Monitoring charge stored in a battery

• US 10,295,600 B2
• Filed: 06/18/2014
• Issued: 05/21/2019
• Est. Priority Date: 06/18/2013
• Status: Active Grant

• Current Assignee: Vlaamse Instelling Voor Technologisch Onderzoek Nv (Vito Nv)
• Original Assignee: Vlaamse Instelling Voor Technologisch Onderzoek Nv (Vito Nv)
• Inventors: Coenen, Peter
• Primary Examiner(s): Charioui, Mohamed
• Assistant Examiner(s): Hagos, Eyob
• Application Number:
  US14/897,962
• Publication Number:
  US 20160139210A1
• Time in Patent Office: 1,798 Days
• CPC Class Codes: B60L 2240/545 Temperature
• CPC Class Codes: B60L 2240/547 Voltage
• CPC Class Codes: B60L 2240/549 Current
• CPC Class Codes: B60L 2240/80 Time limits
• CPC Class Codes: B60L 2260/44 by parameter estimation
• CPC Class Codes: B60L 3/0046 relating to electric energy storage systems, e.g. batteries or capacitors
• CPC Class Codes: B60L 58/13 Maintaining the SoC within a determined range
• CPC Class Codes: B60L 58/16 responding to battery ageing, e.g. to the number of charging cycles or the state of health [SoH]
• CPC Class Codes: G01R 31/3648 comprising digital calculation means, e.g. for performing an algorithm
• CPC Class Codes: G01R 31/367 Software therefor, e.g. for battery testing using modelling or look-up tables
• CPC Class Codes: G01R 31/3842 combining voltage and current measurements
• CPC Class Codes: Y02T 10/70 Energy storage systems for electromobility, e.g. batteries

Abstract

A method and corresponding device, for monitoring a charge of a battery. The method obtains a battery current measurement value and a battery voltage measurement value, and applies a current integration method to update a primary charge estimate value representative of the charge stored in the battery by taking into account the battery current measurement value. The method further determines an ancillary charge estimate value representative of the charge stored in the battery using a battery model taking into account the battery voltage measurement value, and determines an error value for the ancillary charge estimate value, in which the error value expresses the reliability of the battery model. The method also applies a correction to the primary charge estimate value as function of the ancillary charge estimate value and the error value. Interpretation of the correction applied in this manner allows determination of current sensor offset and battery capacity change.

Claims

1. A method for determining whether to change a battery in an electric or hybrid vehicle or in electric power back-up units, the method comprising:

obtaining a battery current measurement value (k) using a current sensor having a predetermined current offset and recorded at regular time intervals and obtaining a battery voltage measurement value using a voltage sensor and recorded at regular time intervals, said battery current measurement value (k) being a battery current (IL) entering or leaving a battery having battery poles including said current offset and said battery voltage (VL) being a battery voltage (VL) across the battery poles,

using a processor to apply a current integration method to update a primary charge estimate value representative of the charge stored in the battery by integrating the battery current measurement value including the current offset over time, the current integration method being implemented as a prediction step of a one-dimensional linear quadratic estimation method for recursively updating an estimated quantity which reflects an underlying system state corresponding to the charge stored in the battery, taking into account a sequence of noise input data, based on a previously determined primary charge estimate value as an a-posteriori state estimate of a previous time instance, where the primary charge estimate value (Qn) is considered as an a-priori state estimate of a scalar hidden state variable representing the true, yet unknown, electrical charge stored in the battery,

determining an ancillary charge estimate value representative of the charge stored in the battery using a battery model taking into account the battery voltage measurement value (VL),

wherein the determining of the ancillary charge estimate value (zk) representative of a relative charge state of the battery being a state-of-charge (SOC) value expressed as a fraction of a total charge capacity of the battery, the determining of the ancillary charge estimate value (zk) being performed by using a battery model with the battery voltage measurement value (VL) as a parameter, being an electromotive force (EMF) of the battery from a predetermined calibration curve of the EMF and the SOC,

determining an error value for said ancillary charge estimate value, said error value expressing the reliability accuracy of said battery model taking the battery voltage measurement value (VL) into account,

