Method and system for estimating battery capacity in a torque-generating system

Abstract

A method for estimating battery performance in a torque-generating system (10) having a battery pack (12) with a plurality of battery cells (12C), the method comprising: calculating a voltage spread via a controller (25) as a difference between an average and a minimum cell voltage of the battery pack (12); and increasing a calibrated voltage control limit by an offset based on the size of the voltage spread when the minimum cell voltage is less than the calibrated voltage control limit; characterized by recording the offset in a memory location of the controller (25) referenced by operating conditions of the battery pack (12); estimating a performance of the battery pack (12) using the recorded offset when the battery pack (12) operates under the same conditions as the operating conditions; and performing a control action of the torque-generating system (10) by the controller (25) using the estimated performance capability.

Description

TECHNICAL FIELD

[0001] The present invention relates to a method according to the preamble of claim 1 and to a system according to the preamble of claim 7 for estimating battery performance in a torque-generating system, as essentially known from DE 10 2014 110 296 A1. BACKGROUND

[0002] High-voltage direct current battery packs can be used to power electric motors in a variety of different torque-generating systems. For example, output torque from an electric motor can be used to rotate a transmission input element in a hybrid electric or plug-in electric vehicle—that is, a vehicle with a battery pack that can be recharged using either a charging outlet or an off-board power supply. The individual battery cells of a given battery pack gradually age and discharge over time. As a result, battery performance parameters such as open-circuit voltage, cell resistance, and state of charge may change from calibrated / new values. Battery discharge is therefore monitored by a dedicated battery controller to estimate the remaining amount of battery capacity.The battery performance estimates can then be used to make decisions about powertrain mode selection or other control measures.

[0003] Existing battery control methodologies attempt to protect relatively weak battery cells in a battery pack, that is, those battery cells that exhibit the lowest cell voltages during a discharge event or the highest cell voltage during a charge event. Typically, battery protection is achieved by controlling the battery pack voltage such that the lowest cell voltages are forced to remain above a minimum allowable pack voltage. In other words, the battery pack voltage is reactively adjusted in response to a given battery cell approaching or falling below a voltage control threshold or floor. However, changing a voltage control threshold, particularly during a power pulse, can reduce the accuracy of battery performance estimates. SUMMARY

[0004] The present approach aims to improve the accuracy of existing overall battery performance estimates, and in particular those made in conjunction with reactive voltage control strategies of the type mentioned above. The present method involves recording a history of past reactive voltage control limit adjustments under different operating conditions, such as different battery pack states of charge (SOC) and temperatures, and subsequently estimating the battery pack's performance under the same operating conditions using the recorded history.

[0005] For example, the method may involve estimating maximum charge and discharge performance over various upcoming time windows, looking forward, for example, 2 seconds and 10 seconds into the immediate future. The performance estimates may be made at a maximum charge / discharge current and also at a maximum charge / minimum discharge voltage threshold. The controller then selects the current amount with the smaller relative magnitude and uses this performance estimate in a subsequent control action, e.g., in powertrain control or electric range estimation and route planning.

[0006] In one particular embodiment, a method for estimating battery performance in a torque-producing system having a battery pack with a plurality of battery cells is disclosed. The method includes calculating a voltage spread via a controller as a difference between an average cell voltage and a minimum cell voltage of the battery pack, and then increasing a calibrated voltage set limit, e.g., by an offset based on the magnitude of the voltage spread, when the minimum cell voltage is less than the calibrated voltage set limit. The method also includes recording the offset in a memory location of the controller referenced by the operating conditions of the battery pack.The method continues estimating the battery pack's performance using the detected offset when the battery pack is again operating under the same conditions as the operating conditions for the corresponding memory location. The controller then executes a control action of the torque-generating system using the estimated performance.

[0007] The operating conditions of the battery pack may include a state of charge, a temperature of the battery pack, or other suitable operating conditions.

