A cloud platform-based cell consistency calculation and prediction method

Abstract

The application discloses a kind of consistency calculation and prediction method of battery based on cloud platform, comprising: judging the charging state of vehicle according to cloud data;When the battery voltage meets the condition and the duration meets t, record the battery voltage, and record the current charging capacity;According to current charging capacity, judge and calculate;Calculate the minimum locking capacity of battery and the capacity of each battery smaller than the highest capacity battery;Record the capacity of each battery smaller than the highest capacity battery and calculate the Kth predicted capacity difference value, calculate the battery capacity difference in the future period of time;Judge the predicted battery capacity difference, if it meets the condition, trigger alarm and push early warning information to service personnel.The application predicts the capacity difference value in the future period of time through vehicle historical data, if the predicted value is greater than the threshold value, cloud platform sends early warning information to vehicle service personnel, and maintains vehicle in time.

Description

Technical Field

[0001] This invention relates to the field of lithium iron phosphate battery cell technology, and in particular to a cell consistency calculation and prediction method based on a cloud platform. Background Technology

[0002] With the increasing maturity and sophistication of 5G communication technology, 5G communication and cloud computing are being increasingly applied in new energy vehicles. 5G and cloud computing offer advantages such as high-speed communication and massive computing power, solving the problem of certain complex algorithms being unable to run on-vehicle controllers. Lithium iron phosphate (LFP) batteries, due to their high safety, are gaining increasing market share in the new energy vehicle market, leading to increasingly stringent requirements for LFP battery management systems. The voltage plateau region of LFP batteries presents a challenge for cell consistency calculations.

[0003] Currently, the main methods for calculating cell consistency are the OCV-SOC (Open Circuit Voltage, OCV; State of Charge, SOC) lookup table method and the real-time single-cell capacity calculation method. The OCV-SOC lookup table method obtains the SOC of each cell string by looking up the OCV-SOC table using the terminal voltage after resting, and calculates the SOC difference between cells in each string to determine cell consistency. Because lithium iron phosphate batteries have a plateau region, such as... Figure 2 As shown, in the plateau region, the SOC cannot be accurately obtained by looking up the OCV-SOC table using the terminal voltage, making this method unsuitable for lithium iron phosphate cells. The real-time single-cell capacity calculation method involves the on-board controller calculating the remaining capacity of each cell string in real time, and then determining the inter-cell differences based on the remaining capacity differences between each string. This method is applicable to different types of batteries, but calculating the remaining capacity of each cell string for hundreds or more battery packs requires significant memory resources, placing memory requirements on the controller selection and increasing the hardware cost of the on-board controller.

[0004] For example, a method and apparatus for predicting cell consistency in a lithium-ion power battery system, disclosed in Chinese patent literature (publication number CN113589174A), includes: obtaining the charge-discharge curves of the highest-capacity and lowest-capacity single cells in the lithium-ion power battery system under test; and calculating the voltage difference at the end of charging and / or discharging of the highest-capacity and lowest-capacity single cells during the test period based on preset operating conditions and the charge-discharge curves of the highest-capacity and lowest-capacity single cells, thereby determining the predicted cell consistency of the lithium-ion power battery system under test during the test period. While this invention offers advantages such as predicting battery consistency degradation at any time during the lifespan, ease of operation, high feasibility, early problem detection, reduced off-site failure rate of the battery system, significant savings in verification and maintenance costs, and consideration of the impact of actual operating conditions on battery consistency degradation, making it more practically valuable, it does not solve the aforementioned problems. Summary of the Invention

