BATTERY CHARGE STATE ESTIMATING

Abstract

A system for estimating the state of charge of a battery assembly comprises a sensor cell and an estimating circuit. The sensor cell is coupled in series with the battery assembly. The battery assembly has an assembly-type battery chemistry, and the sensor cell has a sensor-type battery chemistry, and the assembly-type battery chemistry differs from the sensor-type battery chemistry. The estimating circuit is capable of acquiring a sequence of current sensor cell states of charge based on a sensor cell model and a sequence of sensor voltages across the sensor cell; calculating a sequence of current battery assembly states of charge based on the sequence of current sensor cell states of charge; and calculating an estimated battery assembly state of charge and an estimated sensor cell state of charge by parallel filtering the sequence of current battery assembly states of charge and the sequence of current sensor cell states of charge.

Description

INTRODUCTION

[0001] The present disclosure relates to a system and method for estimating the state of charge of a battery assembly.

[0002] Lithium iron phosphate batteries have a long service life and good thermal stability, making them suitable for use in electric vehicles and hybrid vehicles. However, lithium iron phosphate batteries also have a relatively flat open-circuit voltage versus state of charge curve. This relatively flat open-circuit voltage makes estimating the battery's state of charge difficult and often leads to an error of up to 10%.

[0003] Accordingly, experts continue their research and development work in the field of estimating the state of charge of a battery. DESCRIPTION

[0004] A system for estimating the state of charge of a battery assembly is provided herein. The system includes a sensor cell and an estimation circuit. The sensor cell is coupled in series with the battery assembly. The battery assembly has an assembly battery chemistry, the sensor cell has a sensor battery chemistry, and the assembly battery chemistry is different from the sensor battery chemistry. The estimation circuit is connected to the battery assembly and the sensor cell.The estimation circuit is operable to acquire a sequence of current sensor cell charge states of the sensor cell based on a sensor cell model of the sensor cell and a sequence of sensor voltages across the sensor cell, calculate a sequence of current battery assembly charge states of the battery assembly based on the sequence of current sensor cell charge states, and calculate an estimated battery assembly charge state of the battery assembly and an estimated sensor cell charge state of the sensor cell by filtering the sequence of current battery assembly charge states and the sequence of current sensor cell charge states in parallel.

[0005] In one or more embodiments of the system, the sensor cell model of the sensor cell includes the sequence of the current charge states of the battery assembly as an extended state variable.

[0006] In one or more embodiments of the system, the sequence of current states of charge of the battery assembly is represented by SOC BA (k+1) = (CAP / CAP SCBA )(SOC SC (k) - d%) + (Idt / CAP BA )) + ε(k). Where SOC BA (k+1) is the current state of charge of the battery assembly at time k+1, k is a plurality of measurement times, CAP SC is a capacitance of the sensor cell, CAP BA a capacity of the battery assembly, SOC SC (k) is the current state of charge of the sensor cell at time k, d% is a minimum charge offset, Idt is a sum of a current flowing through the battery assembly, and ε(k) is a noise model at time k.

[0007] In one or more embodiments of the system, an Extended Kalman Filter is used for filtering.

[0008] In one or more embodiments of the system, the extended Kalman filter is a fast extended Kalman filter for calculating the estimated sensor cell state of charge.

[0009] In one or more embodiments of the system, the estimation circuit is further operable to estimate a capacity degradation coefficient of the sensor cell using the extended Kalman filter executed at a slower rate than the fast extended Kalman filter.

[0010] In one or more embodiments of the system, the battery chemistry is a lithium iron phosphate chemistry, a lithium iron manganese phosphate chemistry, or a sodium ion chemistry.

[0011] In one or more embodiments of the system, the sensor battery is a nickel-manganese-cobalt chemistry, a nickel-cobalt-aluminum chemistry, a lithium-ion manganese chemistry, or a lithium-cobalt chemistry.

[0012] In one or more embodiments of the system, the battery unit is a battery pack or a battery module.

