EP4449140A1

EP4449140A1 - Method for resetting the state of charge of a battery system

Info

Publication number: EP4449140A1
Application number: EP22801848.7A
Authority: EP
Inventors: Akram EDDAHECH; Kodjo Senou Rodolphe MAWONOU
Original assignee: Stellantis Auto SAS
Current assignee: Stellantis Auto SAS
Priority date: 2021-12-15
Filing date: 2022-10-17
Publication date: 2024-10-23
Also published as: CN118401848A; FR3130388B1; FR3130388A1; WO2023111409A1

Abstract

The present invention relates to a method for resetting the state of charge of an electrochemical battery system, comprising a phase of evaluating an authorisation for resetting following a relaxation phase, the method comprising the steps of: measuring (E1) a no-load voltage of the battery system following the relaxation phase; determining (E2) the duration of the relaxation phase; performing a first test (E3) consisting in comparing said duration with a first threshold (S1); performing a second test (E7) based on the measurement (E1) of the no-load voltage when said duration (D) is greater than the first threshold (S1); and determining the state (E8, E9) of a resetting authorisation (S_AUTH) based upon the result of the first test (E3) and the second test (E7), a first state (E8) of the authorisation authorising the resetting and a second state (E9) of the authorisation preventing the resetting.

Description

DESCRIPTION

Title: METHOD FOR RESETTING THE STATE OF CHARGE OF A BATTERY SYSTEM

The present invention claims the priority of French application No. 2113544 filed on 15.12.2021, the content of which (text, drawings and claims) is incorporated herein by reference.

The field of the invention relates to a method for recalibrating the state of charge of a battery system following a relaxation phase comprising a phase for evaluating a recalibration authorization.

In the automotive field, in particular electrified vehicles, the battery system comprises a control unit in charge of supervising the battery, and in particular continuously estimating the state of charge, commonly designated by the English acronym SOC for “State of Charge”. In addition to the importance of delivering reliable range information to the driver, it is essential for supervision because it guarantees the proper functioning of the battery, optimizes its use, as well as all the vital energy functions of the vehicle. Conventionally, when the battery wakes up, the state of charge is initialized to a value determined on the basis of a measurement of the no-load voltage, commonly designated by the acronym OCV for "Open Circuit Voltage". The open circuit voltage is the voltage across the battery terminals at zero current. The voltage stabilizes following a relaxation period lasting at least about thirty to forty minutes depending on battery chemistry and temperature. This stabilized voltage is interesting because it makes it possible to estimate the state of charge with precision by means of maps obtained during testing matching state of charge values and values of the no-load voltage.

Then, in operation, the state of charge is updated by current measurements and by calculation of the energy variation. However, after a certain period of operation and for certain driving profiles, of short duration for example or presenting current peaks, this calculation may be affected by a bias due to errors in the accuracy of the current sensors and the chain. digital information processing, in particular. It is therefore necessary to periodically update accurately the state of charge value to ensure optimal operation of the battery system. This operation, called retiming or initialization, is carried out by means of correspondence mapping and is generally carried out when the battery system wakes up.

The document FR3094796A1 is known describing a process for initializing the state of charge of a battery system. This method teaches to estimate a state of charge on the basis of the value delivered by the no-load voltage map as soon as it is detected that the relaxation duration is greater than a target duration. This operation is automatic because it is considered that the no-load voltage is in a condition to provide an accurate estimate. And, when the relaxation duration has not reached the target duration, the readjustment is carried out by taking into account an estimation of error between the value of the no-load voltage for the target duration and the value for the effective duration at initialization.

This last technique requires significant memory resources for recording error calibration maps. Moreover, for certain electrochemical cells, in particular of the Lithium Iron Phosphate (LFP) type, the calculation of the readjustment of the state of charge can be affected by a significant bias of the order of 15 to 30% of points, even when the target relaxation time has been reached. Indeed, the maps for this type of cell are characterized by plateau areas where a variation of half a millivolt leads to a variation in the state of charge of 10 to 15% of points. In addition, the correspondence maps vary significantly depending on the polarization of the current having passed through the battery before the relaxation phase. This difference also leads to a significant bias during registration.

