Battery management system misconfiguration detection

Abstract

A system and method for determining that a battery management system (BMS) comprising one or more sensors that measure state of charge of a battery of a vehicle or equipment is misconfigured is provided herein. The method comprises measuring two or more voltages of the battery at different states of charge of the battery to calculate a rate of change of the battery with respect to the State of Charge (SoC) of the battery. The calculated rate of change is compared one or more measured parameters of the BMS and determining that the BMS is misconfigured if the measured parameters and the expected parameters indicate a mismatch. Methods of calibrating the BMS are also provided.

Description

[0001] Description

[0002] BATTERY MANAGEMENT SYSTEM MISCONFIGURATION DETECTION

[0003] Technical Field

[0004] The present disclosure relates to determining that a Battery Management System (BMS) used for calculating a state of charge of a vehicle comprising a rechargeable battery is misconfigured. The present disclosure is suited for electric vehicles, vehicles with lead-acid batteries and other equipment comprising lead-acid batteries such as generators (e.g. Commercial and Industrial Generator Sets).

[0005]

[0006] Vehicles comprise batteries that are required to be charged for operation of the vehicle. The battery also provides a power source for any other electrical equipment, also called ancillary electrical devices, that the vehicle has such as water heaters, air conditioning, lights, radio, etc.

[0007] Battery Management Systems (BMS) mounted on a starter battery are configured to measure voltage, current and optionally temperature of the battery. Coulomb counting (CC) can be used to estimate charge lost from the battery using current readings from the BMS measurements.

[0008] There are a number of ways to estimate a State of Charge (SoC) of a battery from battery parameters obtained via signals from battery sensors. The most common method used to calculate SoC requires measurement of parameters including voltage at full charge (Umax), voltage at full discharge (Umin), and battery capacity (C20). The relationship between these measurable parameters is described for example in Coleman, Martin, Chi Kwan Lee, Chunbo Zhu, and William Gerard Hurley. " State-of-charge determination from EMF voltage estimation: Using impedance, terminal voltage, and current for lead-acid and lithium-ion batteries." IEEE Transactions on industrial electronics 54, no. 5 (2007): 2550-2557. Whilst capacity measurements are typically readily available,voltage information is often not available to an end user or if provided by a manufacturer may not be accurate for a specific battery.

[0009] Due to significant battery-to-battery variability and variation, for example due to age of the battery, measuring SoC based on the above parameters can be inaccurately calculated.

[0010] Accordingly, there remains a need for determining that the sensors are calibrated correctly.

[0011] According to a first aspect, there is provided a method for determining that a battery management system (BMS) of a vehicle or equipment is misconfigured, the vehicle or equipment comprising a lead acid rechargeable battery. The method comprises measuring a first voltage, OCVX1, of the battery at a first state of charge (SoC), SoCxl, of the battery and a second voltage, OCVx2,of the battery at a second SoC, SoCx2, of the battery, where SoCx1≠ SoCx2. OCVx1at SoCx1and OCVx2at SoCx2are used to calculate a rate of change of voltage of the battery with respect to charge of the battery. The calculated rate of change is used to compare one or more measured parameters of the BMS with one or more expected parameters of the BMS. If said comparison indicates a mismatch between the calculated BMS parameter(s) and the expected BMS parameter(s), the method further comprises determining that the BMS is misconfigured and requires recalibration.

[0012] In one embodiment, the BMS parameters comprise one or more of: a voltage at full charge, Umax, of the battery, and a voltage at full discharge, Umin, of the battery. In other embodiments, a BMS parameter that is compared comprises an expected rate of change (e.g. a theoretical OCV-SoC curve) based on an expected voltage at full charge, Umax, of the battery, an expected voltage at full discharge, Umin, of the battery, and an expected capacity, C20, of the battery. Comparing includes direct value comparison as well as a “distance” between an expected (e.g. theoretical) OCV-SoC curve and a curve fit of the measured values.According to a second aspect, there is provided a vehicle or equipment comprising a battery, a battery management system comprising one or more sensors and a controller configured to perform the method according to the first aspect. The vehicle is an electric vehicle in some embodiments. In other embodiments, the vehicle comprises a lead acid rechargeable battery.

[0013] These and other aspects and features of the present disclosure will be more readily understood after reading the following description in conjunction with the accompanying drawings.

[0014]

[0015] of the

[0016] A more complete understanding of the subject matter may be derived by referring to the detailed description and claims when considered in conjunction with the following figures, wherein like reference numbers refer to similar elements throughout the figures.

[0017] Figure 1 illustrates a flow diagram of a method for determining that a battery is misconfigured according to an embodiment of the present disclosure;

[0018] Figure 2 illustrates a graphical representation of the linear relationship between SoC and OCV of a battery according to an embodiment of the present disclosure;

[0019] Figure 3 illustrates a flow diagram of a method for calibrating a battery sensor for measuring a state of charge (SoC) of a battery using Umax, for example in response to a determination that the battery is misconfigured;

[0020] Figure 4 illustrates a flow diagram of a method for calibrating a battery sensor for measuring a state of charge (SoC) of a battery using Umaxand Uminfor example in response to a determination that the battery is misconfigured;

[0021] Figure 5 illustrates an example of an electric vehicle in the form of a large mining truck, or haul truck.

[0022] Detailed

[0023]

[0024] The following detailed description is merely illustrative in nature and is not intended to limit the embodiments of the subject matter or the applicationand uses of such embodiments. Any implementation described herein as an example is not necessarily to be construed as preferred or advantageous over other implementations. Furthermore, there is no intention to be bound by any expressed or implied theory presented in the preceding technical field, background, or the following detailed description.

