Method for monitoring an electrochemical energy storage system, computer program and associated devices
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- CENT NAT DE LA RECH SCI (C N R S)
- Filing Date
- 2026-01-23
- Publication Date
- 2026-07-30
Smart Images
Figure EP2026051685_30072026_PF_FP_ABST
Abstract
Description
DESCRIPTION Title: Method for monitoring an electrochemical energy storage system, computer program and associated devices Domain
[0001] The present invention relates to a method for monitoring an electrochemical energy storage system.
[0002] The invention also relates to a computer program and devices implementing such a method.
[0003] The invention applies to the field of methods for monitoring an electrochemical energy storage system.
[0004] It is known to use methods for monitoring electrochemical energy storage systems.
[0005] An electrochemical energy storage system, such as a battery or supercapacitor, relies on reversible chemical reactions to convert chemical energy into electrical energy and vice versa. It consists primarily of two electrodes (the positive and negative electrodes) and an electrolyte that transports ions between them. During discharge, the negative electrode releases electrons that flow to the positive electrode via an external circuit, generating an electric current. Simultaneously, positive ions migrate within the electrolyte to maintain electrical balance. During recharging (for rechargeable systems), the process is reversed using an external power source, restoring the initial chemical energy. Furthermore, the classically used characterization methods in electrochemistry are galvanometry, voltammetry, EIS (Electrochemical Impedance Spectroscopy), PITT (Potentiostatic Intermittent Titration Technique), and GITT (Galvanostatic Intermittent Titration Technique). These well-known methods are unsteady, as the electrochemical elements are in a state of permanent disequilibrium. Regarding coulomb measurement, Q = fi(t)dt + Q initial Coulomb), which is based solely on current measurement, there is a problem of knowing the initial state.
[0006] However, such methods do not provide complete satisfaction.
[0007] Indeed, these methods consist of studying electrochemical energy storage systems in states of permanent imbalance, by imposing disturbances (of current, potential or frequency) and analyzing the dynamic responses.
[0008] Electrochemical energy storage systems are nonlinear, exhibiting hysteresis, abrupt parameter changes during chemical phase transitions, and large time constants. While the transfer function (output / input) in a linear system is independent of the excitation type, in a nonlinear system like electrochemical energy storage systems, the transfer function (output / input) depends on the type of excitation applied. Consequently, the excitation type introduces measurement bias when characterizing the components of an electrochemical energy storage system.
[0009] Furthermore, in complex systems such as batteries or supercapacitors, the measured responses may combine the effects of different phenomena, making data analysis more difficult.
[0010] One object of the present invention is to remedy at least one of the drawbacks of the prior art.
[0011] Another aim of the invention is to provide a more precise solution for monitoring an electrochemical storage system.
[0012] Another objective of the invention is to provide a monitoring solution for an electrochemical storage system, allowing for a better representation of the actual behavior of said system. Description of the invention
[0013] To this end, the invention relates to a method for monitoring an electrochemical energy storage system, forming a storage system, the method being implemented by computer and comprising the following steps: -from a voltage measurement V(t) across the terminals of the storage system, selection of a parametric model associated with the measured voltage 7(t) from a set of parametric models, each associated with a corresponding voltage; -adjustment of the selected parametric model, based on a comparison between the measured voltage value 7(t) and an output of said parametric model applied to a current value equal to the current flowing through the storage system; -calculation of a state of charge and / or a state of health of the storage system from at least one parameter of the adjusted parametric model.
[0014] One aim of the invention is to provide a more precise solution for monitoring an electrochemical storage system.
[0015] This accuracy is obtained by adjusting the parameter(s) of the selected parametric model from a comparison between a measured voltage value V(t) and an output 7'(t) calculated by the selected parametric model.
[0016] This iterative adjustment allows for consideration of dynamic variations in the selected parametric model parameter(s). As a result, the state of charge (SOC) and / or state of health (SOH) of the storage system is calculated with greater accuracy, reflecting changes in the actual behavior of the storage system over time.
[0017] More precisely, the voltage measured across the terminals V'(t) of the element is disturbed by the current and internal resistances, by the history of previous transients or variations; by the rebalancing or recoveries taking place by superposition of effects.
