Control procedure for the detection of excessive current

Abstract

Method for controlling the charging and discharging power in a battery pack, the method comprising: - Defining a threshold for each time window of a plurality of time windows, where the threshold is smaller for longer time windows; - Measuring the current in or out of the battery pack using a current sensor; - Calculating the average current for each time window of the multitude of time windows from the measured current; - Comparing the average current with the threshold for each time window of the multitude of time windows; and - Control of charging or discharging power if the average current for one of the multiple time windows exceeds the threshold for the time window.

Description

BACKGROUND OF THE INVENTION 1. Field of the invention

[0001] The present invention relates generally to a method for controlling the power input and output of a battery pack and in particular to a method for controlling the charging and discharging power of a battery pack for an electric vehicle or a hybrid vehicle, which monitors the RMS value integrated over time and controls the power when a current threshold is exceeded. 2. Discussion of the state of the art

[0002] Electric and hybrid vehicles are rapidly gaining popularity in today's automotive markets. They offer several desirable attributes, such as reduced emissions, lower consumption of petroleum-based fuels, and potentially lower operating costs. A key component in both electric and hybrid vehicles is the battery pack. Battery packs in these vehicles typically consist of numerous interconnected cells capable of delivering a large amount of power on demand. The battery pack also represents a significant portion of the vehicle's overall cost.To maximize the range of a vehicle and the lifespan of a battery pack, the current flowing into and out of the battery pack must be regulated so that the battery pack is not charged or discharged too quickly, which could otherwise lead to overheating damage and other problems.

[0003] Several methods have been developed and implemented to monitor battery packs and prevent overheating. These methods include monitoring the temperature within the battery pack to detect overheating and monitoring the state of charge to identify excessively rapid charging or discharging. However, both of these methods have drawbacks. Economically, the temperature within a battery pack can only be monitored at various discrete points. Consequently, there may be areas within the battery pack that are hotter than indicated by the temperature sensors. In such cases, some cells in the battery pack may already be damaged before a high temperature is detected.Furthermore, a time delay occurs when a battery pack heats up, which carries the risk that power regulation only takes place after damage has already occurred. State-of-charge monitoring is typically performed for each individual cell in a battery pack. However, state-of-charge measurements can have a cumulative error, making it difficult to accurately determine whether battery pack cells are currently being damaged. Like temperature measurements, state-of-charge measurement therefore only indicates events that have already occurred.

[0004] To protect battery packs more effectively from damage, there is a need for measurement and control technology. A method that monitors the charging and discharging current over a specific time window can be used to proactively determine whether the power needs to be regulated to prevent damage to the battery pack cells, thus overcoming the limitations of existing methods. Such a method can also be implemented in a way that improves a vehicle's drivability by further extending the battery pack's lifespan.German patent application DE 11 2008 000 536 T5 describes an input / output control device for a secondary battery, comprising an estimating unit for estimating the battery current input to or output from the secondary battery, a current measuring unit for measuring the battery current to output a measured value, and a control unit for controlling the input / output energy based on the estimated value and the measured value. German patent application DE 697 28 733 T2 describes a control system for a battery installed in a vehicle, wherein the control system comprises a motor-generator coupled to the vehicle's drive shaft, a battery installed in that vehicle, and a power converter provided between that battery and the motor-generator, wherein the power converter transmits energy in both directions. SUMMARY OF THE INVENTION

[0005] In accordance with the teachings of the present invention, a method for controlling the charging and discharging of a battery pack for an electric vehicle or for a hybrid vehicle is disclosed, the method comprising: defining a threshold for each time window of a plurality of time windows, wherein the threshold for longer time windows is smaller; measuring the current in or out of the battery pack with a current sensor; calculating the average current for each time window of the plurality of time windows from the measured current; comparing the average current with the threshold for each time window of the plurality of time windows; and controlling the charging power or discharging power if the average current for any of the plurality of time windows exceeds the threshold for the time window.

[0006] Further features of the present invention will become clear from the following description and the attached patent claims in conjunction with the attached figures. BRIEF DESCRIPTION OF THE FIGURES Fig. Figure 1 is a schematic representation of a power management system for an electric vehicle battery pack; Fig. Figure 2 is a graph showing the battery pack current over time, illustrating how the moving average effective current is calculated; Fig. Figure 3 is a graph of the effective current thresholds as a function of the time duration; Fig. Figure 4 is a block diagram for a system error calculation module that can be used to calculate a “faulty” power amount that can be used in a proportional-integral (PI) control module; Fig. Figure 5 is a block diagram of a PI control module, which is located in the controller of the Fig. 1 can be implemented; and Fig. 6 is a flowchart for a process that originates from the controller in the Fig. 1 can be used to control the charging and discharging in a battery pack. DETAILED DESCRIPTION OF THE EXAMPLES OF EXECUTION

[0007] The following discussion of exemplary embodiments of the invention, which relates to a current monitoring and control method for battery packs, is purely exemplary and should not be considered limiting to the invention or its application or use. For example, the method is described in an application to battery packs of electric or hybrid vehicles, but the method is also applicable to battery packs in commercial vehicles, such as forklifts and golf carts, and also to battery packs not intended for vehicles.