[0025] The present disclosure addresses a problem of determining misconfiguration of battery sensors. In one example, the method is suited for batteries in electric vehicles although the scope is not so limited and can be applied to any lead-acid battery in a vehicle such as a starter battery or the like or other equipment comprising rechargeable lead acid batteries such as generators and gensets. A genset is a portable power supply source consisting of an engine and a generator. There is significant battery-to-battery variability and variation due to age that affects battery parameters, as well as user error that can affect accurate readings of battery parameters. The solution provided herein provides a method of determining misconfiguration of an electric vehicle battery and associated sensors specific to the battery that is installed in a given vehicle which can be identified by a battery management system in the field, rather than a laboratory setting. The machine alternator, or other power source like the lithium-ion pack in a Battery Electric vehicle (BEV), and controlled loads to a sensor which is part of the Battery Management System (BMS) are used to calculate whether the battery is configured or not in the field, rather than in a lab, using on-machine data.

[0026] Ensuring timely detection of incorrect configurations using the methods described herein can prevent incorrect SoC calculations and can help to avoid undesired system behaviour such as overcharging of a battery by a smart charging system or false triggers to a low-battery alert system. Incorrectly measured SoC caused by battery misconfiguration can also lead to leaving a battery in low charge for extended periods which can lead to permanent sulfation and battery damage. This can be prevented if the battery misconfiguration is detected and mitigated for.

[0027] According to a most commonly used algorithm for measuring battery state of charge (SoC), Open Circuit Voltage (OCV) measurements are takenor calculated from sensor data of the BMS. Calibrating the battery sensors uses a voltage at full charge of the battery, Umax, and a voltage at full discharge of the battery, Umin, using an equation based on the Nernst equation. In electrochemistry, the Nernst equation is a chemical thermodynamical relationship that permits the calculation of the reduction potential of a reaction (half-cell or full cell reaction) from the standard electrode potential, absolute temperature, the number of electrons involved in the redox reaction, and activities (often approximated by concentrations) of the chemical species undergoing reduction and oxidation respectively.

[0028] Electric work vehicles, such as electric excavators, electric mining trucks, electric articulated trucks, electric wheel-loaders and the like are often used at worksites where the electric vehicles may be used by different operators and / or be supplied with different battery packs. For example, the worksite operator may rent batteries or have subscription agreement for a third-party to supply batteries for the work machines at the worksite. This variation in batteries being used in a machine means that there is a desire for effective battery management system that can detect when a battery is misconfigured. Optionally, in response to a misconfiguration being detected, a user can be alerted to the misconfiguration and a recalibration can be performed. In some examples, the misconfiguration and recalibration are both performed in the field on the electric work machine using data available from the battery management system.

[0029] Figure 1 illustrates a flow diagram of a method 100 for determining that a BMS of an electric vehicle, a vehicle comprising a lead-acid battery or other equipment comprising a lead-acid battery is misconfigured. In particular, the method illustrated comprises determining that a battery sensor for measuring an SoC of a battery is misconfigured. SoC describes the battery's current charge level as a percentage of its capacity. It is important for determining how much energy is left in the battery, for example to calculate when the battery will need to be recharged. Capacity is a measure of the entire energy potential of a battery and it is usually expressed in ampere-hours (Ah). It provides information on how much charge the battery can deliver at a particular discharge rate.The method may be performed by a controller, for example by an electronic control unit (ECU) or the BMS which controls and monitors the battery. One or more battery sensors are coupled to the battery, as part of the BMS, and are configured to take measurements therefrom. Sensors include those suitable for measuring charge, voltage and optionally temperature of the battery. In one example, the battery sensor is a shunt-resistor-based battery sensor, although other alternatives are available that are suitable for making measurements according to the present disclosure.

[0030] Measurements taken by the battery sensor(s) include OCV which is a parameter of a battery cell and measured in Volts. The OCV of a battery cell is the potential difference between the positive and negative terminals when no current flows and the cell is at rest. An OCV curve is dependent on the chemistry of the cell and describes the charge and discharge of the battery, which can be nonlinear. Hysteresis in charge vs discharge curves can result in errors of SoC if the BMS uses cell voltage to estimate SoC. However, calibration methods described herein helps to calibrate the sensors for more accurate measurements.

[0031] The method 100 for determining that a BMS of a vehicle or other equipment is misconfigured comprises a number of steps as described below.

[0032] In a first step 110, the method comprises measuring a first voltage, OCVX1, of the battery at a first SoC, SoCxl, of the battery, and a second voltage, OCVx2, of the battery at a second SoC, SoCx2, of the battery, where SoCx1≠ SoCx2. The measurements are preferably taken within the same discharge cycle of the battery. In an example comprising an electric vehicle, readings of OCV in can be taken during a “parked mode” of the electric vehicle, for example automatically when the parked mode is detected. Sensors that form part of the BMS are configured to measure the voltage and change in SoC. In some embodiments, SoCx1is at full charge of the battery, however, it can be measured when the battery is not fully charged.

[0033] It should be noted that voltage of the battery ~ OCV of the battery at low discharge currents (< 0.25 A) or the voltage of the battery when the battery is disconnected from any load. This is described for example at Perez, Richard." Lead-acid battery state of charge vs. voltage." Home power 36 (1993): p. 66-69. If the discharge currents are high (> 0.25 A), approximations of OCV can be calculated as a function of voltage and current, which are both measurable, and modelled battery impedance. Known models like power-law relationships between loaded current and OCV can be used to estimate the OCV for different discharged values of the battery in Ah.