[0018] The voltage 7(t) is therefore not a directly and easily usable measurement. The present invention allows direct access to the voltage y'(t) through parametric models that adjust the voltage V(t) across the terminals in real time. This adjustment takes into account several measurement biases.
[0019] Another objective of the invention is to provide a monitoring solution for an electrochemical storage system that allows for a better representation of the actual behavior of said system.
[0020] By leveraging voltage and current variations related to the use of the storage system, the invention dynamically adjusts the various parametric models within the overall set of parametric models, thus enabling a precise characterization of the storage system adapted to real-world conditions. This results in improved estimation of states of charge (SOC) and states of health (SOH). The invention allows for a more reliable analysis of the storage system and, consequently, its monitoring.
[0021] Therefore, it is the updating of the parameter(s) of the parametric model that allows us to track the aging of the storage system.
[0022] The parameter(s) of the parametric model, as a function of voltage, allow the state of charge and the health of the storage system to be monitored as the voltage varies. The dynamics of the voltage and current during the operation of the storage system allow the parameter(s) of the parametric model to be updated dynamically.
[0023] Advantageously, the invention may further comprise the following step: - For each among a plurality of distinct predetermined voltage steps:■from a measurement of current i(t) flowing through the storage system in response to the application of said voltage step across the terminals of the storage system at a reference date, fitting of an initial parametric model from a comparison between the measured current value i(t) and an output of said parametric model applied to a voltage value equal to the current flowing through the storage system.
[0024] According to the invention, the nominal voltage of said step corresponds to the average voltage (v initial + V final ) / 2.
[0025] The invention analyzes the current response following a voltage step or jump, for example, between 80mV and 5mV, according to voltage ranges where the variability of the initial parametric model parameter(s) is significant. The invention avoids measurement biases by: - freeing itself from the large time constants by starting from one equilibrium to arrive at another equilibrium; - overcoming strong non-linearities by adapting the pitch of the tension jumps; - overcoming hysteresis by performing transients that always increase or decrease in voltage.
[0026] The invention has very high accuracy and repeatability, which makes it possible to obtain a dynamic initial parametric model at each level of the voltage during charging and discharging.
[0027] Furthermore, the invention proposes a pseudo-stationary method which, by moving from one equilibrium to a new equilibrium, allows for a good initialization of the voltage values of the parameter(s) of the initial parametric models.
[0028] Advantageously still, at least one parametric model associated with a measured voltage 7(t) from among a set of parametric models can receive, as input data, at least one parameter from the initial fitted parametric model.
[0029] This allows the parametric model to be calibrated with more accurate representative parameters of the storage system.
[0030] Preferably, the parameter(s) of the initial parametric model are in the form of a parametric table. The set of parameters of the initial parametric model constitutes the initial parametric table of the storage system.
[0031] For example, to take into account the non-linear evolution of the parameter(s) as a function of temperature, it is possible to have several initial parametric tables.
[0032] The initial parametric table(s) include the evolution of nonlinearities and time constants of the storage system. In the case of hysteresis, two initial parametric tables can be established, one for when the storage system is charging and the other for when the storage system is discharging, with switching between the initial parametric tables depending on the direction of the current.
[0033] The initial parametric model can have an order between two and five.
[0034] Preferably, for a storage system corresponding to a supercapacitor, a model structure of at least second order allows for high accuracy in fitting the parameters of the initial parametric model. For a storage system corresponding to a battery, an order between third and fifth allows for high accuracy in fitting the parameters of the initial parametric model.
[0035] The initial parametric model can be a Foster structure model or a Cauer structure model.
[0036] Foster's structural model allows direct access to time constants and incremental capacity. Cauer's structural model, on the other hand, allows calculations to determine time constants and obtain incremental capacity by summing capacities. This feature is advantageous because these model structures simplify calculations and data generation by the parametric model. Consequently, parametric identification performed by the parametric model is faster and more efficient.
[0037] Advantageously, the invention may further include a step of estimating the storage system with an extended Kalman filter.
[0038] The system estimation according to the invention allows for obtaining the values of the parameter(s) of the storage system. The use of an extended Kalman filter linearizes the equations concerning the estimated state of charge and / or health. Furthermore, in the event of a loss of a voltage measurement, the use of an extended Kalman filter allows for voltage estimation using only the current information.
