A portable method for analyzing the derivative of an electrical quantity of at least one electrochemical element, including the method and associated devices.

Abstract

Embedded method for analyzing the derivative of an electrical quantity of at least one electrochemical element, method and associated devices. The present invention relates to a method for analyzing the derivative of a first electrical quantity relating to at least one electrochemical element (12) with respect to a second electrical quantity during a charge or discharge at constant current or constant power, the method comprising the steps of: - obtaining a time signal of the first electrical quantity, - sampling the signal according to a first sampling time interval, - filtering the first sampled signal, - sampling the filtered signal according to a second sampling time interval strictly greater than the first sampling time interval, - differentiating the second sampled signal with respect to the second electrical quantity.and - detection of a value of a parameter relative to at least one predefined pattern in the derived signal. Figure for the abbreviation: Figure 1,

Description

Title of the invention: A portable method for analyzing the derivative of an electrical quantity of at least one electrochemical element, and associated methods and devices

[0001] The present invention relates to a method for analyzing the derivative of a first electrical quantity relating to at least one electrochemical element (in particular voltage or an accumulated charge) with respect to a second electrical quantity (in particular voltage or an accumulated charge) during a charge or discharge at constant current.

[0002] The present invention also relates to a method for estimating a parameter relating to the state of health of at least one electrochemical element of a battery using the analysis method.

[0003] The present invention also relates to the devices involved in the implementation of such processes, namely: a computer, a management system and an associated battery.

[0004] Typically, a battery comprises one or more current storage cells, also called electrochemical generators, cells, or elements. A battery is an electricity-producing device in which chemical energy is converted into electrical energy. The chemical energy comes from electrochemically active compounds deposited on at least one face of electrodes arranged in the battery. The electrical energy is produced by electrochemical reactions during a discharge of the battery. The electrodes, arranged in a container, are electrically connected to current output terminals that ensure electrical continuity between the electrodes and an electrical load to which the battery is connected.

[0005] To increase the electrical power delivered, several sealed accumulators can be connected together to form a battery. Thus, a battery can be divided into modules, each module being composed of one or more accumulators connected together in series and / or in parallel. For example, a battery may comprise one or more parallel branches of accumulators connected in series and / or one or more parallel branches of modules connected in series.

[0006] A charging circuit is generally provided to which the battery can be connected to recharge the accumulators.

[0007] Furthermore, an electronic management system comprising measurement sensors and an electronic control circuit, more or less sophisticated depending on the application, can be associated with the battery. Such a system makes it possible, in particular, to organize and control the charging and discharging of the battery, to balance the charging and discharging of the different cells of the battery with respect to each other.

[0008] The battery health status is useful information for the electronic battery management system to optimize its use and lifespan. The battery health status is often referred to by the abbreviation SOH, which stands for "State of Health".

[0009] The state of health (SOH) allows us to estimate the aging of the battery between a new state and an end-of-life state, or more generally, between an initial state and a final state.

[0010] In this sense, the state of health (SOH) defines the battery's ability to supply current.

[0011] A technique for determining the state of health (SOH) is one in which the battery's temperature, voltage, and possibly current values ​​are monitored to determine an SOH value based on aging laws. Such aging laws are obtained from laboratory tests. Applying the aging laws to the monitored values ​​thus provides an estimate of the battery's aging.

[0012] However, this technique does not take into account the various mechanisms that make the aging of the battery accumulators non-linear and / or inhomogeneous.

[0013] Moreover, such a technique does not allow for managing the disparity in aging between the elements.

[0014] It is also known to analyze the derivative of the voltage of each element with respect to ampere-hours during a charge or discharge at a relatively low constant current. The lower this current, the more information the derivative contains regarding the aging of the element in question, at the expense of the computation time and maintenance time involved. Values ​​between C / 3 and C / 10 are often used in embedded systems to limit the time taken by a maintenance cycle.

[0015] From this derivative, a detection of peaks is carried out whose type corresponds to phase transitions, in particular crystallographic phases, which take place within the element during charging or discharging.

[0016] For example, in LFP / graphite chemistry, the LiC6 to / from LiC12 transition is a phase transition involving a notable peak. The transition between LiC12 and LIC6 (LIC6 to LIC12 direction) occurs under load, while the transition between LiC6 and LIC12 (LIC6 to LIC12 direction) occurs under discharge.

[0017] Such an analysis is often referred to as DVA analysis, the abbreviation DVA referring to the corresponding English term "Differential Voltage Analysis".

[0018] For a given type of aging, it is possible to associate the position of certain peaks (in Ah elapsed since the beginning of the maintenance cycle) with the capacity of the element using a previously established table.