There is therefore a need to overcome the aforementioned problems. One objective of the invention is to improve the accuracy of the state of charge value during a reset phase following a relaxation of the battery, in particular for LFP type batteries. Another object of the invention is to prevent resetting when the update value is likely to be affected by a significant bias. Another objective is to propose a technique for resetting the state of charge requiring less memory resources than the known solutions.

More specifically, the invention relates to a method for recalibrating a state of charge parameter of an electrochemical battery system comprising a phase evaluation of an authorization of the resetting following a relaxation phase of said battery system, said evaluation phase comprising the following steps: the measurement of an open circuit voltage of the battery system following the relaxation phase, the determination of the duration of the relaxation phase, the execution of a first test consisting in comparing said duration with a first threshold, the execution of a second test based on the measurement of the voltage at empty when said duration is greater than the first threshold and the determination of the state of an authorization of the resetting according to the result of the first test and of the second test, a first state of the authorization authorizing the resetting and a second state of the authorization inhibiting the readjustment.

According to a variant, the method further comprises a step of calculating an estimation error of the state of charge based on the measurement of the open-circuit voltage, the second test consists in comparing said error with a second threshold and the authorization is driven in the first state if the error is less than the second threshold and in the second state if the error is greater than the second threshold.

According to a variant, the step of calculating the state of charge estimation error consists in calculating a difference between a first state of charge and a second state of charge, the first and the second state of charge being delivered by a map of correspondence between the open circuit voltage and the state of charge.

According to a first error calculation mode, the first state of charge is determined from a first value of the measured no-load voltage and the second state of charge is determined from said first value and a margin predetermined error representative of the measurement accuracy.

According to a variant of the first calculation mode, the measurement step includes filtering of the open circuit voltage during the measurement and the first value is the average value of the open circuit voltage delivered by the filtering.

According to a second method of calculating the error, the first state of charge is determined from a second value of the measured no-load voltage representative of the maximum voltage measured and the second state of charge is determined from a third value of the measured no-load voltage representative of the measured minimum voltage.

According to a variant, the method further comprises a step of determining the polarization of the current flowing through the battery system prior to the relaxation phase and a step of selecting the map from among at least one charge map and one discharge map in function of polarization.

According to a variant, the method also comprises the following steps for determining the polarization of the current: during the relaxation phase, a step for measuring two successive values of the voltage, a step for comparing said two successive values, in the event of detection of a decreasing voltage, the recording of a first piece of information representative of a charging current polarization, and in the event of detection of an increasing voltage, the recording of a second piece of information representative of a current polarization discharge.

According to a variant, when the duration is less than the first threshold, the authorization is controlled in the second state.

Further contemplated is an electrified vehicle having an electrochemical battery system including a control unit, which control unit is configured to implement the retiming method according to any of the preceding embodiments.

The method according to the invention has the advantage of taking into account the undesirable effects affecting the accuracy of an open circuit voltage measurement. In particular, it avoids the effects of voltage hysteresis between a charge and a discharge prior to the relaxation phase, the effects of temperature on the estimation of the state of charge and finally the effects of plateau zones with very little voltage variation for no-load voltage maps. The invention thus improves the precision of an estimation of the instantaneous state of charge during an update following a relaxation phase. Finally, the invention avoids the registration operation if the latter is likely to present a significant bias. Other characteristics and advantages of the present invention will appear more clearly on reading the following detailed description comprising embodiments of the invention given by way of non-limiting examples and illustrated by the appended drawings, in which:

[Fig.1] represents an electrified motor vehicle comprising a battery system configured to implement the invention.