[0013] A method for estimating the state of charge of a battery assembly is provided herein. The method includes detecting a sequence of current sensor cell charge states of a sensor cell with an estimation circuit based on a sensor cell model of the sensor cell and a sequence of sensor voltages at the sensor cell. The sensor cell is connected in series with the battery assembly, the battery assembly has an assembly battery chemistry, the sensor cell has a sensor battery chemistry, and the assembly battery chemistry is different from the sensor battery chemistry.The method includes calculating a sequence of current battery assembly charge states of the battery assembly based on the sequence of current sensor cell charge states and calculating an estimated battery assembly charge state of the battery assembly and an estimated sensor cell charge state of the sensor cell by filtering the sequence of current battery assembly charge states and the sequence of current sensor cell charge states in parallel.

[0014] In one or more embodiments of the method, the sensor cell model of the sensor cell contains the sequence of the current charge states of the battery assembly as an extended state variable.

[0015] In one or more embodiments of the method, the sequence of the current states of charge of the battery assembly is represented by SOC BA (k+1) = (CAP / CAP SCBA )(SOC SC (k) - d%) + (Idt / CAP BA )) + ε(k). Where SOC BA(k+1) is the current state of charge of the battery assembly at time k+1, k is a plurality of measurement times, CAP SC is a capacitance of the sensor cell, CAP BA a capacity of the battery assembly, SOC SC (k) is the current state of charge of the sensor cell at time k, d% is a minimum charge offset, Idt is a sum of a current flowing through the battery assembly, and ε(k) is a noise model at time k.

[0016] In one or more embodiments of the method, an Extended Kalman Filter is used for filtering.

[0017] In one or more embodiments of the method, the extended Kalman filter is a fast extended Kalman filter for calculating the estimated sensor cell state of charge.

[0018] In one or more embodiments, the method includes estimating a capacity degradation coefficient of the sensor cell using the extended Kalman filter executed at a slower rate than the fast extended Kalman filter.

[0019] In one or more embodiments of the method, the battery chemistry is a lithium iron phosphate chemistry, a lithium iron manganese phosphate chemistry, or a sodium ion chemistry.

[0020] In one or more embodiments of the method, the chemistry of the sensor battery is a nickel-manganese-cobalt chemistry, a nickel-cobalt-aluminum chemistry, a lithium-ion manganese chemistry, or a lithium-cobalt chemistry.

[0021] In one or more embodiments of the method, the model of the sensor cell comprises a hysteresis conduction component, a plurality of delayed current components, a plurality of resistance components, and a final voltage component.

[0022] A vehicle is provided here. The vehicle includes a battery assembly, a sensor cell, and an estimation circuit. The battery assembly has a battery chemistry. The sensor cell is connected in series with the battery assembly. The sensor cell has a sensor battery chemistry, and the assembly battery chemistry differs from the sensor battery chemistry. The estimation circuit is connected to the battery assembly and the sensor cell.The estimation circuit is operable to acquire a sequence of current sensor cell charge states of the sensor cell based on a sensor cell model of the sensor cell and a sequence of sensor voltages across the sensor cell, calculate a sequence of current battery assembly charge states of the battery assembly based on the sequence of current sensor cell charge states, and calculate an estimated battery assembly charge state of the battery assembly and an estimated sensor cell charge state of the sensor cell by filtering the sequence of current battery assembly charge states and the sequence of current sensor cell charge states in parallel.

[0023] In one or more embodiments of the vehicle, the estimated state of charge of the battery assembly has an accuracy of up to 3%.

[0024] The above features and advantages, as well as other features and advantages of the present disclosure, are readily apparent from the following detailed description of the best modes for carrying out the disclosure when taken in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a schematic diagram illustrating the context of a system according to one or more example embodiments. Fig. 2 is a schematic diagram of a state of charge map in accordance with one or more example embodiments. Fig. 3 is a flowchart of a method for determining an estimated sensor cell state of charge and an estimated battery assembly state of charge in accordance with one or more example embodiments. Fig. 4 is a diagram of an estimated error for the state of charge of a battery assembly in accordance with one or more example embodiments. Fig. 5 is a flowchart of a method for determining an estimated sensor cell state of charge and a sensor cell capacity degradation coefficient according to one or more example embodiments. DETAILED DESCRIPTION