[0019] In some cases, because the peak disappears with aging (this is notably the case of the LiC6 to / from LiC12 transition) or, conversely, because it appears, the DVA technique can simultaneously use several peaks.

[0020] However, such a process cannot be implemented in an embedded manner.

[0021] Indeed, the voltage measurement of the element is noisy and jagged due to the voltage sensor and the acquisition performed by the analog-to-digital converter. This necessitates the use of complex Savitzky-Golay type filtering processes, which requires the ability to perform heavy calculations and store large amounts of information, as these processes require the entire derived curve for filtering.

[0022] There is a need for a method of analyzing the derivative of a first electrical quantity relating to at least one electrochemical element (in particular a voltage) with respect to a second electrical quantity relating to at least one electrochemical element (in particular a quantity of charge) during a charge or discharge at constant current which is shippable and robust, in particular for the purpose of estimating a parameter relating to the state of health of an electrochemical element.

[0023] To this end, the description relates to a method for analyzing the derivative of a first electrical quantity relating to at least one electrochemical element with respect to a second electrical quantity relating to the same at least one electrochemical element during charging or discharging at constant current or constant power, the analysis method being implemented by a computer of a battery management system for at least one electrochemical element, the analysis method comprising the steps of:

[0024] - obtaining a time signal of the first electrical quantity relating to the minus one electrochemical element,

[0025] - sampling of the time signal obtained according to a first time interval sampling to obtain a first sampled signal,

[0026] - filtering of the first sampled signal, to obtain a filtered signal,

[0027] - sampling of the filtered signal according to a second time interval sampling, to obtain a second sampled signal, the second sampling time interval being strictly greater than the first sampling time interval,

[0028] - derivation of the second sampled signal with respect to the second quantity electrical, to obtain a derived signal with a slow component and a fast component, and

[0029] - detection of a value of a parameter relating to at least one predefined pattern in the derived signal, the detection including obtaining the slow component of the signal derived by applying a filter to the derived signal, the possible presence of a predefined pattern being determined using the slow component obtained.

[0030] According to other advantageous aspects, the analysis method comprises one or more of the following features, taken individually or in all technically possible combinations:

[0031] - the differentiation step includes a substep for calculating the derivative, in order to obtain an intermediate signal and a sub-step of filtering the intermediate signal by applying a filter, to obtain the derived signal.

[0032] - the detection step involves obtaining areas of interest where the predefined pattern is potentially present by determining the times during which a condition relating to the difference between the derived signal and the minimum between the slow component of the derived signal and the derived signal is satisfied.

[0033] - the detection step includes determining whether or not there is actually a predefined pattern in areas of interest obtained by checking if the potentially present predefined pattern respects one or more expected characteristics, a characteristic preferably being related to the width of the area of ​​interest.

[0034] - the parameter relating to at least one predefined pattern whose value is detected during The detection step is the position of the predefined pattern.

[0035] - during the filtering step, a second-order filter is used.

[0036] - the second order filter is the composition of two identical first order sub-filters.

[0037] - each sub-filter has the following filtering function:

[0038] czx has SF (n) - 1+Wsf ,

[0039] Where:

[0040] • Ssp(n) denotes the output of the sub-filter at time n,

[0041] • asF is a parameter,

[0042] • SSF(n-1) denotes the output of the sub-filter at time n-1, and

[0043] • eSF(n) denotes the input of the subfilter at time n.

[0044] - the filter applied to the filtering substep and the filter applied to the filtering step are each composed of the same sub-filters, only the sub-filter parameters vary.

[0045] - one of the two electrical quantities is the quantity of charge and the other quantity Electrical voltage is the voltage.

[0046] The description also relates to a method for estimating a parameter concerning the health status of at least one electrochemical cell of a battery, the estimation method being implemented by a computer of a management system for at least one electrochemical cell of a battery, the estimation method comprising the steps of:

[0047] - implementation of a method for analyzing the derivative of a first quantity electrical quantity relating to at least one electrochemical element with respect to a second electrical quantity relating to at least one electrochemical element during a constant current charge or discharge, to obtain at least one value of a parameter relating to at least one predefined pattern in the derivative of the first electrical quantity, the analysis method being a method such as previously described, and

[0048] - estimation of a parameter relating to the health status of at least one element electrochemical of a battery from at least one parameter relating to at least one predefined pattern in the derivative of the first electrical quantity.

[0049] According to other advantageous aspects, the processes described above include one or more of the following features, taken individually or in all technically possible combinations:

[0050] - at least one electrochemical element has at least one phase transition.

[0051] - at least one electrochemical element comprises a cathodic active material including a lithium iron phosphate, a lithium manganese and iron phosphate or a lithium vanadium fluorophosphate.