[Fig.2] represents a map of correspondence between an open circuit voltage and the state of charge for an electrochemical cell characterized by voltage plateaus in relaxation.

[Fig.3] is a graph illustrating the estimation of the state of charge error conditioning the readjustment authorization in accordance with the method according to the invention.

[Fig .4] represents an embodiment of the resetting method comprising a phase of evaluating a resetting authorization in accordance with the invention.

The invention applies to electrified vehicles, having an electric traction engine. The invention relates more specifically to the supervision of a lithium-ion type electrochemical cell battery system for estimating the state of charge parameter. In the present description, the term approximately means +/-10% of the indicated value and the limits of a range of values are included in the range.

In Figure 1, there is shown schematically a motor vehicle 1 comprising an electrochemical battery system 3 provided to implement the invention and a control unit 2 of the vehicle (commonly also called supervisor, ECU ("Electric Control Unit") or VCU ("Vehicle Control Unit") responsible for controlling the functional modules of the vehicle. The vehicle comprises a traction module (not shown) provided with an electric traction machine powered by the battery system 3.

The battery system 3 comprises a computer control unit 4, designated by the English acronym BMS for "Battery Management System" or TBCU for "Traction Battery Control Unit" and an electrical energy storage means 5 comprising at least one electrochemical cell, for example of the Lithium-ion type. The method according to the invention applies in this preferred embodiment to cells of the Lithium Iron Phosphate type. The invention applies to other types of chemistry, for example of the Nickel Manganese Cobalt (NMC), Nickel Cobalt Aluminum (NCA) type, or even of the Lithium Manganese Oxide (LMO) type.

Conventionally, in the case of a vehicle application comprising an electric or fully electric hybrid automotive powertrain, the battery system 3 is equipped with a module or a plurality of modules allowing the storage and the return of electrical energy . The or each module comprises a single electrochemical cell or a plurality of electrochemical cells. The modules can be electrically connected in series and/or in parallel depending on the desired electrical specifications.

The control unit 4 is suitable for supervising the parameters specific to the battery by means of current and voltage sensors, such as the state of charge expressed by a ratio between the quantity of energy stored at a given moment and the quantity of maximum energy that can be stored at a given time, for example as a percentage, the open circuit voltage expressed in Volts, the charging current expressed in Amperes, the state of health SOH (State of Health), which designates the level parameter aging of the battery expressing a ratio between the maximum quantity of electricity storable at a given instant and the maximum quantity of electricity storable when the battery is new. The control unit 4 further comprises temperature sensors provided for measuring the temperature of the electrochemical cells or a temperature estimator capable of delivering information for the entire battery system.

The control unit 4 is capable of determining an off-load voltage value from measurements taken at the terminals of a cell, in particular to determine the state of charge of the cell. In one embodiment, the control unit is able to select the maximum voltage or the minimum voltage of a group of cells or of all the cells forming the battery 3. This with the aim of choosing an off-load voltage for a cell group during the implementation of the method.

In addition, the control unit 4 includes means for estimating the state of charge while driving from an initial state of charge value and measurements of discharge and recharge current flowing through the battery. Thus, the state of charge is continuously estimated as a function of the instantaneous energy variation. The control unit 4 further comprises means for resetting the state of charge based on a predetermined map of correspondence taking as input an open circuit voltage parameter and delivering an estimate of the state of charge. A correspondence map between the open circuit voltage and the state of charge is recorded in the memory of the control unit 4.

In addition, in order to obtain an accurate estimate as a function of the instantaneous operating conditions of the battery, the control unit 4 is capable of specifically selecting a map as a function of the temperature of the system, of the direction of polarization of the current having passed through the battery before the relaxation phase, and a battery aging state parameter. To this end, the control unit 4 can comprise in memory several correspondence maps selected specifically according to the polarization, the temperature and the state of aging of the battery. These maps are obtained during testing during the calibration of the battery system and stored in the memory of the control unit 4.