[0025] Embodiments of the disclosure provide a sensor cell connected in series with a battery assembly (e.g., a battery pack or a battery module) as a measuring unit to determine the state of charge of the battery assembly. The sensor cell and the battery pack have different chemical compositions, thus forming a mixed chemical system. By incorporating the state of charge of the battery assembly as an extended state variable in a cell model of the sensor cell, an extended Kalman filter (EKF) can be used to estimate the state of charge of both the sensor cell and the battery assembly in parallel. A closed-loop observer can minimize the capacity uncertainties in both chemical systems to increase the accuracy of the state of charge estimation for the mixed system.The extended Kalman filter can be a nonlinear version of a Kalman filter that linearizes data via an estimate of a current mean.

[0026] In Fig. 1 is a schematic floor plan illustrating the context of a system according to one or more example embodiments. The system may include a vehicle 80. The vehicle 80 generally includes an engine 90 and a system 100. The system 100 includes a rechargeable energy storage system 102 and a controller 104. The rechargeable energy storage system 102 includes one or more sensor cells 106 (one shown) and one or more battery assemblies 108 (one shown). The controller 104 includes an estimation circuit 110.

[0027] The rechargeable energy storage system 102 generates a system voltage 120 (e.g., Vt) during the charging process. The system voltage 120 is measurable by the estimation circuit 110. A sensor voltage 122 (e.g., Vs) is generated by the sensor cell 106 and measured by the estimation circuit 110. A current 124 (e.g., I) may be provided by the rechargeable energy storage system 102 in a discharge mode. The current 124 may flow into the rechargeable energy storage system 102 when it is in a charge mode. A correction signal 126 is generated by the battery assembly 108 and transmitted to the estimation circuit 110. The correction signal 126 transmits a temperature of the battery assembly 108, the system voltage 120, and the current 124 to the estimation circuit 110.

[0028] Vehicle 80 may include, but is not limited to, mobile objects such as cars, trucks, motorcycles, boats, trains, and / or airplanes. In various embodiments, vehicle 80 may be an electric vehicle. In other embodiments, vehicle 80 may be a hybrid vehicle. In some embodiments, vehicle 80 may include stationary objects such as billboards, kiosks, and / or marquees. Other types of vehicles 80 may be implemented to meet the design criteria of a particular application.

[0029] The motor 90 is an electric motor. The motor 90 generally serves to rotate the drive wheels of the vehicle 80 and generate torque. The electrical energy consumed by the motor 90 may be provided by the rechargeable energy storage system 102 and / or an alternator of the vehicle 80 under the control of the controller 104.

[0030] The system 100 includes a high-voltage storage system configured to store electrical energy. The system 100 is electrically connected to the motor 90. The system 100 is generally capable of supplying electrical energy to the motor 90 to move the vehicle 80. The system 100 may also be recharged from an electrical source (e.g., a charging station and / or an onboard alternator). In various embodiments, the system 100 may provide an electrical potential of approximately 400 to 800 volts DC (direct current). Other battery voltages may also be used to meet the design criteria of a particular application.

[0031] The rechargeable energy storage system 102 consists of a mixed-chemistry battery pack. The rechargeable energy storage system 102 is electrically connected to the controller 104. In discharge mode, the rechargeable energy storage system 102 is capable of supplying the controller 104 with electrical energy to operate the motor 90. In charge mode, the rechargeable energy storage system 102 is capable of receiving electrical energy via the controller 104 and storing the electrical energy for later use.

[0032] The control unit 104 is a battery control unit. The control unit 104 is electrically connected to the motor 90 and the rechargeable energy storage system 102. The control unit 104 is generally capable of supplying electrical energy to the rechargeable energy storage system 102 in charge mode to charge the rechargeable energy storage system 102. In discharge mode, the control unit 104 can extract electrical energy from the rechargeable energy storage system 102. The electrical power received from the rechargeable energy storage system 102 can be used to drive the motor 90 and / or other loads in the vehicle 80.