[0052] - at least one electrochemical element comprises an anodic active material including graphite.

[0053] The description also relates to a computer designed to analyze the derivative of a first electrical quantity relating to at least one electrochemical element with respect to a second electrical quantity relating to the at least one electrochemical element during a charge or discharge at constant current or constant power, the computer being part of a management system for the at least one electrochemical element of a battery, the computer being designed to:

[0054] - obtain a time signal of the first electrical quantity relating to the at least an electrochemical element,

[0055] - sample the obtained time signal according to a first time interval sampling to obtain a first sampled signal,

[0056] - filter the first sampled signal, to obtain a filtered signal,

[0057] - sample the filtered signal according to a second time interval sampling, to obtain a second sampled signal, the second sampling time interval being strictly greater than the first sampling time interval,

[0058] - derive the second sampled signal with respect to the second quantity electrical, to obtain a derived signal with a slow component and a fast component, and

[0059] - detect a value of a parameter relating to at least one predefined pattern in the derived signal by obtaining the slow component of the derived signal by applying a filter on the derived signal, the possible presence of a predefined pattern being determined using the slow component obtained.

[0060] The description also relates to a computer designed to estimate a parameter relating to the health status of at least one electrochemical cell of a battery, the computer being part of a management system for at least one electrochemical cell of a battery, the computer being designed to:

[0061] - implement a method for analyzing the derivative of a first quantity electrical quantity relating to at least one electrochemical element with respect to a second electrical quantity relating to at least one electrochemical element during a constant current charge or discharge, to obtain at least one value of a parameter relating to at least one predefined pattern in the derivative of the first electrical quantity, the analysis procedure being as previously described, and

[0062] - estimating a parameter relating to the state of health of at least one electrochemical element of a battery from at least one parameter relating to at least one predefined pattern in the derivative of the first electrical quantity.

[0063] The description also relates to a management system for at least one electrochemical cell of a battery, the at least one electrochemical cell having terminals, the management system comprising:

[0064] - a suitable current sensor for applying and measuring the current supplied by said sensor minus one electrochemical element,

[0065] - a voltage sensor suitable for measuring the voltage across the terminals of at least one electrochemical element, and

[0066] - a calculator as previously described.

[0067] The description also relates to a battery comprising:

[0068] - at least one electrochemical element, and

[0069] - a management system as previously described.

[0070] In this description, the expression "specific to" means interchangeably "suited for", "adapted to" or "configured for".

[0071] The invention will become clearer upon reading the following description, given solely by way of non-limiting example, and made with reference to the drawings in which:

[0072] - [Fig. 1] [Fig. 1] is a schematic representation of an example of a battery comprising an electrochemical element,

[0073] - [Fig.2] [Fig.2] is a flowchart of an example of the implementation of a method for analyzing the derivative of the voltage of at least one electrochemical element, and

[0074] - [Fig.3] [Fig.4] [Fig.5] [Fig.6] [Fig.7] [Fig.8] [Fig.9] Figures 3 to 9 illustrate the signals obtained at different stages of the implementation of the process according to [Fig.2].

[0075] A battery 10 is shown in [Fig. 1].

[0076] In a manner known per se, a battery is generally an arrangement of a plurality of electrochemical elements but in the interest of simplifying the subject, a case with a single electrochemical element is described in what follows, knowing that the transposition to other arrangements is immediate.

[0077] The battery 10 comprises an electrochemical element 12 and a management system 14 for the electrochemical element 12.

[0078] As explained previously, an electrochemical element 12 is an electricity-producing device in which chemical energy is converted into electrical energy.

[0079] The electrochemical element 12 therefore delivers a current and a voltage between two terminals.

[0080] Preferably, the electrochemical element 12 is an electrochemical element with a cathodic active material comprising a lithium iron phosphate (LFP) or a lithium manganese iron phosphate (LMFP) or a lithium vanadium fluorophosphate (LVPF).

[0081] According to another embodiment, an electrochemical element 12 comprises an anodic active material comprising graphite.

[0082] Of course, these examples are not limiting and the process described later can be used for any type of electrochemical element 12.

[0083] The management system 14 includes a current sensor 16, a voltage sensor 18, a temperature sensor 20 and a computer 22.

[0084] The current sensor 16 is suitable for measuring the current supplied by the electrochemical element 12 or applied to the electrochemical element 12 depending on whether it is in discharge or charge.

[0085] The voltage sensor 18 is suitable for measuring the voltage across the terminals of the electrochemical element 12.

[0086] The temperature sensor 20 is used to measure the temperature of the electrochemical element 12.