By way of example, in FIG. 2, a correspondence map recorded in the memory of the control unit 4 has been schematically represented. This is a map specific to the temperature conditions of the battery system and to a current polarization. The map combines no-load voltage values, expressed in volts on the ordinate axis, and state of charge values on the abscissa axis, expressed as a percentage. Conventionally, a map is drawn up during a characterization test of an electrical cell or the battery system and is drawn up in accordance with a standard procedure by the supplier of the electrochemical cell or by the manufacturer of the battery system.

As can be seen in FIG. 2, for an LFP-type cell, there are plateau areas, surrounded by dotted lines, where the state-of-charge values are particularly sensitive to the precision of the open-circuit voltage. In the 30% to 60% state of charge and 70% to 95% state of charge zones, a difference of 0.5 millivolts leads to a variation in the estimated state of charge of approximately 10 % to 15%. If the measurement sensors and the processing chain deliver an off-load voltage value with an accuracy of a few millivolts, the mapping is likely to deliver a state of charge value with a variation of up to 30% depending on the operating zone . A readjustment under these conditions is not desirable. The invention has the advantage of evaluating the accuracy of the estimation of the state of charge to authorize or inhibit the resetting operation following a relaxation phase. To this end, the invention implements a phase of evaluation of a resetting authorization which takes into account a margin of error in measuring the off-load voltage, in addition to the temperature of the battery system, the direction polarization of the current before the relaxation phase and the duration of relaxation.

Remember that a relaxation phase is a phase during which the current flowing through the battery is zero. A relaxation phase is, for example, when the battery falls asleep when the vehicle is stationary, parked and the vehicle switched off. The battery system is off. In another relaxation situation, the battery system is on, but there is no current flowing through the battery. The battery system operates in a so-called “exhibition mode” mode used for example in trade fairs and exhibitions. The battery system is energized but remains under zero current.

The duration of a relaxation phase makes it possible to accurately estimate the state of charge and must generally be at least greater than approximately 30 or 40 minutes depending on the temperature conditions.

The control unit 4 contains in memory a duration threshold, expressed in seconds or minutes, the function of which is to authorize or not the readjustment in accordance with the method according to the invention. The threshold value is between 30 minutes and 180 minutes for example. The threshold value is dependent on the instantaneous temperature according to a model developed during testing and which is recorded in the memory of the control unit 4.

The control unit 4 is capable of determining the duration of a relaxation phase, for example by means of the vehicle clock running permanently at the level of the control unit 2 or 4. At the start of a phase relaxation, when the battery system is switched off for example, the control unit 4 collects a first time value, then collects a second value when the resetting authorization evaluation phase is triggered. From these two values, the control unit determines the relaxation time used by the method according to the invention. Alternatively, the relaxation time can be determined using a timer triggered when the battery system is turned off. The control unit 4 is also capable of determining the polarization of the current having passed through the battery prior to a relaxation phase. The polarization is determined at the beginning of the relaxation phase on the basis of the measurement of two successive values of the voltage before the battery 3 falls asleep. Indeed, following a discharge current, the no-load voltage tends to stabilize after a transient increase. Following a recharge current, the no-load voltage tends to stabilize after a transient decrease. The detection of the transient variation of the voltage makes it possible to record a parameter representative of the polarization. A variant is envisaged in which the determination of the polarization is carried out when the battery wakes up during the evaluation of the authorization of the resetting. Another variant provides for detecting the sign of the current just before the relaxation phase and for recording polarization information representative of the sign.

The control unit 4 further comprises means for estimating the error in calculating the state of charge on the basis of the correspondence map. To this end, the control unit 4 contains in memory a parameter of the margin of error for the measurement of the no-load voltage. The margin of error parameter is a calibratable variable representing the precision of the open-circuit voltage measurement, expressed in millivolts. The parameter takes into account the material constraints, for example the accuracy of the measurement sensors and the accuracy of the voltage information processing chain. The margin of error parameter is configured to a value between 0.1 mV and 5 mV. In a preferred embodiment, the margin of error parameter is configured at 3mV.