[0039] According to another aspect of the invention, a computer program is proposed comprising executable instructions which, when executed by computer, implement the steps of the process as defined above.
[0040] The computer program can be in any computer language, such as for example machine language, C, C++, JAVA, Python, etc.
[0041] According to another aspect of the invention, a monitoring device for an electrochemical energy storage system is proposed, forming a storage system, the device being configured to: -from a voltage measurement V(t) across the terminals of the storage system, select a parametric model associated with the measured voltage 7(t) from a set of parametric models, each associated with a corresponding voltage; -adjust the selected parametric model, based on a comparison between the measured voltage value 7(t) and an output of said parametric model applied to a current value equal to the current flowing through the storage system; -calculate a state of charge and / or a state of health of the storage system from at least one parameter of the adjusted parametric model.
[0042] The device according to the invention can be any type of device such as a server, a computer, a tablet, a calculator, a processor, a computer chip, programmed to implement the method according to the invention, for example by executing the computer program according to the invention. Brief description of the figures
[0043] The invention will be better understood upon reading the following description, given solely by way of non-limiting example and made with reference to the accompanying drawings in which:
[0044] Figure 1 is a schematic representation of an electrochemical installation according to one embodiment of the invention.
[0045] Figure 2 is a flowchart of a method for monitoring an electrochemical energy storage system according to an embodiment of the invention.
[0046] Figures 3a) to 3d) are non-limiting examples of initial parametric model structures according to the invention.
[0047] Figure 4 is a schematic representation of an implementation of the process according to the invention.
[0048] It is understood that the embodiments described below are by no means exhaustive. In particular, variants of the invention may be conceived comprising only a selection of the features described below, isolated from the other features described, if this selection of features is sufficient to confer a technical advantage or to differentiate the invention from the prior art. This selection includes at least one preferably functional feature without structural details, or with only a portion of the structural details if that portion alone is sufficient to confer a technical advantage or to differentiate the invention from the prior art.
[0049] In particular, all the variants and embodiments described can be combined with each other if there are no technical obstacles to this combination.
[0050] In the figures and in the rest of the description, elements common to several figures retain the same reference. Detailed description
[0051] An electrochemical installation 2 according to the invention is illustrated by figure 1.
[0052] The electrochemical installation 2 includes an electrochemical energy storage system 4, referred to as storage system 4, and a monitoring module 6 for monitoring storage system 4.
[0053] When the storage system 4 is in a state of charge, it is configured to receive a current i(t).
[0054] When the storage system 4 is in a discharge state, it is configured to deliver a current i(t).
[0055] In particular, storage system 4 is an electrochemical battery or a supercapacitor.
[0056] As is known, the storage system 4 comprises several electrochemical cells. Each cell includes two electrodes, a positive electrode and a negative electrode, separated by an electrolyte that allows the conduction of ions.
[0057] When storage system 4 is in a charging state, it receives a current i(t) at the positive electrode, causing positive ions to move from the positive electrode to the negative electrode and releasing electrons. These electrons then flow through an external circuit, thus providing electrical energy. When storage system 4 is in a discharging state, positive ions move from the negative electrode to the positive electrode, while electrons return to the positive electrode, generating electricity in the form of a current i(t).
[0058] For example, storage system 4 can correspond to electrochemical storage battery elements of various technologies (supercapacitor batteries).
[0059] As illustrated in Figure 1, the storage system 4 is connected to a tracking module 6.
[0060] The monitoring module 6 is configured to receive, as input, a voltage V(t) and a current i(t) measured across the terminals of the storage system 4.
[0061] The monitoring module 6 is also configured to calculate a state of charge and / or a state of health of the storage system 4.
[0062] More specifically, the tracking module 6 is configured to implement a tracking process 8.
[0063] Preferably, the tracking module 6 is configured to implement a calibration step 10.
[0064] As illustrated by Figure 2, the monitoring process 8 includes a calibration step 10, a selection step 12, an adjustment step 14 and a calculation step 16.
[0065] The tracking module 6 is configured, during a calibration step 10, to implement an initial parametric model.
[0066] More specifically, the tracking module 6 is configured, for each among a plurality of predetermined distinct voltage steps, to measure a current i(t) flowing through the storage system 4 in response to the application of said voltage step across the terminals of the storage system 4 at a reference date.