[0087] The calculator 22 is suitable for implementing a method for estimating a parameter relating to the state of health of the electrochemical element 12.

[0088] In the example that will be described, the parameter relating to the state of health is the capacity, so that the calculator 12 here implements a method for determining the capacity of the electrochemical element 12.

[0089] The calculator 22 is an electronic circuit designed to manipulate and / or transform data represented by electronic or physical quantities in registers of the calculator and / or memories into other similar data corresponding to physical data in the register memories or other types of display devices, transmission devices or storage devices.

[0090] As specific examples, the computer 22 includes a single-core or multi-core processor (such as a central processing unit (CPU), a graphics processing unit (GPU), a microcontroller and a digital signal processor (DSP)), a programmable logic circuit, such as an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a programmable logic device (PLD) and programmable logic arrays (PLAs), a state machine, a logic gate and discrete hardware components.

[0091] An example of implementation of the method for estimating the capacitance of the electrochemical element 12 is now described with reference to the flowchart of [Fig.2].

[0092] The estimation method comprises a peak detection phase (PI phase) and a capacity estimation phase (P2 phase).

[0093] According to the example described, the PI detection phase comprises an acquisition step E30, a filtering step E32, a derivation step E34 and a detection step E36.

[0094] During the acquisition step E30, the voltage and current are acquired respectively by the voltage sensor 18 and the current sensor 16.

[0095] An example of a signal obtained at the end of the acquisition step E30 is the noisy signal shown in [Fig.3].

[0096] The filtering step E32 comprises a first sampling substep, a filtering substep and a second sampling substep.

[0097] During the first sampling substep, the calculator 22 samples the voltage signal of the electrochemical element 12 (received from the voltage sensor 18) according to a first sampling time interval.

[0098] In other words, the calculator 22 reads the value of the voltage signal of the electrochemical element 12 at each first sampling time interval.

[0099] According to the example described, the first sampling time interval is between 500 milliseconds (ms) and 2 seconds (s), typically equal to one second.

[0100] The calculator 22 thus obtains a first sampled signal.

[0101] During the filtering substep, the calculator 22 applies a first filter Fl to the first sampled signal.

[0102] According to the example described, the first filter Fl is a second-order filter.

[0103] To improve the embeddability with a simple calculation to perform, the first filter Fl is a cascade of two first-order sub-filters, denoted respectively SF1 and SF2.

[0104] Preferably, each SF subfilter is represented by the following equation:

[0105] v / x agpS^n-D+es^i) àsF\n' ~ l+aSF

[0106] Where: • SSF(n) denotes the output of the SF sub-filter at time n, • üsf is a parameter, • SSF(n-1) denotes the output of the SF sub-filter at time n-1, and • Csptn) denotes the input of the SF sub-filter at time n.

[0107] This allows the memorization to be limited to only two values, namely the output of the first sub-filter SF1 at time n-1 and the output of the second sub-filter SF2 at time n-1.

[0108] Advantageously, the parameter üsf of each sub-filter SF1 or SF2 is here a value between 5 and 15, for example equal to 10.

[0109] Such a filtering substep makes it possible to smooth the voltage signal as shown in [Fig.4].

[0110] During the second sampling substep, the calculator 22 samples the filtered signal according to a second sampling time interval.

[0111] This means that the calculator 22 reads the value of the filtered signal voltage at every second sampling time interval.

[0112] According to the example described, the second sampling time interval is between 30 seconds (s) and 90 s, preferably equal to one minute.

[0113] The second sampling time interval is much greater than the first sampling time interval.

[0114] A second sampled signal is thus obtained.

[0115] The second sampling substep behaves as an additional filter to the filter applied during the filtering substep to further smooth the voltage.

[0116] Such a second substep allows the P2 estimation phase to be implemented at a lower frequency. This reduces the computing power required, thereby improving onboard usability.

[0117] Moreover, with a small number of points, it becomes possible to process a very large number of electrochemical elements 12 in parallel, which makes the process usable on the scale of a complete battery 10.

[0118] Each sampling time interval is adjustable to allow adaptation of the process to the intended application.

[0119] The derivation step E34 comprises a calculation substep and a filtering substep.

[0120] During the calculation substep, the calculator 22 calculates the derivative of the second sampled signal. The calculator 22 thus obtains an intermediate signal.

[0121] For this purpose, in the example described, the calculator 22 calculates the ratio between the voltage variation and the variation in Ampere-hours elapsed.

[0122] An example of an intermediate signal corresponds to the curve marked Cl on [Fig.5].

[0123] During the filtering substep, the computer 22 applies a second filter F2 on the intermediate signal.

[0124] Advantageously, the second filter F2 applied is a second-order filter like the first filter Fl.