The estimation of the state of charge calculation error is calculated by measuring the difference between two state of charge values delivered by the map, for example the difference between the average value of the voltage at vacuum measured and the average value integrating the margin of error parameter. The invention advantageously uses the maps already used for estimating the state of charge. The use of additional maps specific to an error calibration is avoided and memory resources are thus saved.

In FIG. 3, a graph is shown illustrating the estimation of the error delivered by the correspondence map for two levels of margin of error of an open circuit voltage measurement. The ordinate axis represents the SOC_err error of the state of charge, expressed in % of SOC, and the abscissa axis represents the level state of charge of the battery corresponding to the measured no-load voltage. Two curves C1 and C2 are shown and illustrate the SOC_err error resulting from an estimate for a mapping. C1 illustrates the error for the no-load voltage referenced OCVm to which a positive error margin ERR is added, here with a value of around +3mV. C2 illustrates the error for the same voltage measurement OCVm from which the margin of error ERR is subtracted, this time negative, here with a value of about -3mV.

The dotted horizontal lines represent an error threshold conditioning the readjustment authorization. In this non-limiting example, the threshold is calibrated at 4%. As seen in the graph, plateau areas on the ranges of approximately 30%-60% of SOC and 70%-95% yield an estimate of SOC error above the 4% threshold. Above the threshold, the recalibration is inhibited. Below this threshold, readjustment is authorized.

It should be noted that the estimation of the error is dependent on each map selected according to a charging current, discharge current, temperature and state of aging, and the calibrated margin of error.

Finally, the control unit 4 comprises means for authorizing the readjustment of the state of charge. The authorization is for example a signal, authorizing in a first state or inhibiting in a second state, the readjustment of the state of charge calculated on the basis of the correspondence map. The state of the signal is dependent on the value of the error with respect to the calibrated threshold, for example 4%. The signal can be a boolean signal. An authorization is provided for a cell, for each cell, for a group of cells, or for all the cells of the battery system. When a clearance is evaluated for a cell group, the open-circuit voltage measurement used for the clearance evaluation is chosen by selecting the maximum or minimum open-circuit voltage of the cell group as the input parameter taken into account. account by match mapping.

In FIG. 4, the process for resetting the state of charge according to the invention has been represented. The method is implemented by the control unit 4 of the vehicle battery system. The control unit is provided with an integrated circuit computer and electronic memories, the computer and the memories being configured to execute the resetting method according to the invention. But this is not mandatory. Indeed, the computer could be external to the control unit 4, while being coupled to the latter 4. In the latter case, it can itself be arranged in the form of a dedicated computer comprising a possible dedicated program, for example. Consequently, the control unit, according to the invention, can be produced in the form of software (or computer (or even “software”)) modules, or else of electronic circuits (or “hardware”), or even of a combination of electronic circuits and software modules.

The readjustment of the state of charge generally takes place when the battery system wakes up. It is contemplated that it can be operated on following any battery relaxation phase.

Before the relaxation phase, during a shutdown of the battery system control unit, the latter executes preliminary operations for collecting information to calculate the relaxation time on waking and operations for determining the current polarization.

To this end, the method comprises the measurement of a time reference of the start of the relaxation phase allowing the calculation of the duration of relaxation, for example the information of a clock or the triggering of a time delay. When you wake up, a new time measurement will be performed.

In addition, the method comprises the measurement of at least two successive values of the open circuit voltage at the terminals of one or each cell, or of a group of cells of the battery. The variation in voltage at the onset of relaxation indicates the polarization of the current, a decreasing current indicating a charging current and an increasing current indicating a discharging current. Optionally, before the current stops, information representative of its polarization is stored.

Once these preliminary measures have been taken, the battery system is shut down, or in exposure mode. Optionally, the electrochemical cells are electrically disconnected from the consuming electrical systems.

