METHOD FOR MONITORING AND MANAGING A BATTERY PARK

Description

TECHNICAL FIELD OF THE INVENTION

[0001] The present invention relates to a method for monitoring and managing a battery bank. TECHNOLOGICAL BACKGROUND OF THE INVENTION

[0002] A battery can be characterized by state parameters, also called "indicators," and by management parameters. State of charge, battery health, and temperature are examples of indicators. Management parameters include, for example, battery charging and discharging cut-off thresholds, the battery charging current value, and parameters related to battery thermal management.

[0003] Currently, these various characteristic parameters of the battery are determined from measurements carried out on test benches and in climatic chambers. For example, to determine the parameters used in the algorithms for calculating the state of charge, capacity tests are performed under controlled conditions at different charge and discharge rates, and at different temperatures.

[0004] However, this method has several drawbacks. First, the time and resources required to implement these measurements are significant. Furthermore, the measurements are performed on new batteries, whereas battery aging influences the values ​​of their characteristic parameters. In addition, the conditions under which the measurements are carried out do not accurately represent the actual usage conditions of the batteries, which are inherently variable.

[0005] Another drawback is that measurements are taken on individual electrochemical cells, whereas a battery is composed of several cells connected in series and / or parallel. Consequently, differences in behavior are expected at the scale of the entire battery. For example, thermal behavior differs, and battery temperature affects its instantaneous performance and endurance. Furthermore, in the case of a complete battery, the power consumption of the management electronics influences the battery's self-discharge. Here, self-discharge refers not to the phenomenon of a decrease in the intrinsic charge level of a battery left at rest (related to so-called parasitic reactions), but rather to a low current draw from the battery leading to a decrease in its charge level.There are also problems with dispersion and balancing between cells which have negative effects on instantaneous performance and battery endurance.

[0006] In summary, determining the characteristic parameters of a battery in the laboratory requires significant time and resources, and even then, the values ​​obtained differ from the actual values ​​for the battery in use. Furthermore, this discrepancy will increase over time as the battery ages. SUMMARY OF THE INVENTION

[0007] It follows from the above that there is a need for a simpler and more reliable method for determining characteristic parameters of a battery.

[0008] The present invention aims to meet this need by proposing a method for monitoring and managing a battery bank comprising a plurality of identical batteries, the method being characterized in that it comprises the following steps: collect data relating to only a portion of the batteries in the park; store the collected data in a database; determine from the data stored in the database a characteristic parameter of each of the batteries in the park.

[0009] The prior art is known from documents EP 2 790 262 A1 and US 2013 / 262067 A1, which describe the monitoring of all batteries in a battery bank. The present invention, however, makes it possible to determine a characteristic parameter for all batteries in the bank from data collected on only a subset of the batteries. The number of measurements required is therefore minimized, saving time. Furthermore, the measurements are performed under real-world conditions, thus improving the reliability of the determined characteristic parameter.

[0010] The method according to the invention may also include one or more of the following characteristics considered individually or according to all technically possible combinations.

[0011] According to one implementation method, the process further includes a step of transmitting the characteristic parameter to each of the batteries in the park.

[0012] According to one implementation method, the characteristic parameter is an indicator revealing a battery state or a battery management parameter.

[0013] In one implementation method, the characteristic parameter belongs to a mathematical model relating to battery behavior. The characteristic parameter of the mathematical model is advantageously updated as new data is collected.

[0014] Depending on one implementation method, the mathematical model is either a battery aging model or an algorithm for calculating the state of charge of batteries.

[0015] In one implementation method, the mathematical model is initially defined based on reference data. One advantage is having a starting model already in place, saving time.

[0016] According to one implementation method, the collected data are selected from the charging temperature, the discharging temperature, the charging current, the discharging current and the voltage across the battery terminals.

[0017] In one implementation method, reference data is initially recorded in the database, and the collected data is compared to this reference data. One advantage is the ability to detect design flaws in battery-integrated systems in case of discrepancies.

[0018] In one implementation method, data is collected opportunistically during normal battery operation. One advantage is that it simplifies data acquisition and saves time by eliminating the need for a specific step.

[0019] In one implementation method, data is collected deterministically using a characterization test of the battery component in question. One advantage is that it allows the desired data to be obtained at the desired time.

[0020] In one implementation method, the characterization test is triggered periodically. One advantage is that it allows the characteristic parameter to be updated with newly collected data.