[0125] Thus, the second filter F2 is also parameterized by a parameter denoted aF2.

[0126] For example, the parameter aF2 of the second filter F2 is equal to 3.

[0127] Applying the second filter F2 allows the intermediate signal to be smoothed for to obtain a derived signal, as represented in [Fig.5] with reference to the indicated curve C2.

[0128] The peak detection step E36 aims to detect peaks in the derived signal.

[0129] According to the example described, the detection step E36 includes an operation to obtain a slow component of the derived signal, that is to say, the basis of the signal whose variation is slow (compared to the transitions which have a rapid variation).

[0130] The slow component thus corresponds to a background average allowing the detection of a peak because this peak will be broken up more significantly, the peaks being part of the fast component of the derived signal.

[0131] Such a obtaining operation includes the application of a third filter F3 on the derived signal.

[0132] The third filter F3 is here of the same type as the first filter Fl and the second filter F2.

[0133] Thus, the third filter F3 is also parameterized by a parameter denoted fl^3. The parameter aF3 is chosen to be high, for example between 10 and 20.

[0134] The calculator 22 then selects the minimum value between the value of the derived signal and the value obtained by applying the third filter F3.

[0135] The union of the selected minimum values ​​forms the background average so that the background average is never higher than the derived signal.

[0136] As can be seen in [Fig.6], the background average shows a lag and flattening relative to the derived signal.

[0137] As will be shown later, the background average is a signal that allows us to detect a peak because this peak will be broken up more significantly, the peaks being part of the fast component of the derived signal.

[0138] The detection step E36 also includes an operation of comparing the derived signal with the background average.

[0139] When the difference in amplitude between the two signals is within a predefined range, a potential localization range of a peak is detected.

[0140] Such an interval is referred to as the area of ​​interest in what follows.

[0141] The predefined interval extends between a minimum threshold and a maximum threshold.

[0142] The value of the thresholds is predefined and depends on the peak considered.

[0143] As an example, for the peak corresponding to the LiC6 transition from / to LiC12, the minimum threshold corresponds on [Fig.6] to a value of 0.3 (unitless) and the maximum threshold is 2.0 (unitless).

[0144] These values ​​are given for the case of a current of C / 10, sampling frequencies of 1s and 60s and parameters for filters of 10 and 3 respectively.

[0145] In practice, the value of the derivative height associated with the peaks depends strongly on these values ​​as well as on the noise level.

[0146] Each area of ​​interest will correspond to time intervals during which the difference in amplitude respects a condition related to the peak sought.

[0147] In the example of [Fig.7], four areas of interest were determined.

[0148] It may be pointed out here that the areas of interest comply with the condition associated with a nearby peak.

[0149] Typically, the difference for the peak of the second area of ​​interest on [Fig.7] (counting from left to right) corresponds to higher minimum and maximum thresholds than the peak of the fourth area of ​​interest.

[0150] Also, the difference in amplitude in the second area of ​​interest meets a first criterion with relatively high minimum and maximum thresholds, while the difference in amplitude in the fourth area of ​​interest meets a second criterion different from the first criterion and involving a relatively low minimum and maximum threshold.

[0151] In the example described, as seen in [Fig.9], there are two potential peaks and therefore two different possible areas of peak presence corresponding to two different local criteria.

[0152] This means that over time, the value of the derived signal is compared with the background average and, as soon as the local criterion is met, an area of ​​interest is opened until the local criterion is no longer met and the area of ​​interest is closed.

[0153] An area of ​​interest is thus delimited by a first delimitation corresponding to the first instant at which the criterion is met and a second delimitation corresponding to the last instant at which the criterion is met.

[0154] To clarify this point, in the possible area of ​​presence of a peak 1, a first local criterion is used and it turns out that the first area of ​​interest and the second area of ​​interest are the only areas where this first local criterion is respected.

[0155] Then, in the possible area of ​​presence of a peak 1, a second local criterion (different from the first criterion) is used, which leads to the identification of the third area of interest and the fourth area of ​​interest as the only areas where this second local criterion is met.

[0156] Following the comparison operation, four areas of interest likely to contain a peak are thus obtained.

[0157] These areas of interest should be considered in light of the fact that in reality, only two peaks are expected.

[0158] It is therefore necessary to determine the areas of interest among the four areas of interest actually exhibiting one of the two expected peaks.

[0159] For this purpose, the detection step E36 then includes an operation to determine the actual presence or not of a peak in the areas of interest obtained.

[0160] For this purpose, it is checked whether the potential peak conforms to one or more expected characteristics.

[0161] For example, the width of the potential peak is compared to an expected width or the position of the potential peak is compared to an expected position range.