[0067] Advantageously, the amplitude of the predetermined voltage steps depends on the nonlinearity of at least one parameter of the initial parametric model. More precisely, in areas where the variability of this parameter is high, the voltage step amplitudes are reduced to improve the accuracy of nonlinearity identification. This feature is advantageous because it optimizes the representation of local variations while minimizing errors related to rapid fluctuations in the parameter(s).
[0068] Due to the strong nonlinearities of the storage system, the largest time constants of the initial parametric model vary by a factor of 200 or more. For the tracking module to accurately represent the behavior of the storage system down to zero frequency, the step duration must exceed the largest of the storage system's time constants. For example, the largest can reach sixty thousand seconds.
[0069] The initial parametric model is configured to receive, as input, the measurement of the current i(t) at the reference time and to generate, as output, a current value i'(t) associated with a voltage value equal to the current i(t). The initial parametric model is fitted by comparing the measured current value i(t) with the value j'(t) generated as output by the initial parametric model.
[0070] More specifically, parameters representative of the behavior of the storage system 4 are adjusted from the comparison between the measured current value i(t) and the value j'(t) generated at the output of the initial parametric model.
[0071] Advantageously, the adjusted parameters correspond to at least one of the following parameters: - the value(s) of the resistances R.0 to RN; - the value(s) of the capacities C0 to CN of the monitoring module 6.
[0072] Such a characteristic is advantageous insofar as the values of the resistances and capacitances allow us to provide information on the aging of the storage system 4. By adjusting their values, the initial parametric model becomes more and more accurate.
[0073] Preferably, the tracking module 6 is configured to identify all the parameters of the initial parametric model. Advantageously, the tracking module 6 is configured to send the adjusted parameters as input data to a master parametric model, the adjusted parameters allowing the master parametric model to be calibrated. This feature is advantageous because it optimizes the accuracy of the master parametric model, since the adjusted parameters are more accurate in representing the storage system 4 at the reference date.
[0074] Optionally, the adjusted parameters are stored in a table, with the parameter values based on predetermined voltage steps. This parameter table, also called a parametric table, is created for each measurement of the voltage value V(t) across the terminals of the storage system 4. The parametric table created based on the voltage value V(t) at the reference date corresponds to the initial parametric table of the storage system 4. The set of parametric tables created allows for the identification and representation of the evolution of parameter nonlinearities as well as the time constants of the storage system 4.
[0075] Optionally, when hysteresis exists, for example when storage system 4 is a supercapacitor, two initial parametric tables are created, one initial parametric table being representative of storage system 4 in a state of charge, the second initial parametric table being representative of storage system 4 in a state of discharge.
[0076] Preferably, the reference date corresponds to the date of first use of the storage system 4.
[0077] Advantageously, the initial parametric model is a Foster structure model or a Cauer structure model. This characteristic is advantageous because this type of model structure allows for the grouping of several essential parameters for monitoring the storage system, such as: - a curve of the incremental capacity AQ / AV; - a model of the dynamic behavior of the storage system 4 - an impedance analysis curve (EIS) in the frequency domain; - the current and voltage time responses of the electrochemical system; -impedance analysis curves; -values of the asymptote of the impedance measurement curves.
[0078] The parametric model is the signature of the electrochemical system. Each parametric model contains the non-linearities and dynamics of the responses.
[0079] These two model structures, for the same order, are equivalent and can be expressed as a pole- and zero-factored transfer function. These two model structures are capacitive models in the sense that their asymptotic behavior at zero frequency corresponds to that of an electrical capacitor. The value of the capacitance (C o for a Foster or 2" i structure model ci for a Cauer structure model) corresponds to the value of the incremental capacity defined by C(v) = dQ / dv of the storage system 4.
[0080] Advantageously, the charging or discharging curve of the storage system 4 allows, by derivation, the calculation of the incremental capacity curve. Conversely, the initial parametric models as a function of the voltage 7(t) contain the incremental capacity and allow, by integration, the recovery of the charging and discharging curves.
[0081] As an example, different initial parametric model structures are presented in figures 3a) to 3d). Figure 3a) represents a capacitive Foster structure, figure 3b) represents a resistive Foster structure, figure 3c) represents a capacitive Cauer structure and figure 3d) represents a resistive Cauer structure.