[0033] The sensor cell 106 consists of a battery cell with a sensor battery chemistry. The sensor cell 106 has a sensor cell type. In various embodiments, the sensor cell type of the battery can be a nickel-manganese-cobalt battery, a nickel-cobalt-aluminum battery, a lithium-ion manganese battery, and / or a lithium cobalt battery. The various types of sensor cell batteries have corresponding battery cell chemistries, including a nickel-manganese-cobalt chemistry 106a, a nickel-cobalt-aluminum chemistry 106b, a lithium-ion manganese chemistry 106c, and / or a lithium cobalt chemistry 106d. Other battery cell types and / or battery cell chemistries can be implemented to meet the design criteria of a particular application.

[0034] The battery assembly 108 consists of a battery module and / or a battery pack. The battery module generally includes multiple battery cells. The battery pack includes multiple battery models or multiple battery cells. The battery assembly 108 comprises a battery type. In various embodiments, the battery type generally includes lithium iron phosphate batteries, lithium iron manganese phosphate batteries, and / or sodium ion batteries. The battery assembly types may include corresponding battery assembly chemistries, including a lithium iron phosphate chemistry 108a, a lithium iron manganese phosphate chemistry 108b, and / or a sodium ion chemistry 108c. Other battery types and / or battery chemistries may be employed to meet the design criteria of a particular application.

[0035] The estimation circuit 110 includes a measurement circuit and a calculation circuit. The estimation circuit 110 is electrically connected to the battery assembly 108 and the sensor cell 106. The estimation circuit 110 is capable of detecting a sequence of current sensor cell charge states of the sensor cell 106 based on a sensor cell model of the sensor cell 106 and a sequence of sensor voltages 122 at the sensor cell 106. The estimation circuit 110 can also calculate a sequence of current battery assembly charge states of the battery assembly 108 based on the sequence of current sensor cell charge states and calculate an estimated battery assembly charge state of the battery assembly 108 and an estimated sensor cell charge state of the sensor cell 106 by filtering the sequence of current battery assembly charge states and the sequence of current sensor cell charge states in parallel.Parallel filtering generally means simultaneous filtering, overlapping filtering, unison filtering and / or near-simultaneous filtering.

[0036] The system voltage 120 may be a voltage across the rechargeable energy storage system 102. The system voltage 120 is a sum of the sensor voltage 122 of the sensor cell 106 and a system voltage of the battery assembly 108. The sensor voltage 122 is a voltage at the sensor cell 106. Both the system voltage 120 and the sensor voltage 122 are measured by the estimation circuit 110.

[0037] Fig. 2 shows a schematic diagram of an exemplary state of charge map 140 in accordance with one or more exemplary embodiments. The state of charge map 140 illustrates a sensor cell state of charge 142 of the sensor cell 106 relative to a cell state of charge 146 of a battery cell 144 having the same battery chemistry as the battery assembly 108. The sensor cell 106 has a sensor cell capacitance CAP SC . The battery cell 144 has a battery cell capacity CAP BC . The sensor cell capacity CAP SC is with the battery cell capacity CAP BC connected by equation 1 as follows: CAPSC=100%S%CAPBC

[0038] Where S% is a percentage of the sensor cell capacity CAP SC , which extends over the entire battery cell capacity CAP BC (e.g. from 0% battery cell SOC BC up to 100% battery cell SOC BC )

[0039] The sensor cell 106 has the sensor cell charge state 142 at a measurement (or observation) k, 142 = SOC SC (k). The battery cell 144 has the battery cell state of charge 146 during the measurement (or observation) k, 144 = SOC BC (k). The state of charge of the sensor cell 142 (SOC SC ) is related to the state of charge of the battery cell 146 (SOC BC ) are connected by equation 2 as follows: SOCSC(k)=CAPBCCAPSCSOCBC(k)+d%

[0040] Where d% is a percentage of the sensor cell capacity CAP SC which differs from a zero sensor cell charge state SOC SC down to a zero battery cell state of charge SOC BC extends.