[0034] In some embodiments, a value of SoC may be known, for example if the battery is at full charge. However, in other embodiments, where a value of SoC cannot be directly measured (e.g. battery not at full charge but some battery energy has been spent to an unknown quantity), a difference in energy released by the battery between SoCxland SoCx2, e.g. a change in SoC, is measured using Coulomb counting. Coulomb counting is used to measure a change of energy measured in discharged Ah from the battery between the two readings of OCV.

[0035] In some embodiments, SoCxl, SoCx2, vary by at least 5% SoC. This helps to provides greater reliability in later calculations.

[0036] A temperature of the battery at the time the first and second voltages, OCVx1, OCVx2, are measured is taken in some embodiments. The method further comprises temperature correcting the measurements of the first and second voltages, OCVx1, OCVx2, based on the measured temperature. Parameters are temperature-corrected (typically 3 mV / °C) to room temperature with thermocouple data that is measured by sensors that form part of the BMS and any other sources.

[0037] In a second step 120, OCVx1at SoCx1and OCVx2at SoCx2are used to calculate a rate of change of voltage with respect to SoC. The calculated rate of change can be referred to as a “live slope” which is calculated based on readings of OCV and SoC taken by the integrated machine system in a parked mode of the electric vehicle. The relationship between the values of OCV and SoC as the battery discharges is linear to a first approximation, particularly where the values fall in a range of the SoC of the battery from around 20% SoC to 80% SoC. Figure 2, described in more detail below, illustrates the relationship between the measured values and how they are used to calculate the rate of change.In a third step 130, the method further comprises comparing one or more measured parameters of the BMS against one or more expected parameters of the BMS.

[0038] In one example, the method comprises comparing a measured voltage at full charge, Umax, of the battery with an expected voltage at full charge, Umax, °f the battery. An alternative embodiment comprises comparing a measured voltage at full discharge, Umin, of the battery with an expected voltage at full discharge, Umin, of the battery. In practice, running the battery down to empty can be harmful to the battery itself and reduces battery longevity. Using the intermediate values advantageously allows calibration to be performed without running the battery to empty.

[0039] In one example, the calculated rate of change is compared with an expected rate of change derived from parameters of the BMS comprising an expected voltage at full charge, Umax, of the battery, an expected voltage at full discharge, Umin, of the battery, and an expected capacity, C20, of the battery.

[0040] The expected rate of change is measured based on the Nernst equation and can be estimated using the following equation:

[0041] (Umax− Umin)

[0042]

[0043] Capacity

[0044] Where Umaxis battery voltage at full charge, Uminis battery voltage at full discharge, and capacity is the capacity of the battery. The capacity of the battery is typically a trusted value received for example from a manufacturer of the battery. However, the voltage at full charge and full discharge of the battery, whilst they may be provided by the manufacturer, are less reliably known and more widely variable.

[0045] Tests have shown that a linear relationship between Umax, Uxl, Ux2, Uminas described above is a good approximation suitable for deriving values by extrapolation according to the method described herein. The relationship is based on the thermodynamics of how lead acid batteries work as described for example in Coleman, Martin, Chi Kwan Lee, Chunbo Zhu, and William Gerard Hurley: " State-of-charge determination from EMF voltage estimation: Usingimpedance, terminal voltage, and current for lead-acid and lithium-ion batteries." IEEE Transactions on industrial electronics 54, no. 5 (2007): 2550-2557.

[0046] Comparing the parameters may comprise comparing using a Root Mean Square (RMS) metric which calculates a sum of the difference of squares between the expected and measured parameters. Alternative comparisons include comparing a sum of the absolute differences between the expected and measured parameters.

[0047] According to a fourth step 140, if the comparison is indicative of a mismatch between the measured and expected parameters, determining that the BMS is misconfigured and requires recalibration. For example, determining that there is a mismatch between one or more of: Umax, Umin, and / or between the measured rate of change with an expected rate of change. A condition of the comparison being indicative of a match is when the comparison indicates either that the expected and measured parameters match exactly or that they are approximately equal and within a small margin of error, such as one standard deviation.

[0048] In some embodiments, recalibration is triggered if a difference between the measured parameters and the expected parameters is above a predetermined threshold. For example, in one embodiment recalibration is triggered if a difference between the measured rate of change and the expected rate of change deviates by an amount above a predetermined threshold. In other embodiments, recalibration is triggered if a measured Umaxdeviates from the expected Umaxby a predetermined amount.

[0049] In a fifth optional step 150, the BMS is recalibrated. Recalibration comprises determining estimated values of Umax' and Umin'.

[0050] According to one embodiment, the method comprises calibrating the battery using a first method. The first method comprises charging the battery until it has reached full charge at SoCmax, and measuring a voltage at full charge, OCVmax, of the battery at SoCmaxto determine Umax' ≈ OCVmax. When a battery such as a lead-acid battery is first connected to a constant voltage power supply,the initial charging current is high because the battery is in a discharged state and the voltage difference between the battery and the power source is large. This current will gradually decrease as the battery voltage rises and the internal resistance increases. As the battery charges, the chemical reactions inside the battery generate an opposing voltage, and the battery’s SoC increases. As the SoC rises, the charging current decreases because the voltage difference between the charger and the battery narrows. The charging current will continue to drop until the battery reaches near full charge. Lead-acid batteries are considered fully charged when the charging current drops to a value that is typically 3-5% of the battery’s ampere-hour (Ah) rating. For example, for a 100 Ah battery, full charge would be reached when the charging current drops to 3-5 A. This is the point at which the battery is no longer able to accept a significant amount of charge and the voltage stabilizes at the charging voltage (often around 2.25-2.30 V per cell for lead-acid batteries). SoCmax(full charge) is defined by when the charging current drops below the target threshold (typically 3-5% of the rated Ah). This threshold helps avoid overcharging, which can damage the battery by causing excessive heat and gassing, leading to reduced lifespan.