applying a correction to the primary charge estimate value as a function of the ancillary charge estimate value and the error value by applying an update step of the linear quadratic estimation method to the prediction step of the linear quadratic estimation method for updating the primary charge estimate value (Qk), the correction of the linear quadratic estimation method being determined by applying the ancillary charge estimate value (zk) of the battery as an observation variable,

determining a systematic error of the battery current measurement value (k) taking into account said correction to the primary charge estimate value,

determining an estimated total charge capacity of the battery taking into account said correction to the primary charge estimate value,

monitoring the charge of the battery taking into account the systematic error, wherein the systematic error is fed into a feedback loop to compensate for the integration of the current offset and to update the estimated total capacity of the battery,

determining a battery age or state of health of the battery based on a change of the total capacity of the battery over time, and

replacing the battery based on a poor battery age or state of health of the battery.

2. The method according to claim 1, in which said linear quadratic estimation filter method is selected form a one-dimensional linear quadratic estimation filter method, a Kalman filter, a Sequential Kalman filter, an Information filter, a Cholesky matrix square root algorithm, a Potter'"'"'s square root measurement-update algorithm, the Household algorithm, the Gram-Schmidt algorithm, the U-D measurement update, the U-D time update, or the particle filter.

3. The method according to claim 1, in which applying the current integration method comprises performing a prediction step of a Kalman filter to update said primary charge estimate value, wherein said primary charge estimate value is a Kalman prediction value of a hidden state variable corresponding to the charge stored in the battery, and said Kalman filter further comprises a control variable commensurate with said battery current measurement value (IL).

4. The method according to claim 3, in which said Kalman filter is a one-dimensional Kalman filter.

5. The method according to claim 1, in which said determining the ancillary charge estimate value takes the battery current measurement value (IL) and/or a battery temperature measurement into account.

6. The method according to claim 1, further comprising the step of when each time the battery becomes full or empty, updating the estimated total charge capacity taking into a direction of the correction of the primary charge estimate value.

7. The method according to claim 1, wherein the battery model is an electrical model with a series resistor and a resistor/current parallel network.

8. The method according to claim 1, wherein the error value for said ancillary charge estimate value is assigned a high value when the SOC varies over a wide range for small variations in the EMF and assigned a low value when the battery is near a full or near empty state.

9. A device for monitoring the charge of a battery in an electric or hybrid vehicle or in electric power back-up units, comprising:

a current sensor for providing a battery current measurement value (IL) entering or leaving a battery having poles that is recorded at regular time intervals, the current sensor having a predetermined current offset,

a voltage sensor for providing a battery voltage measurement value (VL) across the battery poles that is recorded at regular time intervals, and

a processing unit comprising;

means for applying a current integration method to update a primary charge estimate value representative of the charge stored in the battery by integrating the battery current measurement value over time including integrating the current offset, the current integration method being implemented as a prediction step of a one-dimensional linear quadratic estimation method for recursively updating an estimated quantity which reflects an underlying system state corresponding to the charge stored in the battery, taking into account a sequence of noise input data, based on a previously determined primary charge estimate value as an a-posteriori state estimate of a previous time instance, where the primary charge estimate value (Qn) is considered as an a-priori state estimate of a scalar hidden state variable representing the true, yet unknown, electrical charge stored in the battery,

means for determining an ancillary charge estimate value representative of the charge stored in the battery using a battery model taking into account the battery voltage measurement value (VL), wherein the means for determining the ancillary charge estimate value (zk) representative of a relative charge state of the battery being a state-of-charge (SOC) value expressed as a fraction of the total charge capacity of the battery, the determining of the ancillary charge estimate value (zk) being performed by using a battery model with the battery voltage measurement value (VL) as a parameter, being an electromotive force (EMF) of the battery from a predetermined calibration curve of the EMF and the SOC,

means for determining an error value for said ancillary charge estimate value, said error value expressing the reliability accuracy of said battery model taking the battery voltage measurement value (VL) into account,

means for applying a correction to the primary charge estimate value as function of the ancillary charge estimate value and the error value by applying an update step of the linear quadratic estimation method to the prediction step of the linear quadratic estimation method for updating the primary charge estimate value (Qk), the correction of the linear quadratic estimation method being determined by applying the ancillary charge estimate value (zk) of the battery as an observation variable,