[0008] Estimating the battery pack's performance may include estimating both a maximum charge performance and a maximum discharge performance over various future time windows. A calibrated voltage control threshold may, in some embodiments, only occur when the magnitude of the voltage spread exceeds a calibrated voltage spread threshold.

[0009] The torque-generating system in the method may include an engine, in which case the control action may include a command to the engine to turn on or off, e.g., by selecting a powertrain mode or an engine stop / start action. If the torque-generating system is a vehicle, the control action may include executing a route planning action of the vehicle.

[0010] Estimating the performance of the battery pack using the recorded offset may involve adjusting a voltage level of the battery pack by an amount proportional to the voltage spread and then calculating the performance using the adjusted voltage.

[0011] A torque-generating system is also disclosed, which includes a controller programmed to carry out the method described above. In addition to the controller, the system may include the battery pack and an electric motor operable to generate output torque when supplied with electricity from the battery pack.

[0012] The foregoing summary is not intended to represent every embodiment or aspect of the present invention. Rather, the foregoing summary merely illustrates some of the novel aspects and features set forth herein. The foregoing features and advantages, as well as other features and advantages of the present invention, will be readily apparent from the following detailed description of the illustrated embodiments and modes for carrying out the present invention, taken in conjunction with the accompanying drawings and the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a schematic diagram of an exemplary vehicle with a rechargeable battery pack and a system for estimating the performance of the battery pack. Fig. Figure 2 is a timing diagram describing voltage thresholds used as part of the method described herein, with time on the horizontal axis and voltage on the vertical axis. Fig. 3 is a flowchart illustrating an exemplary embodiment of the method disclosed herein. DETAILED DESCRIPTION

[0013] With reference to the drawings, wherein reference numerals are used to identify similar or identical components in the several views, Fig. 1, a torque-generating system 10 in the form of an exemplary vehicle is shown. The system 10 includes a rechargeable battery pack 12 and a controller (C) 25. The controller 25 is programmed with computer-executable code embodying steps of a method 100, an example of which is described below with reference to Fig. 3 with additional reference to the illustrative timing diagram in Fig. 2. As part of an overall control methodology, the controller 25 estimates a performance capability of the battery pack 12 as explained above, i.e., over several different future time windows while operating at the maximum charge / discharge current and also at the maximum charge / minimum discharge voltage limit. The controller 25 then selects the performance capability with the smaller relative magnitude and uses it in the overall control of the system 10.

[0014] The controller 25 may be programmed to model an open circuit voltage which, as is known in the art, cannot be measured during dynamic operation of the system 10, such as during travel of the vehicle from Fig. 1, if the torque-generating system 10 is so configured. That is, the terminal voltage of the battery pack 12 changes dynamically due to electrochemical effects within the battery cells 12C of the battery pack 12, with the open circuit voltage being the value the battery voltage will eventually reach when a load applied to the battery pack 12 is removed. However, if the battery voltage drops too far during dynamic operation of the system 10, undesirable electrochemical reactions may occur, and thus the controller 25 uses the aforementioned reactive voltage limit control measures to prevent such an occurrence.

[0015] The fact that conventional reactive voltage limit control measures are reactive, i.e., are performed after a cell voltage level has already dropped too far, can adversely affect the predictive accuracy of the future power estimates described above. The method 100 is therefore intended to improve the accuracy of power estimation in the face of reactive cell voltage limit control measures, and specifically by tracking the history of previous voltage limit adjustments under comparable operating conditions and subsequently offsetting them during the power estimation processes, as described below with reference to Fig. 2 and Fig. 3 will be explained in more detail.