[0162] These characteristics are established by tests leading to a table associating each transition with a characteristic, either in the form of an interval representing the variation of the characteristic during aging or in the form of a value, particularly when the variation of the characteristic is small during aging.

[0163] When the potential peak does not respect an expected characteristic, the area of ​​interest is discarded.

[0164] A first example of a characteristic is the width of the peak.

[0165] A preferred implementation is to compare the width of the area of ​​interest with the expected width.

[0166] When the area of ​​interest is not wide enough, this implies that the area of ​​interest does not have a peak corresponding to a transition.

[0167] With this criterion, in the example described, the first zone and the third zone are eliminated because they are not extensive enough.

[0168] A second example of a characteristic is the nature of the variation of the derived signal in the area of ​​interest.

[0169] For this purpose, as illustrated by the left part of [Fig.8], a peak search can be carried out by matching patterns on three consecutive values ​​of the derived signal.

[0170] The search for the peak shape thus involves comparing 3 values ​​of the derived signal, taken contiguously in time at time n-2, time n-1, and time n. The following iteration sees these values ​​evolve as follows: the value at time n-2 is discarded, the value at time n-1 becomes the value at time n-2, the The value at time n becomes the value at time n-1 and the value at time n is the current value of the derived signal.

[0171] A peak is detected by pattern search when the following inequality is satisfied:

[0172] Derived signal (n-2) < Derived signal (n-1) > Derived signal (n)

[0173] In the example described, the first area of ​​interest does not satisfy this criterion since the derived signal decreases monotonically in this area of ​​interest.

[0174] The detection step E36 thus makes it possible to obtain the position of the peak.

[0175] The position of the peak is an example of a value of a parameter relating to the peak that can be detected in the derived signal.

[0176] The estimation phase P2 aims to estimate the capacity from the position of the peaks detected during the detection phase PL

[0177] The estimation phase P2 includes a correction step E38 and a deduction step E40.

[0178] During the correction step E38, the calculator 22 seeks to obtain the actual position of the detected peaks.

[0179] A first correction is a compensation for the delays introduced by the filters used.

[0180] To do this, the introduced delay is calculated and then converted into Ah using the fact that the current is constant. This converted delay is subtracted from the detected position of the peak.

[0181] A second example of correction is related to the fact that the initial position is not always known since it is generally expressed in terms of initial charge state socinit.

[0182] To achieve this, an AÀhinit correction will be applied according to the following formula:

[0183] «.Ah^p 100 H—iæ - >

[0184] Where: • a and / 5 are parameters of a linear model linking the absolute position Ahabs of the peaks (after compensation) and the capacity Qpic, that is to say:

[0185] Q^aAh^+fi

[0186] These parameters a and P are assumed to be known, for example from a table. • Ahmes designates the measured value of the position of the peaks (that to be compensated).

[0187] In this model, it is also assumed that the relationship between the capacity Qpic and the absolute position Ah^ is linear, which is generally true.

[0188] In some embodiments, a more elaborate model than the previous linear model may be used.

[0189] During the deduction step E40, the calculator 22 uses a table associating the actual position of the peak(s) with the value of the capacitance.

[0190] The PI detection phase just described can be implemented in highly noisy environments through the use of multiple filtering methods. This makes this detection compatible with sensors incorporating an analog-to-digital converter with poor performance. This detection also functions in the presence of bias in one or more sensors or electromagnetic interference problems.

[0191] The PI detection phase thus makes it possible to carry out a precise survey of the positions of the phase transitions of any electrochemical element 12 of the battery 10 in a noisy environment.

[0192] This measurement is compatible with embedded implementation due to the low computational load. In particular, the combination of several filters with two sampling techniques at different frequencies makes it possible to significantly limit the resources involved.

[0193] Furthermore, the PI detection phase can be carried out in real time.

[0194] This also makes it possible to consider implementing the PI detection phase for a large number of electrochemical elements 12.

[0195] Experiments carried out by the applicant have shown that such a possibility of on-the-fly characterization of peaks makes it possible to obtain a good estimate of the capacity of each electrochemical element 12.

[0196] Indeed, since a full discharge or charge is not indispensable, it is possible to determine the capacitance of each electrochemical element 12 without waiting for the electrochemical element 12 with the lowest capacitance to reach its limits.

[0197] Of course, the estimation phase P2 can be implemented to estimate other parameters representative of the aging of the electrochemical element 12.

[0198] For example, it is possible to determine the resistance for each electrochemical element 12.

[0199] In some embodiments, the estimation phase P2 can also provide an uncertainty of the estimate.

[0200] The uncertainty here is related to the measurement of the state of charge and to the fact that the derived signal is calculated only on the basis of samples of the measured signal.