[0021] According to one implementation method, the characterization test is triggered at specific times. One advantage is the ability to configure opportune moments for data collection—that is, moments that will not disrupt the operation of the battery bank. This is the case, for example, when the batteries are not in use, particularly when they are not being discharged to meet a specific need. Charging phases can be considered opportune moments provided that it is possible, within the time available before the next discharge phase, to perform the tests without interfering with the charging process.

[0022] In one implementation method, the batteries are subjected to identical operating conditions, with several characterization tests being successively triggered on different parts of the batteries in the fleet. One advantage is rotating the batteries in the fleet to avoid the same batteries being unavailable at all times. Another advantage is homogenizing the impact of the characterization tests across all the batteries in the fleet. BRIEF DESCRIPTION OF THE FIGURES

[0023] The invention and its various applications will be better understood upon reading the following description and examining the accompanying figures, among which: there figure 1 schematically represents an example of the context for implementing a method for monitoring and managing a battery bank according to the invention; the figure 2 is a functional diagram of an implementation method of the process according to the invention.

[0024] The figures are presented for illustrative purposes only and are in no way limiting to the invention.

[0025] For clarity, identical or similar elements are identified by identical reference symbols in all figures. DETAILED DESCRIPTION OF METHODS FOR IMPLEMENTING THE INVENTION

[0026] There figure 1 shows a battery bank 1 comprising a plurality N of identical batteries 11. Each battery comprises a plurality of electrochemical cells connected in series and / or in parallel.

[0027] The term "identical batteries" refers to batteries manufactured according to the same specifications (i.e., batteries of the same design, both electrically and mechanically), and which are therefore assumed to be identical, or at the very least similar. The various batteries 11 in battery bank 1 thus contain electrochemical cells of the same type.

[0028] The batteries 11 in bank 1 can be intended for mobile use, each battery 11 equipping, for example, an electric vehicle such as a bicycle or a car. Alternatively, the batteries 11 in bank 1 can be intended for static use, for example, to design uninterruptible power supply systems, also known as UPS systems.

[0029] Each battery 11 can be connected to a charging point (not shown). For example, in the case of an electric vehicle, battery 11 is connected to the charging point temporarily when the vehicle is not in use. In the case of stationary use, battery 11 can be permanently connected to the charging point. The charging point can be common to all or some of the batteries 11, or a charging point can be provided for each battery 11. The charging point(s) include a power supply capable of charging the batteries 11.

[0030] In the implementation context illustrated in the figure 1Each battery 11 in battery bank 1 is equipped with an electronic battery management system 12, also known as a BMS (Battery Management System). The management system 12 controls the battery's operation, notably by regulating its charging and discharging. The management system 12 also performs measurements on the battery, such as charging and discharging current and battery temperature. Such batteries 11 Equipped with management systems 12 are for example lithium-ion batteries, sodium-ion batteries or sodium-nickel chloride batteries.

[0031] The management systems 12 communicate with a data collection device 2 configured to collect data provided by the management systems 12. The data collection device 2 is capable of exchanging information with a database 3 via a communication link 4. A human-machine interface 5 is provided to consult the data stored in the database 3 and to send instructions to the database 3. The human-machine interface 5 belongs, for example, to the database 3.

[0032] Advantageously, the management systems 12 and the collection device 2 communicate with each other via a wireless link 6, for example, Bluetooth or using a mobile phone network. Thus, information relating to a battery 11 that is in use can be transmitted to the collection system 2, even when that battery 11 is not connected to its charging point.

[0033] Wired communication methods may also be provided as an alternative. A wired connection can be chosen, for example, when the batteries 11 are not to be moved or to equip charging points for electric vehicles.

[0034] The batteries 11 in bank 1 exhibit usage conditions that may be identical or different. These usage conditions are defined by various elements such as the charge and discharge regimes to which the batteries 11 are subjected, the cycling and rest phases of the batteries 11, and the ambient temperature of the batteries.

[0035] The term "cycling phases" refers to the phases during which the battery undergoes charge and discharge cycles; that is, the phases during which the battery is in use. In this case, the battery is said to be "in cycled mode."

[0036] The term "rest phases" refers to the phases during which the battery is not in use. In this case, the battery is said to be "in calendar mode".