[0041] The battery cell 144 can be scaled up to the battery unit 108, and the state of charge of the battery cell (SOC BC ) can indicate the state of charge of the battery unit (SOC BA) can be scaled up. In various embodiments, an extended state space equation of the battery assembly 108 is represented by Equation 3 as follows: SOCBA(k+1)=(CAP / CAPSCBA)(SOCSC(k)−d%)+(Idt / CAPBA)+ε(k)

[0042] Where SOC BA (k+1) is the current state of charge of the battery assembly at time k+1, k is a sequence of measurement times, CAP SC is a capacitance of the sensor cell, CAP BA a capacity of the battery assembly, SOC SC (k) is the current state of charge of the sensor cell at time k, d% is a minimum charge offset, Idt is a sum of a current flowing through the battery assembly 108, and ε(k) is a noise model at time k.

[0043] A sensor cell model of the sensor cell 106 can be modeled as a hysteresis transit component, a delayed current component, and a resistance component. A shorthand notation can be used for the state of charge (SOC), which is defined in Equation 4 as follows: θ=SOC

[0044] A derivative (θ̇ SC ) of the state of charge of the sensor cell (SOC SC ) can be expressed by equation 5 as follows: θ˙SC=ICAPSC

[0045] The transit hysteresis component (ζ̇) can be expressed by equation 6 as follows: ς˙=IkhysCAPSC(1−sign(I)ς)

[0046] The component of the delayed currents (u i ) can be expressed by equation 7 as follows: τiu˙i=−ui+II*, for I=1,…,n

[0047] Where τ iis the delay time constant, I is the current flowing through the sensor cell 106, and I* is a scaling factor.

[0048] The resistance component (Ṙ ij ) can be expressed by equation 8 as follows: R˙ij=∂∂θRijtab(θSC,T)ICapSC,fu¨ri∈{1,…,n,Ω}

[0049] Where T is the temperature. A resulting terminal voltage (e.g. Vs) of the sensor cell 106 (see Fig. 1) can be expressed by equation 9 as follows: Vs=OCV(θSC)+ςVgap(T)+∑i=1n∑j=1miRijfij(ui,T,θSC)+ ∑j=1mΩRΩjfΩj(I / I∗,T,θSC)

[0050] Here, OCV(θ SC ) is an open circuit voltage, ζV gap (T) is a hysteresis, ∑i=1n∑j=1miRijfij(ui,T,θSC) is a diffusion / transport factor, ∑j=1mΩRΩjfΩj(I / I∗,T,θNCM) is the ohmic / reaction kinetics and Vgap is the remaining voltage difference after relaxation of the sensor cell 106.

[0051] Since observation equation 3 is modeled, the uncertainty of the battery capacity is modeled as process noise ε(k). This process noise can be suppressed by an extended Kalman filter (EKF). The extended Kalman filter is generally used for parallel (or simultaneous, overlapping, joint, and / or near-simultaneous) estimation of the sensor cell's state of charge (SOC). SC and the state of charge of the battery assembly SOC BA used.

[0052] In a generally nonlinear form of the sensor cell model, Equation 5 for the sensor cell 106 can be written as Equation 10 as follows: θ˙SC=g(ui,ICAPSC) where u̇ i = f i (and i, I), for i = 1,..., n (hysteresis, diffusion, electrode / solid electrolyte interphase (SEI) potential, etc.). The terminal voltage from Equation 9 can be written as Equation 11: Vs=OCV(θSC)+φ(ui,θSC,I,T)+∑j=1mΩRΩjfΩj(l)

[0053] The general nonlinear model can include various battery models, equivalent circuits or electrochemical models.

[0054] With reference to Fig. 3 and back to Fig. 1, a flowchart of an example method 160 for determining an estimated sensor cell state of charge and an estimated battery assembly state of charge is shown, according to one or more example embodiments. The method 160 may be implemented in the estimation circuit. The method (or process) 160 generally includes steps 162 through 172, as illustrated. The order of the steps is shown as a representative example. Other sequences of steps may be implemented to meet the criteria of a particular application.

[0055] In step 162, the sensor cell model can be determined. The system voltage 120 (Vs), the sensor voltage 122 (Vs), and the current 124 (I) can be recorded multiple times in step 164 (e.g., k observations). In step 166, the sensor voltage 122 is used to determine the current state of charge of the sensor cell SOC. SC and the current state of charge of the battery assembly SOCBA for each observation. The extended state is calculated in step 168.