[0201] Regarding the uncertainty of these two sources, it is possible to deduce the resulting uncertainty on the deduced capacity.

[0202] A result obtained for the two peaks of the preceding figures obtained by simulation gives the following results for an electrochemical element with a capacitance of 26.25 Ah:

[0203] [Tables 1] Picl Pic 2 Peak position -3.912 Ah -20.21 Ah Inferred capacity 26.45 Ah 26.19 Ah Uncertainty ± 0.26 Ah ± 0.24 Ah

[0204] The fused capacity of the 2 estimates from each of the peaks is 26.32 Ah and the associated uncertainty is ± 0.25 Ah.

[0205] This is calculated by weighting the result to favour the intermediate capacity with the lowest uncertainty (precision), without however ignoring the most uncertain intermediate capacity (robustness).

[0206] The error from the true value is 0.07 Ah, which is within the uncertainty.

[0207] As regards robustness, it is obtained despite the fact that the voltage measurement of the electrochemical element is noisy and especially crenellated due to the acquisition carried out by the analog / digital converter.

[0208] Such a PI detection phase is also usable for other exploitations of detected peaks.

[0209] For example, the PI detection phase can be used to determine the relationship linking the variation of the open-circuit voltage to the state of charge in order to update a state of charge estimation algorithm that may have deviated.

[0210] The PI detection phase can also serve as an input for an inference of the degradation modes of the electrochemical element 12.

[0211] Furthermore, the PI detection phase is not necessarily limited to peak detection.

[0212] In particular, the PI detection phase can detect valleys.

[0213] The right-hand side of [Fig.8] illustrates the scenario for detecting a valley where only The inequality to be verified changes from what was described previously, and is written as follows:

[0214] Derived signal (n-2) > Derived signal (n-1) < Derived signal (n)

[0215] More generally, the PI detection phase can be used to detect any type of patterns in the derivative of the tension.

[0216] For example, the PI detection phase can be used to detect the beginning of a plateau in the voltage signal.

[0217] This can be particularly useful for calculating resistance.

[0218] It should also be noted here that the analysis method described was done in the particular case of a DVA analysis corresponding to the analysis of the derivative of the voltage with respect to Ampere-hours.

[0219] However, it is conceivable to use a similar implementation to analyze the variation of a derivative of different physical quantities.

[0220] According to a particular example, the analysis method can be used to analyze the derivative of the quantity of charge accumulated (Ampere-hours) with respect to the voltage.

[0221] It is indeed sufficient to carry out the same operations but on a different time signal, namely that of the first physical quantity.

[0222] For the example of the DVA, this first physical quantity is the voltage.

[0223] It could also be envisaged to implement the process in the context of discharge or charge with a constant power.

[0224] A power equal to or 1 / 10 of the rated power are examples of power values ​​which may be used in such a constant power implementation.

Claims

Demands

1. A method for analyzing the derivative of a first electrical quantity relating to at least one electrochemical element (12) of a battery (10) with respect to a second electrical quantity relating to the at least one electrochemical element (12) during a charge or discharge at constant current or constant power, the analysis method being implemented by a computer (22) of a management system (14) of the at least one electrochemical element (12), the analysis method comprising the steps of: - obtaining a time signal of the first electrical quantity relating to the at least one electrochemical element (12), - sampling the time signal obtained according to a first sampling time interval to obtain a first sampled signal, - filtering the first sampled signal, to obtain a filtered signal, - sampling the filtered signal according to a second sampling time interval,to obtain a second sampled signal, the second sampling time interval being strictly greater than the first sampling time interval, - differentiation of the second sampled signal with respect to the second electrical quantity, to obtain a derived signal having a slow component and a fast component, and - detection of a value of a parameter relating to at least one predefined pattern in the derived signal, the detection including obtaining the slow component of the derived signal by applying a filter to the derived signal, the possible presence of a predefined pattern being determined using the slow component obtained.

2. Analysis method according to claim 1, wherein the differentiation step includes a substep of calculating the derivative, to obtain an intermediate signal and a substep of filtering the intermediate signal by applying a filter (F2), to obtain the derived signal.

3. An analysis method according to claim 1 or 2, wherein the detection step comprises obtaining areas of interest where the predefined pattern is potentially present by determining the times during which a condition relating to the difference between the derived signal and the minimum between the slow component of the derived signal and the derived signal is verified.

4. An analysis method according to claim 3, wherein the detection step comprises determining the actual presence or absence of a predefined pattern in the areas of interest obtained by checking whether the potentially present predefined pattern meets one or more expected characteristics, a characteristic preferably being related to the width of the area of ​​interest.