[0037] The batteries can be located in the same place or in different locations. In the latter case, the operating conditions can still be identical, for example if the batteries 11 are each located in an environment where the ambient temperature is regulated to the same target value.

[0038] There figure 2 represents a functional diagram of a process 200 for monitoring and managing battery bank 1 illustrated in the figure 1 , according to an embodiment of the invention. The method 200 includes a step 210 of collecting data relating to only a portion of the batteries 11 of the bank 1. The data is, for example, collected and transmitted to the collection device 2 via the management systems 12.

[0039] The method according to the invention can also be applied to batteries without a management system, such as lead-acid batteries. In this case, the data can be collected directly by the collection device 2, for example by means of a monitoring module comprising measurement means similar to those of management systems 12.

[0040] The data collected are preferably the battery temperature during charging, the battery temperature during discharging, the charging current, the discharging current and the voltage across the battery terminals during charging and discharging (at one or more given times t).

[0041] The collected data is then stored in database 3 during a storage step 220. The collected data is added to the data already present in database 3, the latter coming, for example, from previous collection steps.

[0042] Advantageously, reference data is initially recorded in database 3 before the implementation of process 200. This reference data comes, for example, from technical data sheets provided by the battery manufacturer. 11 or can be obtained through characterization tests carried out beforehand on test benches.

[0043] The process 200 also includes a step 230 of determining, from the data present in the database 3, a characteristic parameter of the batteries 11 of the fleet 1. This characteristic parameter is determined for example by computing means belonging to the database 3, such as a microprocessor.

[0044] A "characteristic parameter" is defined as a parameter related to the state or operation of the battery. A characteristic parameter can be a state parameter or "indicator," such as the battery's state of charge (SOC) or state of health (SOH). It can also be a battery management parameter, such as a charging or discharging cutoff threshold. Management parameters may also relate to battery thermal regulation or current management based on the battery's state of charge. Alternatively, a characteristic parameter may belong to a mathematical model describing battery behavior, such as a battery aging model or a battery state of charge calculation algorithm.

[0045] Thanks to the invention, it is possible to determine the desired characteristic parameter of each battery 11 in battery bank 1 using data collected on only a portion of the bank 1. This minimizes the number of measurements required and therefore saves time. Furthermore, the collected data is obtained through measurements taken under real-world operating conditions of the batteries 11, which improves the reliability of the determined characteristic parameter. In particular, the measurements take into account changes in the battery's condition, especially its aging, and the battery's environment, which is not the case in the prior art.

[0046] Another advantage of the invention is the ability to compare reference data with actual data, a discrepancy being indicative of a design problem in the system incorporating the battery, for example related to parasitic resistances or abnormally high heating.

[0047] Advantageously, the process 200 includes a step 240 of transmitting the characteristic parameter to each battery 11 in bank 1. This is particularly relevant when the characteristic parameter is an indicator, a battery management parameter, or a parameter of a mathematical model. Thus, the characteristic parameter of each battery 11 in bank 1 is updated to allow for optimized monitoring and management of the batteries 11. For example, it is possible to know precisely the current state of charge of the battery 11, and consequently the range of the corresponding vehicle.

[0048] Steps 210 to 240 of the process are advantageously implemented several times during the life of the batteries 11, in order to update the characteristic parameter of each battery 11.

[0049] Data can be collected opportunistically during the normal operation of the battery 11, during charging or discharging (partial or complete). In other words, data can be acquired when the necessary conditions for obtaining it are met. For example, the health status of the battery 11 can be determined when the battery 11 undergoes a cycle during its use, consisting of a full charge (SOC = 100%) followed by a full discharge (SOC = 0%).

[0050] The data can also, or alternatively, be collected deterministically by means of a battery characterization test 11. The characterization test preferably includes a full charge and a full discharge. This test can be triggered at regular time intervals, for example, monthly in the case of applications to uninterruptible power supply systems that place little stress on the batteries 11. The characterization test can also be triggered at specific times, for example, when the vehicle equipped with the battery 11 is reconnected to its charging point after use, or conversely, when the vehicle is ready for use.

[0051] When the operating conditions of the batteries 11 are identical, the characterization test is advantageously triggered alternately on different batteries 11 from bank 1. Thus, by rotating the batteries 11 on which the measurements are performed, all the batteries 11 in bank 1 are not unavailable at the same time. Furthermore, rotating the batteries 11 helps to homogenize the impact of the tests on the entire bank 1, particularly with regard to the aging of the batteries 11.