[0056] In step 170, an extended Kalman filter is applied in parallel to sets of the current sensor cell state of charge SOC SC and the current battery assembly charge level SOC BA applied to the observations to estimate the sensor cell charge level SOCSC^ and an estimated state of charge of the battery assembly SOCBA.^ The estimated sensor cell charge level SOCSC^ and the estimated state of charge of the battery assembly SOCBA^ can be stored in the estimation circuit 110 in step 172 for later use.

[0057] In Fig. 4 illustrates a graph 180 of an exemplary estimated error for the state of charge of a battery assembly, according to one or more exemplary embodiments. The graph 180 generally includes a time axis 182 and an error axis 184. The time axis 182 may be specified in units of seconds. The error axis 184 may be specified in units of error percentage. A curve 186 illustrates the estimation error in estimating the state of charge of the battery assembly over time. The curve 186 generally illustrates that the error percentage is less than 1.5% from zero error.

[0058] With reference to Fig. 5 and back to Fig.1, a flowchart of an example method 200 for determining an estimated sensor cell state of charge and a sensor cell capacity degradation coefficient is shown, according to one or more example embodiments. The method 200 may be implemented in the estimation circuit 110. The method (or process) 200 generally includes steps 162 through 168 and 202 through 208, as illustrated. The order of the steps is shown as a representative example. Other sequences of steps may be implemented to meet the criteria of a particular application.

[0059] Steps 162 to 168 may be the same steps as those described in method 160. Once the signals have been measured over several observations, a fast extended Kalman filter is applied to the observed measurements in step 202. In step 204, the estimated state of charge of the sensor cell is SOCSC^ using the fast extended Kalman filter and equations 12, 13 and 14 as follows: q(k+N)=q(k)+ε(k) θSC(k+N)=θSC(k)+q∫0N⋅dtudtCapSC0 θSC^(k)=θSC(k)

[0060] Where q is the capacity decay coefficient, and N is an elapsed time interval. In step 206, a slow extended Kalman filter is applied to estimate the capacity decay coefficient q of the sensor cell. The "slow" extended Kalman filter means that one or a few observations are used in each filter cycle (or duration). The "fast" extended Kalman filter means that the observations in the elapsed time interval are used in each filter cycle. In other words, the slow extended Kalman filter executes at a slower rate than the fast extended Kalman filter. In step 208, the estimated sensor cell state of charge and the sensor cell capacity decay coefficient are stored in the estimation circuit 110 for later use.

[0061] Various embodiments of system 100 generally use sensor cell 106 as the sensing unit to estimate the state of charge of the LFP's battery assembly. A cell model of sensor cell 106 is formulated, incorporating the state of charge of the battery assembly as an extended state variable into the plant model. System 100 applies the extended Kalman filter to estimate the state of charge of both sensor cell 106 and the battery assembly simultaneously (or in parallel). System 100 covers various types of battery cell models and applies the extended Kalman filter to estimate mixed-chemical states of charge. An extended Kalman filter with two time scales (e.g., fast and slow observers) can be used to determine both the estimated state of charge of the sensor cell and the capacity degradation coefficient of the sensor cell.The System 100 also provides battery assembly state-of-charge (SOC) accuracy within 3% with minimal software changes compared to existing battery management systems. This improved accuracy enables cost-effective, better-performing electric vehicles.

[0062] Embodiments of the disclosure generally provide a system for estimating the state of charge of a battery assembly. The system includes a sensor cell, a battery assembly, and an estimation circuit. The sensor cell is connected in series with the battery assembly. The battery assembly has an assembly battery chemistry, the sensor cell has a sensor battery chemistry, and the assembly battery chemistry is different from the sensor battery chemistry. The estimation circuit is connected to the battery assembly and the sensor cell.The estimation circuit is operable to: acquire a sequence of current sensor cell charge states based on a sensor cell model and a sequence of sensor voltages across the sensor cell; calculate a sequence of current battery assembly charge states based on the sequence of current sensor cell charge states; and calculate an estimated battery assembly charge state and an estimated sensor cell charge state by filtering the sequence of current battery assembly charge states and the sequence of current sensor cell charge states in parallel.