[0016] The Fig. The exemplary torque-generating system 10 shown in FIG. 1 may include an electrified powertrain 15 from which one or more electric motors 16 draw electrical energy from the battery pack 12 and deliver motor torque to the drive wheels 17 via one or more front and / or rear drive axles. An optional internal combustion engine 14 may be used in some configurations when the system 10 is embodied as a hybrid electric vehicle or extended-range electric vehicle. Although illustrated as an exemplary passenger vehicle, the system 10 may be embodied as a mobile or static platform whose battery pack 12 may be selectively recharged via regeneration and / or by connection to an external power source (not shown) such as a 120VAC or 240VAC wall outlet or electrical charging station, or via onboard regenerative charging of the type known in the art.

[0017] In all embodiments, the battery pack 12 comprises a plurality of battery cells 12C, each having a corresponding battery cell voltage (arrow V 12C ), which can be individually measured and / or calculated by a corresponding sensor 19, for example, a voltage or current sensor, which can typically be configured as a cell measurement circuit or circuit board. In addition, the controller 25 calculates or otherwise determines an average pack voltage (arrow V AVG ), e.g., by modeling the pack voltage using a circuit model and dividing the modeled pack voltage by the number of battery cells 12C in the battery pack 12, as is well known in the art.

[0018] The controller 25 may be embodied as one or more different devices, each optionally comprising one or more microcontrollers or central processing units (P) and a memory (M), such as read-only memory, RAM, and electrically erasable programmable read-only memory. The controller 25 may be configured to run or execute various software programs, including the method 100, in the overall operation of the battery pack 12 and the torque-generating system 10. In addition, the controller 25 may output a control signal (arrow CC O) to the battery pack 12 to control it, including performing control actions such as electric range estimates, controlling a mode of the powertrain 15, and the like. The controller 25 may also use the predicted performance to schedule upcoming transitions in the powertrain 15 mode, to turn the engine 14 and / or electric motor 16 on or off, for route planning, to indicate remaining electric range, or for any other suitable control action that may benefit from the improved accuracy of the power estimate.

[0019] Central to the present method 100 is the accurate estimation of the maximum charge / discharge current of the battery pack 12. As mentioned above, the controller 25 may estimate the maximum power over any number of upcoming time windows using either a maximum charge / discharge current or a maximum charge / discharge voltage, either or both of which may be measured and / or calculated by the sensor 19, with the controller 25 selecting the smaller of the two values. The controller 25 dynamically adjusts a voltage regulation limit, i.e., a voltage floor, in the logic to protect the weakest of the battery cells 12C. To improve the accuracy of the estimates, the method 100 contemplates tracking and recording previous movements of the voltage regulation limit over time and estimating the power capability using the acquired history of the movements.

[0020] The control 25 from Fig. 1 can selectively adjust the voltage control limit based on the magnitude of a voltage spread. As used herein, the term "voltage spread" refers to the difference between a pack voltage mean (V AVG ) for the battery pack 12 and a minimum cell voltage (V MIN, C ) of the battery cells 12C or alternatively in a battery module or a small battery section of the battery pack 12. As is known in the art, a battery pack voltage is modeled using an equivalent circuit model for the battery pack 12, although the pack voltage may also be calculated or determined in other ways. The pack voltage means (V AVG ) can be calculated by dividing the modeled, calculated or otherwise determined pack voltage (V B ) can be calculated by the number of battery cells 12C in the battery pack 12, ie VAVG=VB#Cells

[0021] The voltage spread (Vs) is then calculated as follows: VS=VAVG−VMIN,C

[0022] As part of the present method 100, the controller 25 tracks Fig. 1 then considers the history of adjustments to the voltage control limit to determine the voltage spread, with such possible adjustments being made in a manner proportional to the magnitude of the voltage spread, and then takes the voltage spread into account during power estimation and other control actions of the battery pack 12. The present approach is rooted in the recognition made herein that the occurrence of a large-scale voltage spread is a relatively repeatable, condition-specific phenomenon; for example, the voltage spread tends to occur under substantially the same operating conditions, such as a recurrence of the same battery temperature and state of charge.The controller 25 therefore tracks trends in the history of voltage control limit adjustments and uses the trends in the history without modeling the battery dynamics, which can be extremely nonlinear at low states of charge. The application of the method 100 to the torque-generating system 10 of FIG. Fig. 1 will now be examined with particular reference to Fig. 2 and Fig. 3 described.