[0051] In one example, the battery is charged continuously by an alternator of the vehicle at Ualternator(~28 V) until the charging current indicates full charge, for example when no more charge is accepted and the charging current measures under a threshold amount, for example 0.1 A. The SoC is specified as being at 100% when fully charged. The measurement of maximum SoC, SoCmax, anchors the calibration to a known and reliable measurement; i.e. when the battery is fully charged, it is known that the SoC is 100% to a degree of accuracy.

[0052] The battery can be left to settle for a resting period such that an equilibrium of the battery voltage is reached. This helps to provide a more accurate measurement of the OCVmax. It is usually required to rest for a period of between 1 to 15 hours, more preferably between 1 and 6 hours, for example 3 hours. The measurement of OCVmax is made at the end of a resting period.

[0053] Battery temperature can also be used in estimating values such as OCVmax, so that the voltage can be temperature corrected (to room temperature,typically measured at 25 degrees Celsius). This is performed by measuring a temperature of the battery at the time the voltages OCVmaxand OCVXare measured and temperature correcting the measurements of the voltages OCVmaxand OCVXbased on the measured temperature. Voltage readings are corrected at approximately 3 mV / °C.

[0054] The first calibration method further comprises partially discharging the battery to a first intermediate SoC of the battery SoCxl', where SoCmax> SoCxl' > SoCmin, wherein SoCminis at full discharge of the battery. After partially discharging the battery, a first intermediate voltage, OCVX1of the battery at SoCxl' is measured. The method further comprises extrapolating from OCVmaxat SoCmaxand OCVx1' at SoCx1', to estimate a voltage at full discharge, Umin' of the battery at SoCmin. In some embodiments, OCVX1' and SoCxl' are different values to those of OCVX1at SoCxlmeasured in the misconfiguration detection above in step 110. However, in some embodiments, they are the same.

[0055] The first intermediate SoC of the battery, SoCxl', is less than full charge, SoCmax, but greater than full discharge, SoCmin, of the battery. It will be appreciated that the first intermediate SoC of the battery, SoCxl', can be at any value less than 100%, however, it is preferable to take a first measurement of the first intermediate SoC of the battery, SoCxl', at a value equal to or above around 50% at least because this reading can be taken earlier compared to measurements which require the battery to be drained beyond 50%. Whilst it may be more reliable to take regular readings at small increments of the SoC as the battery discharges, sufficient accuracy can be achieved by taking a reading at larger increments, for example at 90% and below. It is also suitable to take just one reading of the first intermediate SoC of the battery, SoCxl', to perform the calibration method described herein. Values from 50% to 95% are suitable, with preferable values between 60% to 80%.

[0056] The first calibration method may further comprise, at a second intermediate SoC of the battery SoCx2' which is different to the first intermediate SoC, SoCxl', measuring a second voltage reading, OCVx2'. It will be appreciated that any number of measurements, N, of the SoC can be taken as data points butthat a single intermediate data point in addition to the data point at SoC = 100% is enough to be able to perform the calibration as described herein.

[0057] According to one embodiment, a first intermediate SoC of the battery, SoCxl', is measured at SoC = 80% and a second intermediate SoC of the battery, SoCx2’, is measured at SoC = 60%. It will be appreciated that this is an example only, and that other values of SoC are available.

[0058] Temperature correction based on a measured temperature of the battery at the first, second, and any further measured intermediate voltages can be performed by measuring the temperature of the battery at the time the reading is taken by the one or more sensors and temperature correcting the values accordingly.

[0059] The battery itself can be discharged in different ways between the readings. Two examples of different techniques used to discharge the battery during calibration are discussed herein, other techniques may be available.

[0060] A first battery discharge option comprises discharging the battery in a controlled manner using fixed current discharge between SoCmaxand SoCx. In some examples, the electrical system of the machine or equipment (e.g. a controller) is used to discharge the battery according to a known and fixed current and impedance load. Other loads can be optionally added to alter the discharge current rate. The battery is discharged to a desired SoC value and left to settle to equilibrium before a measurement is taken.

[0061] In combination with Coulomb counting, the SoC of the battery can be calculated. Coulomb counting is a book-keeping method where the charge transferred through the battery during full charge-discharge process is counted by monitoring the current continuously using current sensors of the BMS. Thus, the transferred amount of ampere hours are tracked and consequently the remaining capacity is known when starting from a battery which is fully charged (i.e. at full capacity).

[0062] When the battery is discharged using fixed current discharge, Coulomb counting can be used to track battery discharge and stop battery discharge at a predetermined value of SoCxto measure an intermediate voltage, OCVX. Insome cases, however, SoC of the battery may not be known. In these embodiments, Coulomb counting can be used to track battery discharge and stop battery discharge at a predetermined value of depth of discharge (DoD) of battery capacity to measure the intermediate voltages, OCVX.