[0063] Numerical values ​​of parameters (e.g., of quantities or conditions) in this specification, including the appended claims, are to be understood as being modified in all cases by the term "approximately," whether or not "approximately" actually appears before the numerical value. "Approximately" means that the stated numerical value admits of slight imprecision (with some approximation to the accuracy of the value; approximately or reasonably close to the value; almost). If the imprecision given by "approximately" is not otherwise understood in the art with this ordinary meaning, then "approximately," as used herein, indicates at least variations that might arise from ordinary methods of measuring and using such parameters. Furthermore, disclosure of ranges includes disclosure of values ​​and further subdivided ranges within the overall range.Each value within a range and the endpoints of a range are hereby disclosed as a separate embodiment.

[0064] While the preferred embodiments for carrying out the disclosure have been described in detail, those familiar with the art to which this disclosure relates will recognize various alternative designs and embodiments for carrying out the disclosure within the scope of the appended claims.

Claims

[1] A system for estimating the state of charge of a battery assembly, the system comprising: a sensor cell connected in series with the battery assembly, the battery assembly having an assembly battery chemistry, the sensor cell having a sensor battery chemistry, and the assembly battery chemistry being different from the sensor battery chemistry; and an estimation circuit coupled to the battery assembly and the sensor cell, the estimation circuit being operable to: Detecting a sequence of current sensor cell charge states of the sensor cell based on a sensor cell model of the sensor cell and a sequence of sensor voltages at the sensor cell; Calculating a sequence of current battery assembly charge states of the battery assembly based on the sequence of current sensor cell charge states; and Calculating an estimated battery pack charge level of the battery pack and an estimated sensor cell charge level of the sensor cell by filtering the sequence of current battery pack charge levels and the sequence of current sensor cell charge levels in parallel. [2] The system of claim 1, wherein the sensor cell model of the sensor cell includes the sequence of current charge states of the battery assembly as an extended state variable. [3] The system of claim 2, wherein the sequence of current charge states of the battery assembly is represented by: SOCBA(k+1)=(CAP / CAPSCBA)(SOCSC(k)−d%)+(Idt / CAPBA))+ε(k) where SOC BA (k+1) is the current state of charge of the battery assembly at time k+1, k is a plurality of measurement times, CAP SC is a capacitance of the sensor cell, CAP BA a capacity of the battery assembly, SOC SC(k) is the current state of charge of the sensor cell at time k, d% is a minimum charge offset, Idt is a sum of a current flowing through the battery assembly, and ε(k) is a noise model at time k. [4] The system of claim 1, wherein the filtering uses an extended Kalman filter. [5] The system of claim 4, wherein the extended Kalman filter is a fast extended Kalman filter for calculating the estimated sensor cell state of charge. [6] The system of claim 5, wherein the estimation circuit is further operable to estimate a capacity degradation coefficient of the sensor cell using the extended Kalman filter executed at a slower rate than the fast extended Kalman filter. [7] The system of claim 1, wherein the battery chemistry is a lithium iron phosphate chemistry, a lithium iron manganese phosphate chemistry, or a sodium ion chemistry. [8] The system of claim 1, wherein the sensor battery is a nickel-manganese-cobalt chemistry, a nickel-cobalt-aluminum chemistry, a lithium-ion manganese chemistry, or a lithium-cobalt chemistry. [9] The system of claim 1, wherein the battery assembly is a battery pack or a battery module. [10] A method for estimating the state of charge of a battery assembly, comprising: Detecting a sequence of current sensor cell charge states of a sensor cell based on a sensor cell model of the sensor cell and a sequence of sensor voltages at the sensor cell with an estimation circuit, wherein the sensor cell is coupled in series with the battery assembly, the battery assembly has a battery chemistry of the assembly, the sensor cell has a sensor battery chemistry, and the battery chemistry of the assembly is different from the sensor battery chemistry; Calculating a sequence of current battery assembly charge states of the battery assembly based on the sequence of current sensor cell charge states; and calculating an estimated battery pack charge state of the battery pack and an estimated sensor cell charge state of the sensor cell by filtering the sequence of current battery pack charge states and the sequence of current sensor cell charge states in parallel.

Images