[0008] Battery packs in electric and hybrid vehicles are typically large enough that thermal management must be considered. Unlike the small batteries used in computers and electronic devices, battery packs in electric and hybrid vehicles consist of many cells packed closely together, meaning no single cell has enough free space to allow for effective air cooling. The thermal management problem is exacerbated by the high currents required to power an electric or hybrid vehicle, which generate heat during charging and discharging. High operating temperatures are known to be detrimental to battery packs, such as the lithium-ion battery packs used in many electric and hybrid vehicles.

[0009] It is therefore essential that the cells in a battery pack are protected from overheating during charging or discharging. Electric vehicle battery packs typically have an integrated cooling system to dissipate heat, which is necessary to keep the temperature below a certain threshold. Battery pack control systems have also been developed to limit the charging or discharging current based on either the temperature or the state of charge. However, these existing control methods and systems cannot always protect the battery pack from damage, as they are designed to respond to temperature or state-of-charge measurements that may contain errors and / or time delays.To overcome these problems, a new type of battery pack control method is shown below, which uses the temporal history of the charging or discharging current as the basis for power regulation.

[0010] The Fig. Figure 1 is a schematic diagram of a power management system 10 for the battery pack of an electric vehicle. A vehicle 12 uses one or more electric motors (not shown) for propulsion, with electrical energy for the motors being supplied by a battery pack 14. The battery pack 14 is equipped with a variety of voltage and temperature sensors 16, which monitor the voltage as an indicator of the state of charge and the temperature within the battery pack. Other parameters can also be monitored by the voltage and temperature sensors 16, and the measurements can be taken at a variety of points distributed across the battery pack 14. A power cable 18 carries the total charging and discharging current to and from the battery pack 14. The charging current comes from a charging transformer (not shown), while the discharging current goes to the electric motors. A current sensor 20 continuously measures the charging and / or discharging current.

[0011] A controller 22 is used to monitor and regulate the current into and out of the battery pack 14 using measurements from the current sensor 20. The controller 22 is connected to the voltage and temperature sensors 16 via connections 24 and to the current sensor 20 via connection 26. Connections 24 and 26 can be wired or wireless. The controller 22 is configured to limit the charging or discharging power, if necessary, to prevent overheating damage to the battery pack 14, as discussed in detail below.

[0012] The present invention proposes monitoring the current in (during charging) or out (during discharging) the battery pack 14, using the time-integrated effective (RMS) current as the basis for a control algorithm. The algorithm records the effective current over various time intervals and compares it to threshold values ​​for these intervals. If the effective current exceeds a threshold value, a proportional-integral (PI) control algorithm implemented in the controller 22 regulates the power. The PI control algorithm in the controller 22 continuously regulates the power until the effective current falls below the threshold value.

[0013] The Fig. Figure 2 is a graph 30 that illustrates how the moving average RMS current is calculated for different time windows. In graph 30, the horizontal axis 40 represents time, and the vertical axis 50 represents the square of the current flowing into or out of the battery pack 14. The curve 60 is a graph of the squared current as a function of time, recorded under typical operating conditions of the battery pack 14 in the vehicle 12. To avoid excessive charging or discharging current into the battery pack 14, it is necessary to monitor the effective current for different time windows, ranging from the present to the past. For example, line 42 represents the present, line 44 represents a time 30 seconds before line 42, and line 46 represents a time 60 seconds before line 42.Line 62 represents the mean squared current for the last 30 seconds, and line 64 represents the mean squared current for the last 60 seconds. As can be seen from graph 30, line 62, which represents the 30-second RMS current and the 60-second RMS current, shows... I302 This can be described as a current-squared value, which is shown with the line marker 52. Similarly, line 64, which is the 60-second moving average current and is described as I602 The current squared value, denoted by the dash mark 54, is displayed. The effective current for any time window can be determined by taking the square root of the mean squared current for that time window. The moving mean effective current can be calculated for any time window deemed relevant, where the time windows can include the last 30 seconds, the last 60 seconds, the last 2 minutes, and the last 5 minutes.