[0052] The collected data may come from different groups of batteries 11 subjected to different usage conditions. Database 3 is thus enriched with information corresponding to different scenarios but which is valid for all batteries 11. It is therefore possible, from the information contained in database 3, to predict the behavior of a group of batteries 11 under usage conditions to which this group has not yet been subjected but which have been experienced by another group of batteries, without needing to carry out new measurements.

[0053] To illustrate this point, consider, for example, a fleet of electric vehicles divided into several groups located in different cities. In such a case, it is expected that the ambient temperature will differ from one city to another. Thanks to data sharing, the behavior of the batteries can be known for all the different temperatures already recorded across all the cities.

[0054] A first example of the application of the method according to the invention will now be described. This first example focuses on monitoring the health status of the batteries 11 in bank 1 when they are subjected to identical usage conditions. In this case, it is advantageous to use health status measurements taken on a subset of the batteries 11 to construct a model of health status evolution, also called an "aging model," which is valid for all the batteries 11 in bank 1. As new data are collected, this model can be refined to become more accurate.

[0055] In this first application example, a generic model can initially be parameterized using reference data from database 3. Therefore, a large number of laboratory measurements are not required, as the model is built during use. The model's parameters are modified based on newly collected data.

[0056] State of health (SOH) is expressed as a percentage and corresponds to the ratio between the maximum total discharge capacity Qmax and the initial total capacity Q0 of the battery, the latter being able to be assimilated to the nominal capacity specified by the battery manufacturer or to be measured. SOH = Q max Q 0

[0057] The maximum discharge capacity Qmax represents the amount of electrical energy the battery can deliver when fully charged, at any given point in its lifespan. The initial total capacity Q0 is the battery's initial discharge capacity, that is, when the battery is new. The closer the discharge capacity Qmax is to the initial total capacity Q0 of the battery, the better the battery's health. A loss of health, therefore, represents an irreversible decrease in the battery's capacity.

[0058] The following formulas provide an example of an aging model. State of health (SOH) is related to capacity losses (dQ loss) as follows: SOH = Q max Q 0 = Q 0 − dQ loss Q 0

[0059] Capacity losses (dQ loss) are defined as the sum of capacity losses at rest and those during cycling: dQ loss = ∂ Q loss ∂ t dt + ∂ Q loss ∂ Q th dQ th Or : the term ∂ Q loss ∂ t dt represents the capacity losses during rest phases (a function of time t); the term ∂ Q loss ∂ Q th dQ th represents the capacity losses during the cycling phases (function of the quantity Qth of Ah accumulated during successive discharges).

[0060] An example of the equation for each of the preceding quantities is given by the following formulas: ∂ Q loss ∂ t dt = J cal 1 + A cal Q loss ∂ Q loss ∂ Q th dQ th = J cyc 1 + A cyc Q loss Or : Q loss is the capacity lost by the battery, expressed in ampere-hours (Ah); J cal is a first factor of accelerating battery aging in calendar mode; J cycl is a first factor of accelerating battery aging in cycled mode; A cal is a second factor of accelerating battery aging in calendar mode; and A cycl is a second factor of accelerating battery aging in cycled mode.

[0061] The parameters J cal and A cal depend on the state of charge (SOC) and the temperature (T) at rest, while the parameters J cycl and A cycl depend on the state of charge (SOC) and the current and temperature conditions during cycling, in other words the discharge current I d, the charge current I c, the discharge temperature T d and the charge temperature T c.

[0062] This example of an aging model is described in detail in the document [B. PILIPILI MATADI, Doctoral thesis from the University of Grenoble Alpes, "Study of the aging mechanisms of Li-ion batteries in low temperature cycling and high temperature storage: understanding the origins and modeling of aging", 2017].

[0063] In a conventional manner, the parameters J cal, A cal, J cyc and A cyc are identified from health status evolution (HSI) data obtained through laboratory tests under different usage conditions and at the individual cell level.

[0064] In the process according to the invention, these parameters are determined from data collected during the use of the batteries in the park, thus allowing both a saving of time (that of the tests) and better accuracy (by placing oneself directly in the real usage conditions of the complete batteries and no longer of the individual cells).