[0063] A second option for discharging the battery during the calibration is available, for example if controlled discharge is not suitable due to additional components added to the discharge load or time availability. In a first scenario, the battery is discharged at low current, for example at around < 0.25 A, for an extended period before measuring OCVX. This extended period can take place, for example, when an electric vehicle is parked over the weekend or for up to a few weeks without being used. The battery loses charge due to leakage, for example. A voltage reading of the battery in parked mode can be approximated as the OCV. Coulomb counting can be used in this example by measuring the charge output to give an approximation of SoC. Coulomb counting can be used to estimate the change in SoC as total Coulomb discharge divided by total battery capacity. Discharge of the battery between 5% and 10% of the battery is suitable to perform calibration using this discharge method. Different set ups will take different amounts of time in parked mode to reach a suitable discharge of the battery using the second discharge option.

[0064] In a second scenario, the battery is discharged at high current >0.25 A, e.g. with loaded discharge. Voltage of the battery ~ OCV of the battery at low discharge currents (< 0.25 A). Direct measurement of the OCV via measuring the battery voltage is therefore not available at high current discharge but corrections such as battery impedance models or Puekert’s law can be used to correct for the effect of higher currents if low-current data is not available. An estimate of OCV is calculated as a function of measured values of voltage and current in combination with modelled battery impedance. Known models like power-law relationships between loaded current and OCV can be used to estimate the intermediate OCVXfor different discharged values of the battery in Ah.

[0065] Calibration of Umincomprises extrapolating from OCVmaxat SoCmax and OCVX1' at SoCxl', to estimate a voltage at full discharge, OCVmin, ofthe battery at SoCmin, where Umin« OCVmin. A relationship between the data points Umaxand Uxl' can be calculated and used to extrapolate an intercept, Umin, of the graph at SoC = 0% = SoCmin. The intercept data point, Umin, at SoC = 0% determines a value of a voltage at full discharge, OCVmin, of the battery at full discharge, SoCmin. Extrapolation can be visualised, for example, by considering a relationship between voltage and SoC of the battery by plotting a graph of measured SoC against measured OCV. Such a graph is illustrated and further described below in relation to Figure 2.

[0066] A second calibration method is also available which comprises using the calculated rate of change to extrapolate values of Umax' and Umin'. Estimated values of Umax' and Umin' are taken by extrapolating based on the rate of change to SoC of the battery at full charge, SoCmax, and full discharge, SoCmin, respectively. This can be performed when values of the SoC are available at the time the voltage measurements are taken.

[0067] The above method can be repeated to improve confidence in the measured values a number of times until there is confidence that the values of Umax' and Uminareaccurate. This process is detailed further in relation to Figures 3 and 4 below.

[0068] The two calibration methods can be used in combination with each other without limitation.

[0069] The values of Um' and Umi' derived via the first or second calibration method can be individually compared to the expected values of Umaxand Umin. The method comprises comparing the estimated voltage at full charge, Umax', to the expected voltage at full charge, Umax, and if the comparison indicates Umax' ≈ Umax, determining that Umaxis calibrated; or if the comparison indicates Umax'

[0070]

[0071] Umax, determining that Umaxis misconfigured and requires recalibration. Umax' ~ Umaxwhen the values approximately match one another, for example if they have the same value or are within a threshold difference of one another.Alternatively or additionally, the method comprises comparing the estimated voltage at full discharge, Umin', to the expected voltage at full discharge, Umin, if the comparison indicates Umin' ≈ Umin, determining that Uminis calibrated; or if the comparison indicates that Umin' Umin, determining that Uminis misconfigured and requires recalibration. Umin' ~ Uminwhen the values approximately match one another, for example if they have the same value or are within a threshold difference of one another.

[0072] Umax comparisons could be performed every time full charge is detected, although this is not a strict requirement.

[0073] Measuring the rate of change can be repeated several times on different discharge cycles of the battery, for example. In one example, the rate of change can be measured again after a calibration has been performed to check that the calibration has been successful.

[0074] A method of repeating the measurements for calculating a second measured rate of change comprises measuring a third voltage, OCVx3, of the battery at a third SoC, SoCx3, and a fourth voltage, 0CVx4, of the battery at a fourth SoC, SoCx4, where SoCx3SoCx4. In some embodiments, SoCx3= SoCxl, and SoCx4= SoCx2, however, these values are not limited as such and do not have to be taken at the same SoC as the first measurements. Using OCVx3at SoCx3and OCVx4, at SoCx4, a second rate of change of voltage with respect to SoC is calculated. The second measured rate of change is compared to the first measured rate of change and (i) if the comparison indicates that a difference between the values of the first and second rate of change is below a predetermined threshold, increasing a confidence level for the values of the expected voltage at full charge, Umax, and the expected voltage at full discharge, Umin, or (ii) if the comparison indicates that the difference is above the predetermined threshold, setting the confidence level to zero and determining that the BMS is misconfigured and requires recalibration. In alternative embodiments, different BMS parameters to the rate of change (such as Umax)arecompared to one another in the comparison stage of the method.

[0075] The expected voltage at full charge, Umax, and the expected voltage at full discharge, Umin, are calibrated after achieving a match between themeasured and expected BMS parameters a predetermined number of times. In one example, a match is achieved when the measured value of the rate of change and the expected value of the rate of change match a predetermined number of times.

[0076] A match is defined when the compared values are below a predetermined threshold. Calibrated values improve SoC calculations made by the BMS and helps to avoid charging related errors.

[0077] Figure 2 illustrates a graphical representation of the linear relationship between SoC and OCV of a battery. SoC (%) is plotted along the x axis and OCV (V) is plotted along the y axis. The values in Figure 2 are of an example battery used for illustrating the concept only. It will be appreciated that batteries vary considerably, and that Figure 2 is merely demonstrative in this regard.