[0014] Using the method described above for calculating the moving average effective current for different time windows, it is necessary to define one or more threshold values ​​that are used by the controller 22 to determine whether power limitation needs to be applied at any given time. Although a single effective current threshold would be the simplest to implement, both vehicle performance and battery pack lifetime can be better optimized by using a multi-threshold method. Fig. Figure 3 is a graph 70 that illustrates how several effective current thresholds can be defined. Time is plotted on the horizontal axis 80 of graph 70, while effective current is plotted on the vertical axis 90. Threshold lines 102, 104, 106, and 108 represent effective current thresholds that can be used in conjunction with the moving average effective currents, as defined above and shown in graph 30. In the example shown in graph 70, the line marker 82 represents a time of 30 seconds. Therefore, threshold line 102 represents the threshold for a moving average effective current. I302 as defined above, where the effective current threshold is represented by line marker 92. Line marker 84 represents a time of 60 seconds, so that threshold line 104 represents the threshold of an effective current averaged over 60 seconds. I602 represents an effective current threshold that is designated by the line marking 94.

[0015] The purpose of the multiple threshold model is to ensure that battery pack 14 can handle high charging or discharging power for short periods without sustaining damage, but the power must be reduced somewhat if the process is carried out for an extended period. If only a single threshold were used, defined by threshold line 102, battery pack 14 would be damaged if the threshold current were exceeded for several minutes. Furthermore, using only a single threshold, defined by threshold line 108, could unnecessarily impair the vehicle's short-term performance by preventing the use of higher currents for short periods.The multiple threshold model allows for both optimization of battery pack durability and optimization of vehicle performance.

[0016] Threshold lines 106 and 108 are defined similarly. Each is associated with a longer, moving average time window and has a progressively lower threshold current value. In the example of graph 70, threshold line 106 extends to marker 86 at a time of 2 minutes and has a threshold current value labeled with marker 96. Threshold line 108 is a continuous operating threshold without a time limit and has a threshold current value labeled with marker 98. More or fewer thresholds than the four shown in graph 70 can be defined.

[0017] The above discussed Fig. 2 and Fig. Figure 3 illustrates the concept of the moving average current and the associated current thresholds. A detailed discussion of moving average currents and the control algorithms used to limit power when necessary follows.

[0018] To obtain the mean square current for a certain time window, the following theoretical formula can be used: I2=1t2−t1∫t1t2(I(t)2)dt where I is the current, t is the time, and t1 and t2 are the start and end times of the time window.

[0019] The effective current can be calculated from equation (1): Irms=∫t1t2(I(t))2dtt2−t1

[0020] To implement the control algorithm in an electronic processor, such as the controller 22, a discrete method for calculating the floating effective currents is required. The mean square of the current at a given time step can be calculated as follows: I¯k2=1n∑i=k−n+1kIi2 where k is the time step size for the current, n is the number of measurement points, i is the increment, and I is the current.

[0021] The moving effective current over n measurement points can be calculated by using an exponentially weighted moving average filter calibrated with weights defined by the integral, as follows: I¯k2=wI¯k−12+(1−w)Ik2 where w=nn+1 It is implied that (1−w)=(1n+1).

[0022] The Fig. Figure 4 is a block diagram for a system error calculation module 120, which can be used to calculate an "error" power amount used in a PI control module. Squared current data from current sensor 20 is taken on line 130. The squared current signal on line 130 is of the type shown in curve 60 on graph 30. In box 132, the exponentially weighted moving average filter, as calculated in equations (3) and (4), is applied to calculate a moving average RMS current for any desired time window. The moving average RMS current is used to construct function 134, multiplying it by the time from box 136 to obtain an I̅. 2 to produce a t-value. The I̅ 2The t-value is passed to summing point 138, where the appropriate threshold for the time window is subtracted from box 140. This results in a function u in box 142, representing the "error" or excess in the threshold, expressed as the squared current time. The function u is passed to the product function 144, where it is divided by the time from box 146, resulting in an I̅. 2The error value is calculated as follows. In box 148, the square root is taken, resulting in an I̅ error value. In the product function 150, the I̅ error value is multiplied by the output voltage of battery pack 14 from box 152, resulting in a system "error" as a function of time in box 154, expressed as power in watts. The error value in box 154 can be converted to kilowatts by dividing it by 1000. A sign convention is introduced such that the system error function e(t) is negative if a threshold is exceeded during battery pack charging and e(t) is positive if a threshold is exceeded during discharge, i.e., while driving.