[0065] Prior to this, default values ​​for the coefficients J cal, A cal, J cyc, and A cyc were recorded in the centralized database for the N batteries in the fleet. These so-called reference values ​​are obtained, for example, from cell / battery data provided by the manufacturer or from laboratory measurements.

[0066] At step 210 of the process (cf. Fig. 2 ), we collect data relating to n park batteries (1 ≤ n < NIn this first example, the data collected consists of usage data—the charging current Ic, the discharging current Id, the charging temperature Tc, the discharging temperature Td, and the voltage during charging and discharging—and state of health (SOH) data. The usage data allows, in particular, the calculation of the state of charge (SOC). The state of health data is obtained by measuring the state of health under different state of charge and temperature conditions, preferably at regular intervals (e.g., every three months).

[0067] The collected data is then recorded in the centralized database during step 220.

[0068] In step 230, the coefficients Jcal, Acal, Jcyc, and Acyc of the aging model are calculated from the collected data, for example by performing a curve fit (the parameters Jcal, Acal, Jcyc, and Acyc are linked to the SOH via dQ loss and depend on usage data and the SOC). This calculation may involve statistical processing of the collected data, particularly to weight outliers and / or reduce processing time.

[0069] The values ​​of the parameters J cal, A cal, J cyc and A cyc stored in the database are replaced by those obtained in step S3, then transmitted to each battery in the park during step 240 (for example to be recorded in a memory of the electronic management system).

[0070] Steps 210 to 240 are advantageously repeated to refine the identification of model parameters (new usage conditions collected and / or new health status values ​​collected).

[0071] Based on such an aging model, it is possible to perform predictive maintenance by determining, in particular, when to replace the batteries 11.

[0072] A second application of the method according to the invention will now be described. This second example focuses on determining the parameters of an algorithm for calculating the state of charge of batteries 11 when they are subjected to different usage conditions. To do this, it is necessary to know the available capacity as a function of operating parameters such as current and temperature. Since the state of charge has been measured separately, it is possible to reconstruct, from data obtained from different batteries used under different conditions, a map of the battery life as a function of usage conditions.

[0073] The battery's state of charge (SOC, expressed as a percentage) can be determined from the following equations: SOC = Q max I ref T ref − Q d + Q c Q max I ref T ref × 100 Q d = α I d T d . I d . t Q c = β I c T c . I c . t Or : Iref and Tref are respectively the current and temperature taken as reference values ​​(for example, Tref = 25°C and Iref = Qn / 2 with Qn the nominal capacity of the battery); Qmax(Iref, Tref) is the maximum capacity of the battery (in Ah) after a full charge of the battery at current Iref and temperature Tref (Qmax varies depending on the SOH); Qd is the discharged capacity, expressed in ampere-hours (Ah), during a discharge phase carried out at a current Id and a temperature Td; Qc is the charged capacity, expressed in ampere-hours (Ah), during a charge phase carried out at a current Ic and a temperature Tc; α and β are discharge and charge equivalence coefficients respectively, in other words, coefficients allowing consideration of the impact of the operating current and temperature conditions on the available discharge capacity or the chargeable capacity, respectively;and t is a given period of time.

[0074] This example of a state of charge algorithm is described in detail in the document [A. DELAILLE, Doctoral thesis from the University of Paris 6, "Development of methods for evaluating the state of charge and the state of health of batteries used in photovoltaic systems", 2006].

[0075] In a conventional manner, the coefficients α and β of the state of charge algorithm are determined from the following relationships, by performing laboratory capacity tests on individual cells under different current and temperature conditions and at a single state of health (SOH = 100%). α = Q max I ref T ref Q max I d T d β = Q max I ref T ref Q max I c T c Qmax(Iref, Tref) is the integral of the reference current Iref during a complete discharge or charge of the battery at the reference temperature Tref. Qmax(Ic, Tc) is the integral of the discharge current Id during a complete discharge of the battery at the charging temperature Td. Qmax(Ic, Tc) is the integral of the charge current Ic during a complete charge of the battery at the charging temperature Tc.

[0076] In practice, the coefficients α and β are matrices of values ​​that depend respectively only on the quantities Id, Td and Ic, Tc (Iref, Tref and Qmax(Iref, Tref) having been previously determined). It is therefore sufficient to collect several pairs of values ​​Id, Td to determine α and several pairs of values ​​Ic, Tc to determine β.