[0023] Fig. Figure 2 shows a timing diagram describing the electrical values ​​measured or controlled as part of the method 100, with voltage (V) on the vertical axis and time (t) on the horizontal axis. The average pack voltage (V AVG ) and minimum cell voltage (V MIN, C) for the battery pack 12 decrease rapidly at t1, t2, t3, t4, t5, and t6, indicating a discharge event in which a large outflow of electrical current from the battery pack 12 occurs. The average pack voltage (V AVG ) increases immediately after such discharge events, signaling a charging event in which the battery cells 12C from Fig. 1. Zones 30 show a wide voltage spread. This means that while the average pack voltage (V AVG ) and the minimum cell voltage (V MIN, C ) at other times do not match completely, under certain conditions the difference or offset can be pronounced, which, if sufficiently large in relation to a calibrated voltage spread threshold, can act as a control trigger for the reactive adjustment of a voltage control limit (V CL ) can serve as part of the method 100.

[0024] The below with reference to Fig. 3, generally involves determining an individual cell voltage for each of the battery cells 12C, e.g., by direct measurement by a corresponding sensor 19 as mentioned above, and then increasing the voltage control limit (V CL ) if the lowest cell voltage compared to the other battery cells 12C in the battery pack 12 is lower than the voltage setting limit (V CL ). The tax measure can be seen in t1, t5 and t6 of the non-restrictive representation in Fig. 2. As mentioned above, the method 100 includes recording a number of the increment in the memory (M) of the controller, for example, in a particular memory location referenced by operating conditions such as the state of charge and temperature of the battery pack 12 at the time the increment was ordered.

[0025] Controller 25 then estimates a performance capability of battery pack 12 using a voltage initially adjusted by the recorded amount of previous increases under the same operating conditions, i.e., in proactive anticipation of the voltage limit control action that will take place after the performance estimation has begun. In this way, controller 25 can use a more accurate pack voltage value to estimate future performance capability. Controller 25 finally performs a control action of system 10 using the more accurately estimated performance capability.

[0026] Between t0 and t1, the calibrated cell voltage control limit (V CL ), which is a predetermined value that is selectively adjusted as needed, is maintained at a level that is a calibrated offset from minimum pack voltage (V MIN, P), ie a minimum permissible voltage for the battery pack 12 as a whole. The cell voltage setting limit (V CL ) serves as a reference point that ultimately triggers certain control actions with respect to the battery pack 12 or the torque-generating system 10.

[0027] At t1, the minimum cell voltage (V MIN, C ), for example, below a level of the calibrated cell voltage control limit (V CL ). The cell spread described above increases significantly in the 30 ranges. In response, the controller 25 adjusts the cell voltage control limit (V CL ) at t1 by a predetermined amount and then maintains the new higher limit until the next case where the cell voltage control limit (V CL ) is violated, which in the example of Fig. 2 occurs at approximately t5. The control 25 raises the voltage control limit (V CL ) at t5 and then again at t6.

[0028] The amount of each adjustment of the voltage control limit (V CL ) may be based on the magnitude of the voltage spread existing at the time of the adjustment. The controller 25 may store the history data in data bins or memory locations corresponding to the operating conditions of the battery pack 12 as a whole when the adjustment is made. When the torque-producing system 10 again operates under operating conditions substantially the same as when the adjustments were made, and an estimate of performance under substantially the same operating conditions is required, the controller 25 performs the performance estimation using the adjusted voltage, knowing in advance from the recorded history that the voltage control limit is likely to increase as a result of the reactive voltage limit control measures.