[0023] The Fig. Figure 5 is a block diagram of a proportional-integral (PI) control module 160, as it would be implemented in the controller 22. The error function e(t) from module 120 is applied to the input of line 162, which branches at junction 164 into a proportional block 166 and an integral block 168. The proportional block 166 calculates a control kernel that is proportional to an error function; that is, the proportional term in block 166 is equal to a gain constant K. p multiplied by the error function e(t). Integral block 168 calculates a control kernel that is a function of the integral of the error signal; that is, the integral term in box 168 is equal to an error constant K. i multiplied by the time integral of the error function e(t). The outputs of the proportional block 166 and the integral block 168 are added in a summing unit 170 to generate a PI control signal.

[0024] In Box 172, a limiting function is applied to the integral term from Box 168. The limiting function in Box 172 serves to restrict the integral term so that it falls between a maximum and a minimum value, where the maximum and minimum values ​​depend on whether the battery pack 14 is being charged or discharged. During charging, the maximum value for the integral term is equal to the negative proportional term, whereas the minimum value is equal to the negative value of a charging power limit P. chlim minus the proportional term. During discharge operation, the maximum value of the integral term is equal to the negative of the discharge power limit P. dchlimminus the proportional term, whereas the minimum value is equal to the negative of the proportional term. The proportional term from box 166 is not affected by the limiter function in box 172. The result of the limiter function in box 172 goes to the summing unit 174, where it either leads to the charging power limit P. chlim or to the discharge power limit P dchlimfrom box 176 is added. The result of the PI control module 160 is the power setting for the current iteration of the procedure on line 178. The power setting on line 178 is either equal to the charging power limit or to the discharging power limit plus the sum of the proportional term from box 166 and the integral term from box 168, whereby the integral term may have been limited in box 172. The sign convention was chosen such that the proportional and integral control terms calculated from the "error" signal have a reducing effect on the power setting; that is, during charging, the power has a positive value, the proportional control kernel and the integral control term are negative, thus reducing the power setting. The opposite applies during discharging.The property of the limiter function is that the output of the power setting has a value somewhere between zero and a suitable power limit (P). chlim or P dchlim ) has.

[0025] Fig. Figure 6 is a flowchart 200 of the entire process used by the controller 122 to monitor the current and regulate the charging and discharging power in the battery pack 14. Flowchart 200 describes the procedures used in the system error calculation module 120 and the PI control module 160. The procedure begins in the start oval 202. In box 204, the moving average current is calculated over box 132 of module 120. In box 206, the mean squared current is multiplied by time to determine the I 2 The t-value is obtained via the product function 134. In Box 208, the error function u is obtained by subtracting the I. 2The t-value is obtained from a suitable threshold via the summing unit 138. Simultaneously, the sign of the error function u is also determined in box 208. The sign convention is chosen such that the charging power is positive and the system error is negative during charging, whereas the discharging power is negative and the system error is positive during discharging. In box 210, the system error function e(t) is calculated by dividing the error function u by time, taking the square root, multiplying by the system voltage, and adopting the appropriate positive or negative sign from box 208. The equations from box 210 represent the product function 140, box 142, the product function 144, and box 146 from module 120.

[0026] In decision diamond 212, the procedure branches depending on the sign of the system error function e(t). If the error function is negative, the procedure proceeds to the loading branch in box 214. In box 214, the proportional term of the PI control module 160 is calculated over box 166. In box 216, the integral term of the PI control module 160 is calculated over box 168. Decision diamonds 218 and 222 serve to limit the integral term so that it remains between certain extrema, as discussed above. In decision diamond 218, it is checked whether the integral term from box 216 is greater than the value of the proportional term from box 214. If so, the integral term is set equal to the proportional term. If not, the integral term is left unchanged, and the procedure continues to decision diamond 222.In decision diamond 222, it is checked whether the integral term from box 216 is smaller than a minimum value for charging operation, where the minimum value is the constant charging power limit P. chlim The value is defined as minus the proportional term from box 214. If this is the case, the integral term is set equal to the minimum value for charging operation. If this is not the case, the integral term is left unchanged, and the procedure continues to box 226. In box 226, the complete PI control signal for charging is calculated by adding the proportional term from box 214 to the integral term from box 216, which may have been modified in box 220 or box 224. In box 228, the value for the charging power is calculated by adding the complete PI control signal from box 126 to the charging power limit P. chlim calculated. The process branch for loading ends in end oval 230.