[0077] In the method according to the invention, the coefficients α and β are determined from these same relationships and from data collected directly during the use of the batteries in the park, thus allowing both a saving of time (that of laboratory tests) and better accuracy (by placing oneself in the real usage conditions of the whole batteries, and no longer of the individual cells, and by also considering the state of health of the batteries).

[0078] Prior to this, default values ​​for the α and β coefficients were recorded in the centralized database for the N batteries in the fleet. These so-called reference values ​​are obtained, for example, from cell / battery data provided by the manufacturer or from laboratory measurements (for SOH = 100%).

[0079] Then, at step 210 (cf. Fig. 2 ), we collect data relating to n batteries in the park (1 ≤ n < N ) ,These n batteries are subjected to different current and temperature conditions. Data is collected initially at the beginning of their lifespan (SOH = 100%, i.e., Qmax = Q0) during a full charge and a full discharge. In this second example, the data collected are the current and temperature conditions Id, Td during discharge, the current and temperature conditions Ic, Tc during charge, and the voltage across the battery terminals during discharge and charge.

[0080] The collected data is then recorded in the centralized database during step 220.

[0081] In step 230, the coefficients α (discharge) and β (charge) of the state-of-charge gauge for the n batteries are calculated from the collected data. This calculation may involve statistical processing of the collected data, particularly to weight outliers and / or reduce processing time.

[0082] The values ​​of parameters α and β stored in the database are replaced by those obtained in step S3, and then transmitted to each battery in the bank during step 240 (for example, to be recorded in the memory of the electronic management system). Thanks to the method according to the invention, the state of charge of any battery in the bank can now be determined accurately.

[0083] Steps 210 to 240 are advantageously repeated for new data collected subsequently at the same health status or for other health status values. Before each repetition, the maximum battery capacity Q max (I ref , T ref ) is advantageously requalified by performing a full charge and a full discharge at the reference current I ref and the reference temperature T ref .

[0084] Naturally, the invention is not limited to the embodiments described with reference to the figures. However, the invention is defined by the attached claims.

Claims

1. A method (200) for monitoring and managing a battery fleet (1) comprising a plurality of identical batteries (11), the method (200) being characterised in that it includes the following steps: - (210) collecting data relating to only part of the batteries (11) in the fleet (1); - (220) storing the data collected in a database (3); - (230) determining, from the data stored in the database (3), a characteristic parameter of each of the batteries (11) in the fleet (1).

2. The method (200) according to claim 1, characterised in that it further includes a step (240) of transmitting the characteristic parameter to each of the batteries (11) in the fleet (1).

3. The method (200) according to one of claims 1 and 2, characterised in that the characteristic parameter is an indicator revealing a state of the battery (11) or a management parameter of the battery (11).

4. The method (200) according to one of claims 1 and 2, characterised in that the characteristic parameter belongs to a mathematical model relating to the behaviour of the batteries (11), the characteristic parameter of the mathematical model being updated as new data are collected.

5. The method (200) according to claim 4, wherein the mathematical model is an ageing model of the batteries (11) or an algorithm for calculating the state of charge of the batteries (11).

6. The method (200) according to one of claims 4 and 5, characterised in that the mathematical model is initially defined from reference data.

7. The method (200) according to any of claims 1 to 6, characterised in that the data collected are selected from the charge temperature, the discharge temperature, the charge current, the discharge current and the voltage across the battery.

8. The method (200) according to any of claims 1 to 7, characterised in that reference data are initially recorded in the database (3) and in that the data collected are compared with the reference data.

9. The method (200) according to any of claims 1 to 8, characterised in that the data are collected opportunistically during normal operation of the batteries (11).

10. The method (200) according to any of claims 1 to 8, characterised in that the data are collected deterministically by means of a characterisation test of said part of the batteries (11).

11. The method (200) according to claim 10, characterised in that the characterisation test is periodically triggered.

12. The method (200) according to one of claims 10 and 11, characterised in that the characterisation test is triggered at predetermined instants.

13. The method (200) according to any of claims 10 to 12, characterised in that the batteries (11) are subject to identical usage conditions, with several characterisation tests being successively triggered on different parts of the batteries (11) in the fleet (1).

14. The method (200) according to any of claims 1 to 13, characterised in that the data collected are derived from different groups of batteries (11) subjected to different usage conditions.