[0029] Also shown in Fig. 2 a calibrated minimum pack voltage (V MIN, P ), which is a calibrated minimum pack voltage for the battery pack 12 as a whole. This value shows each adjustment of the calibrated cell voltage setting limit (V CL ) an upward trend. The total pack voltage of the battery pack 12 is controlled by operation of the controller 25 so that the minimum cell voltage (V MIN, C ) is not less than the minimum pack voltage (V MIN, P ). Thus, the minimum pack voltage (V MIN, P ), as in Fig. 2, automatically and selectively increased over time, in response to the periodic adjustment of the voltage control limit (V CL ), which in turn occurs when a given battery cell 12C exceeds the minimum cell voltage limit (V MIN, C ) falls below.

[0030] An exemplary embodiment of the method 100 is shown in Fig. 3. The method 100 in this embodiment may include collecting voltage spread data for different operating states over time and recording the magnitudes of such adjustments to the voltage control limit (V CL ). Different time windows such as 2s and 10s can be used for this purpose, insofar as the controller 25 in the process of controlling the system 10 from Fig. 1 ultimately makes power estimates for various future or upcoming time windows. That is, the controller 25 can predict how much battery energy will be available for allocation to the electric motor 15 in a future time interval. Using this prediction, the controller 25 can control the operation of the system 10 using more accurate power estimates, without which method 100 would be available.

[0031] After method 100 begins (**), controller 25 determines in step S102 whether controller 25 is running. Step S102 may detect an ignition or turning a key of system 10 to the on position, or may otherwise assess whether controller 25 is powered on and operational, and that conditions are appropriate for continuing method 100. Controller 25 proceeds to step S104 only if such conditions exist.

[0032] Step S104 involves determining whether the minimum cell voltage (V MIN, C ) of the voltage control limit (V CL ), ie, a cell voltage lower limit, which may be recorded in the memory (M) of the controller 25 as a calibrated value. If so, the method 100 proceeds to step S106 and prepares to increase the voltage control limit (V CL). In some embodiments, the adjustment may only occur if the voltage spread exceeds a calibrated voltage spread threshold, e.g., to avoid unnecessary adjustments. The method 100 instead proceeds to step S116 if the minimum cell voltage (V MIN, C ) does not exceed the voltage setting limit (V CL ) or below.

[0033] In step S106, the controller 25 next determines operating conditions of the battery pack 12, e.g., the state of charge and temperature of the battery pack 12, as mentioned above, by measuring or evaluating via another control module (not shown), and then identifies the corresponding memory location (M). A plurality of memory locations may be allocated, for example, in the memory (M), e.g., as one or more lookup tables. For illustrative purposes, the battery temperature may be divided into 10°C bands, e.g., from -40°C to 40°C, and the state of charge may be divided into 10% bands ranging from 0 to 100% state of charge, with various combinations of these bands being stored in the lookup tables.

[0034] Based on the current state of charge and temperature of the battery pack 12, the controller 25 can then select the appropriate memory location for the operating conditions where the voltage spread and voltage limit adjustments are considered. Step S106 can be used because, as mentioned above, the voltage spread is related to the state of charge and temperature of a battery pack 12 in a given torque-generating system 10 over time and can also be adequately related to other operating conditions. The controller 25 proceeds to step S108 once the memory locations have been identified.

[0035] In step S108, the controller 25 updates Fig. 1 the characteristic voltage spread information for the identified memory location from step S110. Step S108 entails determining the magnitude of the voltage spread or cell spread, as explained above, and recording it in the corresponding memory location. The method 100 then proceeds to step S110.

[0036] Step S110 involves determining whether all storage locations have been populated, particularly those that may be underpopulated. For example, if the torque-generating system 10 is a vehicle, an operator may infrequently discharge the battery pack 12 or continue to charge the battery pack 12 to a high state of charge and maintain the high state of charge for various reasons, including range concerns, short commuting distances, or charging habits. The method 100 proceeds to step S112 if all storage locations are not populated, and alternatively, to step S114 if all storage locations have been populated.