[0027] In decision diamond 212, if the error function is positive, the procedure proceeds to the discharge branch in box 234. In box 234, the proportional term of the PI control module 160 is calculated via box 166. In box 236, the integral term of the PI control module 160 is calculated via box 168. Decision diamonds 238 and 242 serve to limit the integral term so that it is kept between certain extrema. In decision diamond 238, the integral term from box 236 is checked to see if it is smaller than a maximum value for discharge operation, where the maximum value is the constant (negative) discharge power limit P. dchlimThe integral term is subtracted from the proportional term in box 234. If this is the case, the integral term is set equal to the maximum value for the discharge operation. If this is not the case, the integral term is left unchanged, and the process proceeds to decision diamond 242. In decision diamond 242, the integral term from box 236 is checked to see if it is larger than the negative proportional term from box 234. If this is the case, the integral term is set equal to the negative proportional term. If this is not the case, the integral term is left unchanged, and the procedure proceeds to box 246. In box 246, the entire PI control signal for discharge is calculated by adding the proportional term from box 234 to the integral term, where the integral term comes from box 236 and may have been modified via boxes 240 or 244.In box 248, the value for the discharge power is obtained by subtracting the entire PI control signal from box 246 from the discharge power limit P. dchlim calculated. The discharge branch of the process also ends in the end oval 230.

[0028] The overall effect of the procedure from flowchart 200 is that, if a threshold is exceeded first, the controller 22 begins to reduce the charging or discharging power at a rate based on the gain constants K. p and K iIn the PI control module 160, if the moving average effective current remains above the threshold, the power is further reduced. In some cases, the charging or discharging power is set to zero. With little or no current flow, the effective current is expected to fall below the threshold after a certain period, and the power will be increased. The power can then be increased again before the effective current falls below the threshold, as is naturally determined by the PI control module 160, and can be adjusted by selecting the gain constant K. p and K i and the performance limits P chlim and P dchlim be custom-made.

[0029] Once the controller 22 limits the power because a threshold has been exceeded, the specific power limitation can be implemented in a variety of ways, such as adjusting the voltage ratio between the motor and the charging transformer and the battery pack 14, reducing the current by adding electrical resistance to the power cable 18, or using other methods. The spatial placement and physical implementation of a given power limiter can be determined based on cost, reliability, packaging, or other factors. In any case, the control methodology disclosed above, which monitors the current over various time windows, ensures reliable control of the battery pack's charging and discharging power, thus protecting the battery pack 14 from damage while simultaneously providing optimal performance for the driver.

Claims

[1] Method for controlling the charging and discharging power in a battery pack, the method comprising: - Defining a threshold for each time window of a plurality of time windows, where the threshold is smaller for longer time windows; - Measuring the current in or out of the battery pack using a current sensor; - Calculating the average current for each time window of the multitude of time windows from the measured current; - Comparing the average current with the threshold for each time window of the multitude of time windows; and - Control of charging or discharging power if the average current for one of the multiple time windows exceeds the threshold for the time window. [2] Method according to claim 1, wherein the calculation of the mean current for a plurality of time windows includes weighting with an exponentially weighted moving average filter. [3] Method according to claim 1, wherein the control of the charging power or discharging power includes reducing the charging power or discharging power if the average current for one of the plurality of time windows is above a threshold for the time window. [4] Method according to claim 1, wherein the control of the charging power or discharging power includes calculating an error function and using a proportional-integral control algorithm to calculate a power setting. [5] Method according to claim 4, wherein the use of a proportional-integral control algorithm for calculating a power setting includes: - Calculating a proportional term based on the error function; - Calculating an integral term based on the error function; - Limiting the integral term, if necessary, to keep it between a maximum and a minimum value; - adding the proportional term and the integral term one after the other; and - Adding a power limit value to the sum of the proportional term and the integral term to obtain the power setting. [6] Method according to claim 5, wherein limiting the integral term includes defining the maximum and minimum values ​​for the charging power differently from the maximum and minimum values ​​for the discharging power. [7] Method according to claim 5, wherein the addition of a power limit includes the power limit for the charging power being defined differently from the power limit for the discharging power. [8] Method according to claim 1, wherein the battery pack is a lithium-ion battery pack. [9] Method according to claim 1, wherein the battery pack is used in an electric vehicle or a hybrid vehicle. [10] A power management system for a battery pack for supplying electrical energy to a device, the power management system comprising: - a current sensor for measuring current flowing into or out of the battery pack; - a controller for adjusting the charging and discharging power in the battery pack, wherein the controller responds to the measurement of signals from the current sensor and is configured to limit the power to prevent damage to the battery pack if the current averaged over each time window of a plurality of time windows exceeds a threshold for one of the plurality of time windows, wherein the threshold is smaller for longer time windows; and - Means to limit the charging and discharging power in the battery pack based on signals from the controller.

Images