[0082] Preferably, the order of the selected parametric model is from two to five depending on the spectral richness of the storage system 4.
[0083] A schematic representation of an implementation of process 8 according to the invention is illustrated by figure 4.
[0084] The tracking module 6 is configured to, during the selection step 12, select a parametric model 18.
[0085] More specifically, the monitoring module 6 is configured to select a parametric model 18 associated with a voltage 7(t) 20 measured across the terminals of the storage system 4 from a set of parametric models. Each parametric model is associated with a corresponding voltage, the voltage values varying from a Vmin to a Vmax and corresponding to the operating range of the storage system 4.
[0086] The selected parametric model 18 is configured to receive, as input, the measured voltage value 7(t) and to generate, a value 7'(t) associated with a current value i(t) measured across the terminals of the storage system 4.
[0087] Each parametric model includes the parameters representative of the operation of the storage system 4.
[0088] The tracking module 6 is configured to, during the adjustment step 14, adjust the selected parametric model 18.
[0089] More specifically, the tracking module 6 is configured to adjust the selected parametric model 18 from a comparison between the measured voltage value 7(t) and the voltage value V' f).
[0090] More specifically, the tracking module 6 is configured to adjust the parameters of the selected parametric model 18 so as to make the voltage value 7'(t) tend towards the voltage value V(t).
[0091] The parameters of the parametric table of the selected parametric model 18 are adjusted in real time by an optimization algorithm 22.
[0092] Advantageously, optimization algorithm 22 corresponds to an online recursive identification algorithm. This characteristic is advantageous because recursive identification allows for the updating of the various parametric tables of the different parametric models, and consequently, the parameters of the different parametric models.
[0093] For example, the adjustment of a parametric table is carried out during a recharge of the storage system 4 at constant current i(t), which allows the incremental capacity to be adjusted.
[0094] For example, the adjustment of a parametric table is carried out when the current i(t) is identified as disturbed by the tracking module 6 during the use of the storage system 4.
[0095] For example, the adjustment of a parametric table is carried out during a recharge of the storage system 4 with a square wave current i(t) enriching the spectrum to increase the spectral richness of the impedances of the tracking module.
[0096] The monitoring module 6 is configured to, during calculation step 16, calculate a state of charge and / or a state of health of the storage system 4 in output S of the parametric model.
[0097] More specifically, the monitoring module 6 is configured to calculate a state of charge and / or a state of health of the storage system 4 from the adjusted parametric table comprising the adjusted parameters of the selected parametric model 18.
[0098] The state of charge and / or the state of health of the storage system 4 is calculated from the initial capacity Q initialeof storage system 4. "Initial capacity" refers to the maximum amount of load that storage system 4 can store at a reference date or when fully loaded. The calculation of the state of charge and / or the state of health of storage system 4 depends on the structure of the selected parametric model 18 and its order.
[0099] Monitoring the aging of the electrochemical system by updating parametric models allows us to maintain accuracy over time for the state of charge and to calculate the state of health of the electrochemical system.
[0100] The initial capacity Q initiale storage system 4 is calculated from the initial parametric table generated during calibration step 10.
[0101] For example, for an initial parametric Foster structure model, the initial capacity is calculated using the following equations: Where c0(v) is the value of the different capacities of the tracking module 6; Av is the voltage difference defined by a parameter between two different voltage values from the initial parametric table.
[0102] Calculating the initial capacity Q initiale is achieved by summing the value of the capacitances C(v) of the tracking module 6 weighted by the voltage difference Av. Indeed, the capacitances C(v) of the storage system 4 are weighted by the voltage difference they cover, which means that each capacitance is multiplied by the voltage range in which it is taken into consideration.
[0103] A value for the capacity of storage system 4 at a given time t is also calculated by the tracking module 6 according to the following equation:
[0104] The capacity at time t is calculated by summing the values of the capacities C(v) of the storage system 4 weighted by the voltage difference Av up to the voltage corresponding to time t. The capacity at time t is calculated from a parametric table extracted from the selected parametric model 18 associated with the voltage value of time t.