Then, upon awakening, the control unit initiates the resetting process. Before powering up, the method provides, in accordance with the invention, a phase for evaluating authorization of the resetting. The evaluation phase comprises a first measurement step E1 of an off-load voltage of the battery system following the relaxation phase. The E1 measurement is made after falling asleep, i.e. upon waking up at the moment the evaluation phase is triggered when the battery is still at zero current. The measurement is preferably carried out for a predetermined duration of approximately 1 second to 2 seconds. The E1 measurement consists of measuring the off-load voltage, for at least one cell, each cell or a group of cells of the battery system.

During the measurement E1, the method preferentially provides for the filtering of the measured voltage. The filtering is for example a calculation processing of the moving average or weighted average over the duration of the measurement. The function of filtering is to eliminate noise on the measurement and to improve the accuracy of the open-circuit voltage value for state-of-charge calculations based on the map.

If necessary, the measurement E1 is also used to determine the polarization of the current during the operating phase of the battery having preceded the relaxation phase.

At a second step E2, the method comprises determining the relaxation duration D. To this end, when the battery wakes up, the method comprises a step of collecting time information, for example information from the clock of the vehicle supervisor or a battery system clock. The duration of relaxation D corresponds to the period delimited by the instant before falling asleep and the instant of awakening.

Then, in a third step E3, the method comprises a first test E3 consisting in comparing the relaxation duration D with a duration threshold S1 representative of a minimum relaxation duration to carry out a readjustment of the state of charge with sufficient accuracy. The threshold is between 30 minutes and 180 minutes. The value of the threshold S1 is determined according to the instantaneous temperature T of the battery system. A model f stored in the memory of the control unit is consulted to estimate the threshold S1, where S1 =f(T). The temperature T is collected by means of a battery system sensor, an external temperature sensor or an estimator, for example. The model is developed in test, then saved in memory of the battery system. By way of non-limiting example, at cold temperatures, for example close to zero degrees Celsius or less, the threshold S1 can reach a value of 60 to 90 minutes. At a temperature of approximately 25 to 40 degrees Celsius, the threshold S1 is between 40 and 60 minutes. For extremely cold temperatures, the threshold can reach a value of 90 to 180 minutes. If it is detected that the relaxation duration D is less than the threshold S1, the process inhibits the resetting and the resetting process goes directly to step E9, which ends the authorization evaluation phase. The result of the test is indicated “NOK” in FIG. 4, for “Not OK” in English. At this step E9, the recalibration authorization signal S_AUTH is driven into a recalibration inhibition state. It is judged that the no-load voltage is not sufficiently stabilized to accurately estimate the state of charge. The authorization signal S_AUTH, in this example a Boolean signal, is driven to the low state or state “0”. The battery system control unit, once the evaluation phase is complete, retains the last state of charge value stored before falling asleep.

If it is detected that the relaxation time D is greater than the threshold S1 (the test result is indicated "OK" in FIG. 4), the evaluation phase includes a second test phase corresponding to steps E4, E5, E6 , aimed at calculating the estimation error of the state of charge SOC_err from the measurement E1 of the no-load voltage. And finally, this evaluation phase includes the second test E7 which determines the status of the registration authorization.

The second test phase aims to check whether the cells are operating at an electrical voltage level corresponding to a plateau, for example. This situation is likely to alter the accuracy of the estimation of the state of charge from the correspondence mapping. In such a case, provision is made to inhibit the recalibration. Otherwise, the method authorizes the readjustment.

More specifically, step E4 has the function of determining the polarization of the current having passed through the battery prior to the relaxation phase. According to one mode, the polarization is identified by comparing two successive measurements of a cell voltage of the battery to detect whether the voltage has varied upwards or downwards. At this stage, the control unit consults the polarization information parameter which was recorded when the system was put to sleep. Alternatively, the comparison of the two voltage values is carried out on waking to determine the polarization. In another mode, the polarization information is obtained by identifying the sign of the current flowing through the battery before the relaxation phase. The bias information is then used to select the state-of-charge and open-circuit voltage correspondence map which is used to calculate the state of charge error. Step E4 can optionally be executed during measurement E1.