[0078] Measurements of the battery voltage (OCV) are taken at two different times where the SoC of the battery is less than full charge. Illustrated in Figure 2, values of the OCV at two different SoC are taken; at SoC = 80% and SoC = 60%. These have been plotted as data points UX1and Ux2respectively in Figure 2.

[0079] In some embodiments, the estimate of measured voltages UX1and Ux2(and any further readings) are temperature corrected. This is achieved by measuring a temperature of the battery at the time the battery voltage is taken and temperature-corrected to room temperature at 25°C (typically at 3 mV / °C).

[0080] When plotted on the graph of SoC (%) along the x-axis and OCV (V) along the y-axis, a linear relationship between the data points Umax, Uxl, Ux2, and Umaxcan be seen. A first rate of change of is calculated from the values of UX1and Ux2taken at yi = OCVX1of the battery at xi = SoCxlof the battery and y2 = OCVx2of the battery at X2 = SoCx2, of the battery.

[0081] When the measurements are taken at known SoC, by extrapolating these data points using the linear relationship, an intercept of the graph at SoC = 100% and at SoC = 0% can be calculated. At the intercept where x = 0 % on a graph having SoC along the x-axis, can be found, which determines a valueof a voltage at full discharge, OCVmin, of the battery at full discharge, SoCmin. A value of SoCmaxis estimated at 100% SoC by extrapolating to x = 100 % SoC.

[0082] An alternative way of measuring Umax' comprises measuring a maximum SoC of the battery, SoCmax, at 100%, i.e. when the battery is fully charged. The battery is rested for a resting period after full charge has been achieved to allow the battery to come to equilibrium. In some examples, the resting period is between 1 and 6 hours at low current. A measurement of the battery voltage, OCVmax, is taken when the battery is fully charged (i.e. at SoCmax) and rested to equilibrium by a sensor coupled to the battery. This voltage is provided as an estimate of the voltage at full charge Umax. A sensor coupled to the battery, e.g. a sensor that forms part of the BMS, measures the OCV in Volts (V).

[0083] In alternative embodiments not illustrated in Figure 2, current discharged from the battery determined via Coulomb counting measured in ampere hours (Ah) is plotted along the x axis instead of SoC.

[0084] Figure 3 illustrates a flow diagram of a method for calibrating a battery sensor for measuring an SoC of a battery using Umax, for example in response to a determination that the battery is misconfigured.

[0085] According to first step 302 a BMS is installed on the battery. This may be due to a change of battery for example.

[0086] According to a second step 304, values of Umin, Umax, and capacity are input to the BMS. These values may be provided by a manufacturer in some examples or may be values determined by a previous calibration exercise. If no values are provided or available, default values can be input for example to initialise an algorithm.

[0087] According to a third step 306 a confidence level on the value of Uminand / or Umaxis set to zero (e.g. no confidence).

[0088] According to a fourth step 308 the battery is charged until full charge of the battery is detected, for example when a “battery full” condition is met. Additionally or alternatively, full charge may be indicated by the charging current dropping below a threshold value. In one example, the alternator is switched on to charge the battery until it is fully charged. The alternator typicallycharges the batteries using Constant Current Constant Voltage (CCCV) charging. CCCV is characterised by high initial current when the voltage is low with decreasing voltage as the voltage gradually increases. A typical voltage used to charge the batteries using an alternator is 28 Volts. In other examples, the battery is charged using other methods such as from an external energy source.

[0089] If the charge current is below a threshold current, Ith, the charge is switched off. If the charge current is above the threshold current, Ith, the battery is continued to be charged until the charge current is less than the threshold current, Ith. In one example, the threshold current, Ith, is 0.1 A.

[0090] At step 310, when the charge is switched off, the battery is allowed to settle for a resting period with low current and / or no load applied until it reaches equilibrium. Typical equilibrium resting times are from one to six hours, though other times may be used. Equilibrium describes a state when the battery voltage has settled and is not fluctuating such that the value of OCV measured is representative of the actual voltage of the battery at the determined SoC. The resting period depends on the internal characteristics of each battery and is modelled using resistance and capacitance values. A time constant is determined which is indicative of an approximate time for the voltage value to reach the OCV. The exact equation is described by Equation (5): Moo, C. S., K. S. Ng, Y. P. Chen, and Y. C. Hsieh; " State-of-charge estimation with open-circuit-voltage for lead-acid batteries." In 2007 Power Conversion Conference JCagoya, pp. 758-762. IEEE, 2007.

[0091] After the resting period, the voltage of the battery, ()CVmax, is measured, wherein OCVmax~ Umax'. Accordingly, Umax' is calculated to a first approximation.

[0092] At step 312, the measured Umax' is compared to the Umaxof step 304. If the difference between the two compared values is below a predetermined threshold, according to step 314, the confidence level is increased for Umax. The comparison indicates that the values match when the difference is below the predetermined threshold. After a number of matches, for example four matches,the confidence level can be set to 100%, i.e. the value of Umnxis known with confidence.

[0093] If the difference between the two compared values is above the predetermined threshold, according to step 316, the confidence level is set to zero for Umax. The comparison indicates that the values indicate a mismatch when the difference is above the predetermined threshold.

[0094] Accordingly, a misconfiguration of Umaxwarning is triggered. A calibration can be triggered as a result of this. In some embodiments, the estimated value of Umax' can replace the expected value of Umaxin the system as stored and used by the BMS.