[0037] Step S112 involves updating the voltage spread information in the less populated memory locations and transitioning to step S114.

[0038] In step S114, the controller 25 determines the Fig.1, a battery pack voltage for use in power estimation, and it does so for a given state of charge / temperature memory location. For example, the controller 25 may extract the recorded history or characteristic voltage spread from a respective memory location according to the existing operating conditions. The method 100 then proceeds to step S116.

[0039] In step S116, the controller 25 estimates a performance of the battery pack 12 by using and recording the voltage from step S114, and then performs a control action as explained above using the estimated performance.

[0040] Therefore, by using the method 100, the controller 25 can more accurately predict when an upward movement of the cell voltage control limit (V L) will occur by assuming a repetition of the trend from the history of past control adjustments under the same operating conditions. While the offset levels for past adjustments of the voltage control limit (V CL ) under the same operating conditions, be it the state of charge and temperature-dependent or other conditions, may not always be identical to upcoming adjustments, the proactive use of the trends from the history in the reactive control of the lower voltage limit is intended to reduce errors in the power prediction and improve the control of the system 10.

Claims

[1] A method for estimating battery performance in a torque-generating system (10) having a battery pack (12) with a plurality of battery cells (12C), the method comprising: calculating a voltage spread via a controller (25) as a difference between an average and a minimum cell voltage of the battery pack (12); and increasing a calibrated voltage control limit by an offset based on the size of the voltage spread when the minimum cell voltage is less than the calibrated voltage control limit; characterized by recording the offset in a memory location of the controller (25) referenced by operating conditions of the battery pack (12); estimating a performance of the battery pack (12) using the recorded offset when the battery pack (12) operates under the same conditions as the operating conditions; and performing a control action of the torque-generating system (10) by the controller (25) using the estimated performance capability. [2] The method of claim 1, wherein the operating conditions of the battery pack (12) include a state of charge or a temperature of the battery pack (12). [3] The method of claim 2, wherein the operating conditions of the battery pack (12) include the state of charge and the temperature of the battery pack (12). [4] The method of claim 1, wherein estimating the performance of the battery pack (12) includes estimating both a maximum charging performance and a maximum discharging performance over a plurality of different future time windows. [5] The method of claim 1, wherein the torque generating system (10) includes a motor (15) and wherein the control action includes a command to the motor (15) to turn on or off. [6] The method of claim 1, wherein estimating the performance of the battery pack (12) using the recorded offset includes adjusting a voltage of the battery pack (12) by an amount proportional to a magnitude of the voltage spread, and then calculating the performance using the adjusted voltage. [7] Torque generating system (10) comprising: a battery pack (12) having a plurality of cells; an electric motor (15) operable to generate output torque when supplied with electricity from the battery pack (12); and a controller (25) in communication with the battery pack (12), the controller (25) being programmed to: to calculate a voltage spread as a difference between an average and a minimum cell voltage of the battery pack (12); increase a calibrated voltage control limit by an offset based on the magnitude of the voltage spread if the minimum cell voltage is less than the calibrated voltage control limit; characterized by , that the controller (25) is further programmed to: record the offset in a memory location of the controller (25) referenced by operating conditions of the battery pack (12); estimate a performance of the battery pack (12) using the recorded offset when the battery pack (12) operates under the same conditions as the operating conditions; and to carry out a control action of the torque-generating system (10) using the estimated performance capacity. [8] The torque generating system (10) of claim 7, wherein the operating conditions of the battery pack (12) include a state of charge and a temperature of the battery pack (12). [9] The torque generating system (10) of claim 7, wherein the controller (25) is programmed to estimate the performance of the battery pack (12) by estimating both a maximum charging performance and a maximum discharging performance over a plurality of different future time windows. [10] The torque generating system (10) of claim 7, wherein the torque generating system (10) is a vehicle (10) and wherein the controller (25) is programmed to perform a route planning action of the vehicle (10) as the control action.

Images