t7, the computer 30 could of course proceed otherwise. It could thus operate by simple thresholding, by statistical filtering or by temporal filtering.

FIG. 6 shows the time evolution of the RMS rms value of the output signal U4, once the aberrant values have been suppressed by the computer 30 (this time with a period T of 6 seconds). This is called corrected RMScor effective value. It remains for the computer 30 to deduce from each corrected RMScor effective value a SOC charge state corresponding to the cell 41. For this, the computer 30 uses a predetermined function or a predetermined table stored in its read-only memory. This is a function Fi, shown in broken lines in Figure 7.

This function F1 is here a polynomial interpolation of order 2 of the curve C1 represented in solid lines in this same FIG. 7. This curve Ci, which associates with each corrected effective value RMScor a corresponding state of charge SOC, is obtained in the laboratory. (by technical solutions known but expensive, hardly embeddable in a vehicle). In summary, at each time step of period T, the computer 30 can then calculate the RMS rms, RMSref, and if the RMS rms value is not aberrant, calculate using the F1 function above the level of SOC load of cell 41.

The present invention is in no way limited to the embodiment described and shown, but those skilled in the art will be able to make any variant within their mind. Thus, even if the temperature variations have a limited influence on the results of the method used, it will however be possible to improve the estimate taking into account the temperature T measured. This can be done in various ways, as is otherwise well known to those skilled in the art, for example by means of a Kalman filter. Measured currents I can also be taken into account, for example by using the coulomb counting method. According to another variant, instead of the SOC state of charge, or in addition to the calculation of this SOC state of charge, the method described here could have been used to calculate other parameters of the cell 41, such as for example its maximum capacity Q or an aging indicator SOH (acronym for "State of Health"). It is recalled that the maximum capacity Q of the cell 41, which is generally expressed in ampere-hours, makes it possible to know the duration during which the cell can supply an electric current of a given amperage. This capacity deteriorates over time, depending in particular on the temperature history of the battery and its history of charging and discharging cycles. The SOH aging indicator is expressed as a percentage. It provides information as to the state of aging of the cell 41 considered. As a general rule, at the time of manufacture of the cell 41, the aging indicator SOH is equal to or slightly less than 100%, then decreases with time, depending on the use made of the cell 41. thus, for example, calculating these two parameters Q, SOH as a function of the state of charge SOC calculated according to the method described above, according to a method well known to those skilled in the art. According to another variant, it would have been possible to use the entire bandwidth of the amplified signal U3 to calculate the SOC state of charge value, in which case the result would have been less accurate.

Claims (10)

REVENDICATIONS1. A method of estimating the value of a characteristic (SOC) of an electrochemical cell (41) of accumulators being charged or discharged, comprising a step of: a) acquiring a voltage signal ( U1) at the terminals of said cell (41), characterized in that it then comprises steps: b) determining an AC component (U2) of the voltage signal (U1) acquired in step a), and c) evaluating the value of said characteristic (SOC) as a function of the determined alternative component (U2). 2. Estimation method according to the preceding claim, wherein said characteristic comprises the state of charge (SOC) of said cell (41). 15 3. estimation method according to one of the preceding claims, wherein, in step b), said alternative component (U2) is determined by filtering the voltage signal (U1) acquired in step a) by means of a high-pass filter (105). 4. Estimation method according to one of the preceding claims, wherein, in step c), the alternative component (U2) determined in step b) is amplified with a gain greater than 50 in absolute value. 5. estimation method according to one of the preceding claims, wherein, in step c), said characteristic (SOC) is evaluated using an output signal (U4) which corresponds to said AC component (U2) cut off 25 in a determined frequency band (AF1). 6. Estimation method according to the preceding claim, wherein said output signal (U4) is obtained by filtering said AC component (U2) by means of a band-pass filter (120, 130). 7. Estimation method according to one of the two preceding claims, wherein, in step c), said characteristic (SOC) is evaluated as a function of the effective value (RMS) of said output signal (U4). 8. Estimation method according to the preceding claim, wherein, prior to the evaluation of said characteristic (SOC), there is provided an operation for eliminating the effective values (RMS) considered commaberrantes. 9. Estimation method according to the preceding claim, wherein during said elimination operation: a substantially constant reference signal (U5) is determined by cutting off said alternating component (U2) in a second frequency band ( AF2) distinct from said frequency band (AF1), - the rms value (RMSref) of said reference signal (U5) is calculated, - the times (t1, t2, t3, t4, t5, t6, t7) are the effective value (RMSref) of the reference signal (U5) has an amplitude peak, and 10 - for the evaluation of said characteristic (SOC), the effective values (RMS) of said output signal (U4) are calculated at said times (t1, t2, t3, t4, t5, t6, t7). 10. The estimation method according to one of the three preceding claims, wherein said characteristic (SOC) is determined using a predetermined function or a predetermined array, matching each RMS value. of the output signal (U4) a value of the characteristic (SOC). 20