[0095] Optionally, steps 306 to 316 can be repeated a number N times until a confidence of 100% in Umnxcan be determined.

[0096] The value of Umincan also be measured to determine a misconfiguration of the battery. Figure 4 illustrates a flow diagram of a method for calibrating a battery sensor for measuring a state of charge (SoC) of a battery of an electric vehicle using Umaxand Uminfor example in response to a determination that the battery is misconfigured. It will be appreciated that the example below relates to an electric vehicle, however, the method can be suitably adapted for vehicles which are not electric vehicles and other equipment comprising lead-acid batteries.

[0097] The first steps 402, 404, 406 correspond with steps 302, 304 and 306 above. In some embodiments, the methods of Figures 3 and 4 can run in parallel or as one combined process for the two methods illustrated in Figures 3 and 4.

[0098] At step 408 an electric vehicle “parked mode” is detected. Parked mode can be detected, for example, if the current measured is less than a threshold current for a given duration. The durations in one example is three hours or more, however, alternative durations suitable for determining that the electric vehicle is parked can also be used.

[0099] At step 410 measurements of the battery are continuously taken for a specified duration. In one embodiment, a voltage output in Volts is measuredalong with a measured electric charge lost from the battery using Coulomb counting. A calculated change in SoC can be calculated by using Coulomb counting (CC) as a ratio of the charge lost to the battery capacity.

[0100] At step 410 a best-fit slope of battery voltage vs charge (either by CC or change in SoC depending on the available measurements) comprising the calculated rate of change for the specified duration is calculated. The measured value of the rate of change provides a first estimate of the voltage with respect to change in SoC (or CC if this is plotted along the x-axis instead of SoC).

[0101] At step 412 the calculated rate of change is compared with an expected rate of change. If the values match within a threshold range, a confidence level of the values of Umaxand Uminis increased at step 414. The comparison indicates that the values match when the difference is below the predetermined threshold. After a number of matches, for example four matches, the confidence level can be set to 100%, i.e. the value of Umnxis known with confidence. Confidence can be determined simply as confident or not confident, without referring to percentages in alternative embodiments.

[0102] If the difference between the two compared values is above the predetermined threshold, according to step 316, the confidence level is set to zero for Umaxand Umin. The comparison indicates that the values indicate a mismatch when the difference is above the predetermined threshold.

[0103] Accordingly, a misconfiguration of Umaxand Uminwarning is triggered. A calibration can be triggered as a result of the misconfiguration. In some embodiments, the estimated value of Umax' can replace the expected value of Umaxin the system as stored and used by the BMS. The calculated values of Umax' and Umin stored in the system can be used in calculating the SoC of the battery in normal use of the electric vehicle with improved accuracy. Calibration can be performed again, for example if the readings become unreliable or if the battery is replaced with a new one.

[0104] Figure 5 illustrates an example of an electric vehicle 500 in the form of a large mining truck, or haul truck. It will be appreciated that this is an example only and that the electric vehicle 500 could be any kind of electric vehicle orelectric work vehicle. The electric vehicle 500 comprises a frame 505, an electric motor 510 supported by the frame 505, and a drivetrain 515 being operatively driven by the electric motor 510.

[0105] The electric motor 510 is powered by chemical energy stored in a rechargeable battery pack. No limitation is intended herein for a composition or topology of the rechargeable battery pack, which may be lead-acid, lithium-ion, nickel-metal hydride, etc. A battery pack may comprise a single power storage module, or a plurality of power storage modules. A battery pack may be referred to elsewhere as a battery.

[0106] The drivetrain 515 of the electric vehicle features the wheels and tires as shown, or it may engage the ground in a separate fashion, such as by employing crawler belts, tracks, treads, and the like, in order to propel the electric vehicle 500.

[0107] The electric vehicle 500 further comprises a vehicle inlet 520 or charging port that is operatively and electrically connected to the rechargeable battery pack and designed to receive electrical power from an external source. In some embodiments, the vehicle inlet 520 may embody a common connector and in other embodiments the vehicle inlet 520 may use a proprietary connection format unique to the type, make, or model of the electric vehicle 500. It should be understood that, while the vehicle inlet 520 is shown to be located on a lower section of the frame 505, this is for illustration only, and the vehicle inlet 520 may be located elsewhere on the vehicle 500 without limitation. Where the electric vehicle 500 is a work machine, the placement and orientation of the vehicle inlet 520 may differ significantly across different vehicle types, makes, and / or models, which may themselves range in size and shape.

[0108] The electrical components of the electric vehicle 500 illustrated in Figure 5 may comprise an external power source, an onboard charger (OBC), a battery system comprising a battery and a battery management system (BMS), ancillary loads from one or more ancillary electrical devices, machine engine control unit (ECU), and one or more sensors. A controller, which may be part of the BMS or the ECU, is also provided to control the method described above. Theexternal power source 310 may comprise an AC wall socket and is connectable to the onboard charger during charging of the electric vehicle. The onboard charger may comprise a plurality of chargers. Other components may include an electric powertrain (ePT) for powering the electric motor and driving the vehicle, and sensors in communication with the onboard charger and the battery and BMS which can be used in controlling the above described method and measuring quantities required.

[0109] Although not illustrated, the present disclosure is also suitable for vehicles and equipment comprising rechargeable lead-acid batteries (such as gensets) as described above. Any examples described herein using electric vehicles can be suitably adapted for use in vehicles and equipment comprising rechargeable lead-acid batteries.