More specifically, the method further comprises a step E5 of selecting the correspondence map as a function of the polarization information determined in step E4 and the temperature information T of the battery system at the time of the evaluation of authorization. For a discharge current bias, the method selects a correspondence map between the state of charge and the no-load voltage which is specific to a discharge current. For a load current bias, the method selects a load current specific mapping.

The selection has the advantage of taking into account the voltage hysteresis resulting from the bias and the temperature to improve the accuracy of the calculation of the state of charge. This improves the robustness of the evaluation phase of the registration authorization.

Then, the method further comprises a step of calculating E6 the estimation error of the state of charge SOC_err, where the calculation E6 consists in calculating a difference between a first state of charge SOC1 and a second state of charge SOC2 . The two levels of charge states SOC1 and SOC2 are delivered by the correspondence map which was selected during step E5. This difference is then compared with the calibrated state-of-charge error threshold S2, for example at 4% of state-of-charge points.

In a first calculation mode, the first state of charge SOC1 is determined from the average value 0CV_m of the open circuit voltage measured during step E1 and the second state of charge SOC2 is determined from this average value 0CV_m and the margin of error ERR representative of the measurement precision. The average value 0CV_m is the value resulting from the filtering of step E1.

More precisely, SOC1 is equal to the following relation, SOC1 = g(OCV_m), and, SOC2 is equal to the following relation, SOC2 = g(OCV_m + ERR), where g is the correspondence function of the mapping used for the calculation, 0CV_m is the mean measured value of the no-load voltage and the margin of error, here 3mV. In addition or alternatively, SOC2 is equal to the following relation, SOC2= g(OCV_m - ERR). Preferably, to determine the error estimate, the method selects the maximum value in absolute value between:

[SOC1 - g(OCV_m + ERR)] and [SOC1 - g(OCV_m - ERR)]

According to a variant of the first mode of calculation, it is envisaged that SOC1 can be determined from the maximum value or the minimum value of the open circuit voltage measured during step E1.

In a second calculation mode, the state of charge error estimate SOC_err is determined by the mapping from the difference between the value SOC1 corresponding to the maximum value of the measured no-load voltage, and the value SOC2 corresponding to the minimum value measured during step E1. In another variant, SOC1 is the mean value and SOC2 is one of the peak values of the measured no-load voltage, maximum or minimum. Thus, the invention makes it possible to establish an error specific to each cell or each group of cells. This has the advantage of taking into account the manufacturing deviation of the cells.

Then, the method includes the step of the second test E7 which consists in comparing the error SOC_err calculated in the previous step with the second threshold S2 of state of charge to determine the state of the authorization.

If the SOC_err error is less than the threshold S2, then the method authorizes in step E8 the readjustment of the state of charge. The result of the test is indicated "OK" in Figure 4. The authorization signal S_AUTH, in this example a Boolean signal, is driven high. The battery system control unit, once the evaluation phase is complete, activates the readjustment of the state of charge from the correspondence map and the E1 measurement of the no-load voltage. The invention thus improves the robustness of the state of charge information.

To illustrate this situation, with regard to figures 2 and 3, this case occurs, after a sufficient relaxation time, for a complete recharge reaching 95%-100% of the state of charge of the battery, for a state of charge lower than 30%, or between 60 and 70%.

If the error is greater than S2, then the method inhibits the resetting at step E9. The result of the test is indicated "NOK" in Figure 4. It is estimated that the estimation error is too large, at least with the risk of being greater than the estimation deviation based on consumption. electric. The signal authorization S_AUTH, in this example a Boolean signal, is driven low. The battery system control unit, once the evaluation phase is complete, retains the last stored state of charge value.