5. An analysis method according to any one of claims 1 to 4, wherein the parameter relating to at least one predefined pattern whose value is detected during the detection step is the position of the predefined pattern.

6. Analytical method according to any one of claims 1 to 5, wherein, during the filtering step, a second-order filter (Fl) is used.

7. Analysis method according to claim 6, wherein the second-order filter (Fl) is the composition of two identical first-order sub-filters (SF1, SF2).

8. Analysis method according to claim 7, wherein each sub-filter (SF1, SF2) has the following filtering function: ^SF Vn) ~ l+aSF Where: • SSF(n) denotes the output of the sub-filter (SF1, SF2) at time n, • asF is a parameter, • SSF(n-1) denotes the output of the sub-filter (SF1, SF2) at time n-1, and • esF^n) denotes the input of the sub-filter (SF1, SF2) at time n.

9. Analytical method according to claims 2 and 8, wherein the filter applied to the filtering substep and the filter applied to the filtering step are each composed of the same subfilters, only the subfilter parameters of which vary.

10. Analytical method according to any one of claims 1 to 9, wherein one of the two electrical quantities is the quantity of charge and the other electrical quantity is the voltage.

11. A method for estimating a parameter relating to the health status of at least one electrochemical element (12) of a battery (10), the estimation method being implemented by a computer (22) of a management system (14) of at least one electrochemical element (12), the estimation method comprising the steps of: - implementing a method for analyzing the derivative of a first electrical quantity relating to at least one electrochemical element (12) with respect to a second electrical quantity relating to at least one electrochemical element (12) during a charge or discharge at constant current, to obtain at least one value of a parameter relating to at least one predefined pattern in the derivative of the first electrical quantity, the analysis method being a method according to any one of claims 1 to 10, and - estimating a parameter relating to the state of health of at least one electrochemical element (12) of a battery (10) from at least one parameter relating to at least one predefined pattern in the derivative of the first electrical quantity.

12. A method according to any one of claims 1 to 11, wherein at least one electrochemical element has at least one phase transition.

13. A method according to any one of claims 1 to 12, wherein at least one electrochemical element (12) comprises a cathodic active material comprising a lithium iron phosphate, a lithium manganese iron phosphate, or a lithium vanadium fluorophosphate.

14. A method according to any one of claims 1 to 12, wherein at least one electrochemical element (12) comprises an anodic active material comprising graphite.

15. A calculator (22) adapted to analyze the derivative of a first electrical quantity relating to at least one electrochemical element (12) of a battery (10) with respect to a second electrical quantity relating to the at least one electrochemical element (12) during a charge or discharge at constant current or constant power, the calculator (22) being part of a management system (14) of the at least one electrochemical element (12), the calculator (22) being adapted to: - obtain a time signal of the first electrical quantity relating to the at least one electrochemical element (12),

16.

17. - sample the obtained time signal according to a first sampling time interval to obtain a first sampled signal, - filter the first sampled signal to obtain a filtered signal, - sample the filtered signal according to a second sampling time interval to obtain a second sampled signal, the second sampling time interval being strictly greater than the first sampling time interval. - to derive the second sampled signal with respect to the second electrical quantity, to obtain a derived signal exhibiting a slow component and a fast component, and - detect a value of a parameter relating to at least one predefined pattern in the derived signal by obtaining the slow component of the derived signal by applying a filter to the derived signal, the possible presence of a predefined pattern being determined using the slow component obtained. Calculator (22) for estimating a parameter relating to the health status of at least one electrochemical cell (12) of a battery (10), the calculator (22) being part of a management system (14) for at least one electrochemical cell (12), the calculator (22) being for: - to implement a method for analyzing the derivative of a first electrical quantity relating to at least one electrochemical element (12) with respect to a second electrical quantity relating to at least one electrochemical element (12) during a constant current charge or discharge, in order to obtain at least one value of a parameter relating to at least one predefined pattern in the derivative of the first electrical quantity, the analysis method being a method according to any one of claims 1 to 10, and - estimate a parameter relating to the state of health of at least one electrochemical element (12) of a battery (10) from at least one parameter relating to at least one predefined pattern in the derivative of the first electrical quantity. Management system (14) of at least one electrochemical element (12) of a battery (10), the at least one electrochemical element (12) having terminals, the management system (14) comprising: - a current sensor (16) suitable for applying and measuring the current supplied by said at least one electrochemical element (12),

18. - a voltage sensor (18) suitable for measuring the voltage across at least one electrochemical element (12), and - a calculator (22) according to claim 15 or 16. Battery (10) including: - at least one electrochemical element (12), and - a management system (14) according to claim 17.