means for determining a systematic error of the battery current measurement value taking into account said correction to the primary charge estimate value and for determining an estimated total charge capacity of the battery taking into account said correction to the primary charge estimate value,

wherein said processing unit has means for determining whether to change the battery by monitoring the charge of the battery taking into account the systematic error wherein the systematic error is fed into a feedback loop to compensate for the integration of the current offset and to update the estimated total capacity of the battery, and a means for providing a fuel gauge based on the estimated total charge capacity of the battery and the state of charge of the battery, wherein a battery age or state of health of the battery is determined based on a change of the total capacity of the battery over time, and replacing the battery based on a poor battery age or state of health of the battery.

10. The device according to claim 9, wherein a Kalman filter is adapted to update said primary charge estimate value, wherein said primary charge estimate value is a Kalman prediction value of a hidden state variable corresponding to the charge stored in the battery, and said Kalman filter further comprises a control variable commensurate with said battery current measurement value.

11. The device of claim 9, the filter method is selected from a Kalman filter, a Sequential Kalman filter, an Information filter, a Cholesky matrix square root algorithm, a Potter'"'"'s square root measurement-update algorithm, the Household algorithm, the Gram-Schmidt algorithm, the U-D measurement update, the U-D time update, or the particle filter.

12. A device for monitoring the charge of a battery in an electric or hybrid vehicle or in electric power back-up units, comprising:

a current sensor for providing a battery current measurement value (IL) entering or leaving a battery having poles that is recorded at regular time intervals, the current sensor having a predetermined current offset,

a voltage sensor for providing a battery voltage measurement value (VL) across the battery poles that is recorded at regular time intervals, and

a processing unit comprising;

wherein said processing unit is configured to apply a current integration method to update a primary charge estimate value representative of the charge stored in the battery by integrating the battery current measurement value over time including integrating the current offset, the current integration method being implemented as a prediction step of a one-dimensional linear quadratic estimation method for recursively updating an estimated quantity which reflects an underlying system state corresponding to the charge stored in the battery, taking into account a sequence of noise input data, based on a previously determined primary charge estimate value as an a-posteriori state estimate of a previous time instance, where the primary charge estimate value (Qn) is considered as an a-priori state estimate of a scalar hidden state variable representing the true, yet unknown, electrical charge stored in the battery,

wherein said processing unit is configured to determine an ancillary charge estimate value representative of the charge stored in the battery using a battery model taking into account the battery voltage measurement value (VL),

wherein said processing unit is configured to determine an ancillary charge estimate value (zk) representative of a relative charge state of the battery being a state-of-charge (SOC) value expressed as a fraction of the total charge capacity of the battery, and wherein the determining of the ancillary charge estimate value (zk) is performed by using a battery model with the battery voltage measurement value (VL) as a parameter, being an electromotive force (EMF) of the battery from a predetermined calibration curve of the EMF and the SOC,

wherein said processing unit is configured to determine an error value for said ancillary charge estimate value, said error value expressing the reliability accuracy of said battery model taking the battery voltage measurement value (VL) into account,

wherein said processing unit is configured to apply a correction to the primary charge estimate value as function of the ancillary charge estimate value and the error value by applying an update step of the linear quadratic estimation method to the prediction step of the linear quadratic estimation method for updating the primary charge estimate value (Qk), the correction of the linear quadratic estimation method being determined by applying the ancillary charge estimate value (zk) of the battery as an observation variable,

wherein said processing unit is configured to determine a systematic error of the battery current measurement value taking into account said correction to the primary charge estimate value and for determining an estimated total charge capacity of the battery taking into account said correction to the primary charge estimate value,

wherein said processing unit is configured to determine when to change the battery by monitoring the charge of the battery taking into account the systematic error, wherein the systematic error is fed into a feedback loop to compensate for the integration of the current offset and to update the estimated total capacity of the battery, and to determine a battery age or state of health of the battery based on a change of the total capacity of the battery over time, and replace said battery based on a poor battery age or state of health of the battery.