[0110] Industrial

[0111]

[0112] Measuring the State of Charge (SoC) of electric batteries is required to determine how much energy is left in the battery, for example to determine when the battery will need to be recharged. Determining that the BMS is misconfigured can be helpful in alerting an end user of faulty readings and triggering a calibration accordingly. Calibrating the sensors is important for ensuring that the determined SoC is accurate. Battery-to-battery variability can cause measurements to be inaccurate thereby giving an incorrect indication of the charge left to an end user. A method and system for detecting BMS misconfiguration and subsequently calibrating battery sensors to improve SoC calculations is disclosed herein. The misconfiguration and calibration described herein can be performed in the field, rather than in a lab under strict testing conditions.

[0113] While the preceding text sets forth a detailed description of the embodiments of the present disclosure, it should be understood that the scope of protection is defined by the words of the appended claims. The detailed description is to be construed as exemplary only and does not describe every possible embodiment since describing every possible embodiment would be impractical, ifnot impossible. Numerous alternative embodiments could be implemented which would still fall within the scope of the claims defining the scope of protection.

Claims

Claims1. A method for determining that a battery management system (BMS) of a vehicle or equipment is misconfigured, the method comprising:measuring a first voltage, OCVX1, of the battery at a first state of charge (SoC), SoCxl, of the battery and a second voltage, OCVx2, of the battery at a second SoC, SoCx2, of the battery, where SoCx1≠ SoCx2,using OCVx1at SoCx1and OCVx2at SoCx2to calculate a rate of change of voltage of the battery with respect to charge of the battery;using the calculated rate of change to compare one or more measured parameters of the BMS with one or more expected parameters of the BMS; andif said comparison indicates a mismatch between the measured and the expected one or more parameters, determining that the BMS is misconfigured and requires recalibration.

2. The method of claim 1, wherein the BMS parameters comprise one or more of: a voltage at full charge, Umax, of the battery, a voltage at full discharge, Umin, of the battery, and / or an expected rate of change based on an expected Umax, an expected Umin, and an expected capacity, C20, of the battery.

3. The method of claim 1 or claim 2, further comprising triggering a calibration of the BMS comprising:charging the battery such that the SoC is SoCmax, measuring a maximum voltage, OCVmax, of the battery at SoCmaxto determine Umax', wherein OCVmaxis measured after the battery has been left to settle for a resting period.

4. The method of claim 3, further comprising:partially discharging the battery to a first intermediate SoC of the battery SoCx1', where SoCmax> SoCx1' > SoCmin, wherein SoCminis at full discharge of the battery;measuring a first intermediate voltage, OCVx1', of the battery at SoCx1'; andextrapolating from OCVmaxat SoCmaxand OCVx1' at SoCx1', to estimate a voltage at full discharge, Umin' of the battery at SoCmin.

5. The method of claim 4, further comprising using Coulomb counting to track battery discharge; andstopping battery discharge at a predetermined value of SoCx1or a predetermined SoC change to measure OCVx1.

6. The method of claim 1, further comprising calculating an estimated voltage at full charge, Umax', of the battery comprising:extrapolating Umax' at full charge of the battery, SoCmax, based on the measured rate of change.

7. The method of any of claims 3 to 6 further comprising comparing the estimated voltage at full charge, Umax', to the expected voltage at full charge, Umax; andif the comparison indicates Umax' ≈ Umax, determining that Umaxis calibrated; orif the comparison indicates Umax' ≠ Umax, determining that Umaxis misconfigured and requires recalibration.

8. The method of any preceding claim, further comprising: calculating an estimated voltage at full discharge, Umin', comprising extrapolating based on the measured rate of change to calculate Umin' at full discharge of the battery, SoCmin,comparing the estimated voltage at full discharge, Umin', to the expected voltage at full discharge, Umin,if the comparison indicates Umin' ≈ Umin, determining that Uminis calibrated; orif the comparison indicates that Umin'Umin, determining that Uminis misconfigured and requires recalibration.

9. The method of any preceding claim, wherein recalibration is triggered if a difference between the measured parameter and the expected parameter deviates by an amount above a predetermined threshold.

10. The method of any preceding claim, further comprising measuring a temperature of the battery at the time OCVX1, OCVx2, are measured; andtemperature correcting the measurements of OCVX1, OCVx2, based on the measured temperature.

11. The method of any preceding claim, wherein SoCxl, SoCx2, vary by at least 5% SoC.

12. The method of any preceding claim, further comprising: measuring a third voltage, OCVx3, of the battery at a third SoC, SoCx3; measuring a fourth voltage, OCVx4, of the battery at a fourth SoC, SoCx4,using OCVx3at SoCx3and OCVx4at SoCx4to calculate a second rate of change of voltage with respect to SoC;comparing the second measured rate of change to the first measured rate of change; and(i) if the comparison indicates that a difference between the values of the first and second rate of change is below a predetermined threshold,increasing a confidence level for the values of the expected voltage at full charge, Umax, and the expected voltage at full discharge, Umin,(ii) if the comparison indicates that the difference is above the predetermined threshold, setting the confidence level to zero and determining that the BMS is misconfigured and requires recalibration.

13. The method of claim 12, further comprising determining that the expected voltage at full charge, Umax, and the expected voltage at full discharge, Umin, are calibrated after achieving a match between the measured values a predetermined number of times.

14. The method of claim 13, wherein the match is determined when the compared values are below a predetermined threshold.

15. A vehicle or equipment comprising:a battery;a battery management system comprising one or more sensors configured to measure a voltage of the battery and a state of charge of the battery, and a controller configured to perform the method of any one of the preceding claims 1 to 14.

Images