[0105] Then, the monitoring module 6 calculates the state of charge and / or the health status of the storage system 4. Regarding the calculation of the state of charge, this is calculated according to the following equation: Q(t) soc = Qinitial
[0106] The state of charge is calculated by taking the ratio between the capacity at time t and the initial capacity Q initiale -
[0107] Health status is calculated by monitoring module 6 according to the following equations: SOH = QMax Qinitial
[0108] The health status is therefore calculated by taking the ratio between the maximum capacity of the storage system 4 and the initial capacity Q initiale calculated. Maximum capacity refers to the total amount of energy that the storage system 4 can store when fully charged. Maximum capacity is calculated using the following equation:
[0109] In the case of an initial parametric Cauer structure model, the initial capacity is calculated according to the following equations: Qinitial 3600 Where Q(v) is the value of the sum of the different capacities of the storage system 4; Av is the voltage difference defined between two different voltage values from the initial parametric table.
[0110] Calculating the initial capacity Q initialeis achieved by summing the different capacity values of storage system 4 and then summing the value of the sum of the capacities of storage system 4 weighted by the voltage difference Av.
[0111] A value for the capacity of storage system 4 at a given time t is also calculated by the tracking module 6 according to the following equation:
[0112] The capacity at time t is calculated by summing the different capacity values of the storage system 4, then summing the sum of the capacities of the storage system 4, weighted by the voltage difference Av defined between two different voltage values from the parametric table, up to the voltage corresponding to time t. The capacity at time t is calculated from a parametric table extracted from the selected parametric model 18 associated with the voltage value at time t.
[0113] Then, the monitoring module 6 calculates the state of charge and / or the health status of the storage system 4. Regarding the calculation of the state of charge, this is calculated according to the following equation: Q(t) soc = Qinitial
[0114] The state of charge is calculated by taking the ratio between the capacity at time t and the initial capacity Q initiale -
[0115] Health status is calculated by monitoring module 6 according to the following equations: SOH = QMax Qinitial
[0116] The health status is therefore calculated by taking the ratio between the maximum capacity of the storage system 4 and the initial capacity Q initiale calculated. The maximum capacity is calculated according to the following equation:
[0117] Of course, the invention is not limited to the examples that have just been described.
Claims
DEMANDS 1. A method for monitoring an electrochemical energy storage system, forming a storage system (4), the method being implemented by computer and comprising the following steps: - from a voltage measurement V(t) (20) across the terminals of the storage system, selection of a parametric model (12) associated with the measured voltage 7(t) from a set of parametric models, each associated with a corresponding voltage; -adjustment of the selected parametric model (14), from a comparison between the value of the measured voltage 7(t) and an output of said parametric model applied to a current value equal to the current flowing through the storage system; -calculation of a state of charge and / or a state of health (16) of the storage system (4) from at least one parameter of the adjusted parametric model.
2. The method according to claim 1, the method further comprising the following step: - For each among a plurality of distinct predetermined tension levels: ■ From a measurement of the current i(t) flowing through the storage system in response to the application of said voltage step across the terminals of the storage system at a reference date, adjustment of an initial parametric model from a comparison between the measured current value i(t) and an output of said parametric model applied to a voltage value equal to the current flowing through the storage system.
3. Method according to claim 2, wherein at least one parametric model associated with a measured voltage 7(t) from among a set of parametric models receives as input data at least one parameter of the initial adjusted parametric model.
4. A method according to any one of claims 2 to 3, wherein the initial parametric model has an order between two and five.
5. A method according to any one of the preceding claims 2 to 4, wherein the initial parametric model is a Foster structure model or a Cauer structure model.
6. A method according to any one of the preceding claims, further comprising a step of estimating the storage system with an extended Kalman filter.
7. Computer program comprising executable instructions which, when executed by computer, implement the steps of the process according to any one of the preceding claims.
8. A monitoring device for an electrochemical energy storage system, forming a storage system (4), the device being configured to: - from a voltage measurement V(t) (20) across the terminals of the storage system, select a parametric model (12) associated with the measured voltage 7(t) from a set of parametric models, each associated with a corresponding voltage; -adjust the selected parametric model (14), from a comparison between the value of the measured voltage 7(t) and an output of said parametric model applied to a current value equal to the current flowing through the storage system; -calculate a state of charge and / or a state of health (16) of the storage system from at least one parameter of the adjusted parametric model.