Provision is made according to the invention to implement an authorization signal S_AUTH for one cell or each cell of the battery system. It is envisaged that a single authorization signal is implemented for a group of cells. The measured no-load voltage is then the maximum or minimum voltage chosen from among the voltages of the group of cells or a reference voltage of a single cell of the group.

The invention applies to any battery-powered device, electrified vehicles with electric traction, car, truck, bus, motorcycle, bicycle, scooter, or even aircraft comprising an on-board electrochemical battery system.

Claims

1 . Method for recalibrating a state of charge parameter of an electrochemical battery system (3) comprising a phase for evaluating an authorization for recalibration following a relaxation phase of said battery system (3) , said evaluation phase comprising the following steps:

- the measurement (E1) of an off-load voltage of the battery system (3) following the relaxation phase,

- the determination (E2) of the duration (D) of the relaxation phase,

- the execution of a first test (E3) consisting in comparing said duration (D) with a first threshold (S1),

- the method being characterized in that it further comprises the execution of a second test (E7) based on the measurement (E1) of the no-load voltage when said duration (D) is greater than the first threshold (S1 ) and the determination of the state (E8, E9) of an authorization of the resetting (S_AUTH) according to the result of the first test (E3) and of the second test (E7), a first state (E8) of the authorization authorizing the resetting and a second state (E9) of the authorization inhibiting the resetting.

2. Resetting method according to claim 1, characterized in that it further comprises a step of calculating (E6) a state of charge estimation error (SOC_err) based on the measurement (E1) of the no-load voltage, in that the second test (E7) consists in comparing said error (SOC_err) with a second threshold (S2) and in that the authorization (S_AUTH) is controlled in the first state (E8) if the error is less than the second threshold (S2) and in the second state (E9) if the error is greater than the second threshold (S2).

3. Resetting method according to claim 2, characterized in that the step of calculating (E6) the state of charge estimation error (SOC_err) consists in calculating a difference between a first state of charge ( SOC1) and a second state of charge (SOC2), the first and second states of charge (SOC1, SOC2) being delivered by a map of correspondence between the open circuit voltage and the state of charge.

4. Resetting method according to claim 3, characterized in that the first state of charge (SOC1) is determined from a first value (OCV_m) of the measured no-load voltage and in that the second state of charge ( SOC2) is determined from said first value (OCV_m) and from a predetermined margin of error (ERR) representative of the measurement precision.

5. Resetting method according to claim 4, characterized in that the measurement step (E1) comprises a filtering of the no-load voltage during the measurement (E1) and in that the first value (OCV_m) is the average value of the no-load voltage delivered by the filter.

6. Resetting method according to claim 3, characterized in that the first state of charge (SOC1) is determined from a second value of the measured no-load voltage representative of the maximum voltage measured and in that the second state load (SOC2) is determined from a third value of the measured no-load voltage representative of the measured minimum voltage.

7. Resetting method according to any one of claims 3 to 6, characterized in that it further comprises a step of determining (E4) the polarization of the current flowing through the battery system (3) prior to the phase of relaxation and a step (E5) of selecting the map from among at least one charge map and one discharge map as a function of the polarization.

8. Resetting method according to claim 7, characterized in that it further comprises the following steps to determine the polarization of the current:

- during the relaxation phase, a step of measuring two successive values of the voltage,

- a step of comparing said two successive values,

- in the event of detection of a decreasing voltage, the recording of a first item of information representative of a charging current polarization, and in the event of detection of an increasing voltage, the recording of a second piece of information representative of a discharge current polarization.

9. Resetting method according to any one of claims 1 to 8, characterized in that when the duration (D) is less than the first threshold (S1), the authorization is controlled in the second state (E9).

10. Electrified vehicle (1) comprising an electrochemical battery system (3) comprising a control unit (4), characterized in that the control unit (4) is configured to implement the resetting method according to any of claims 1 to 9.