[0005] The present invention aims to overcome the problem that existing technologies cannot effectively calculate and predict the consistency of lithium iron phosphate batteries, and provides a cloud-based method for calculating and predicting the consistency of batteries.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A cloud-based method for calculating and predicting battery cell consistency includes the following steps: S1: Acquire data uploaded by the entire vehicle to the cloud platform, record the start charging time T0, and determine the vehicle charging status based on the data. If the conditions are met, proceed to S2; otherwise, continue executing S1; S2: Record the current battery cell voltage U. i and charging capacity Q i，y When the cell voltage U i >=U i When the charge is +x and the duration satisfies t, record the cell voltage and the current charging capacity Q. i，y+1 S3: Based on the current charging capacity Q i，y+1 Perform judgment and calculation. If the condition is met, proceed to S4; otherwise, continue to S2. S4: Calculate the minimum locked capacity of the battery cell Qchgmin, and calculate the capacity Qchgdiff of each battery cell that is smaller than the highest capacity battery cell based on the minimum locked capacity. i S5: Record the current Qchgdiff i Qchgdiff i K is obtained by calculating Qchgdiff of degree K-1. i The Kth predicted capacity difference value PQchgdiff is obtained. i K, according to Qchgdiff iK and PQchgdiff i K calculates and predicts the cell capacity difference PQchgdiff over a future period. i (K+1); S6: For the predicted cell capacity difference PQchgdiff i (K+1) makes a judgment. If the condition is met, an alarm is triggered and a warning message is pushed to the service personnel. If the condition is not met, S1 is executed. This invention proposes a cell consistency calculation and prediction method based on a cloud platform. By using 5G communication and a cloud platform, cell consistency calculation and prediction are performed. By calculating the reciprocal of the charging capacity corresponding to a certain voltage increment during charging, the charging capacity of a single cell is locked based on the maximum reciprocal. The capacity difference Qchgdiff between cells is obtained by comparing the locked charging capacities of different cells. By using historical vehicle data and the currently calculated Qchgdiff, Qchgdiff is intelligently predicted for a future period. If the predicted Qchgdiff value is greater than a threshold, the cloud platform sends a warning message to the vehicle service personnel for timely vehicle maintenance.

[0008] As a preferred embodiment of the present invention, the data uploaded by the vehicle to the cloud platform includes the voltage of each battery cell, the highest temperature, the lowest temperature, the vehicle status, and the vehicle charging capacity information.

[0009] As a preferred embodiment of the present invention, the specific process of judging the vehicle charging status based on data in S1 is as follows: judging whether the vehicle charging duration is ≥ TIMEchg, whether the minimum temperature is ≥ Tchgmin, and whether the maximum temperature is ≤ Tchgmax. If the conditions are met, proceed to S2; otherwise, continue to execute S1. Here, TIMEchg is the vehicle charging duration threshold, Tchgmin is the minimum temperature threshold, and Tchgmax is the maximum temperature threshold, which are set by the operator according to the vehicle condition.

[0010] As a preferred embodiment of the present invention, S3 includes the following steps: S31: Determine if y+1≥n, then let △Q i,b =Q i，y+1 -Q i，y+1-n , △DQ i,b =1 / △Q i,b Enter S32; otherwise, let △DQ i,b =0, continue executing S2; S32: Determine if △DQ i,b ≥△DQ i,b-1 Then let Qchg i =Q i，y+1 Otherwise, let Qchg i =Q i，y S33: Determine if the current charging time Tcur-T0 ≥ Tdelt1 and the current cell voltage > VOL1. If so, record △DQ. iIf the number of times >c is NUMDQ, proceed to S34; otherwise, continue to S2; S34: Record the charging time as T. 1i , judge NUMDQ i If the value is greater than or equal to Numv-z, proceed to S35; otherwise, continue with S2. S35: Determine the current charging time Tcur-T. 1i If ≥Tdelt2, then output record Qchg. i Otherwise, continue with S2. z is a value such as 0, 1, 2, 3... to prevent ΔVOL error from affecting the NUMDQ judgment. i The impact.

[0011] As a preferred embodiment of the present invention, Numv = ΔVOL / x, where Numv is the number of voltage increments satisfying x, and ΔVOL is the difference between two voltage plateau regions of the battery cell in the charging state.

[0012] As a preferred embodiment of the present invention, S4 specifically involves: calculating the minimum locking capacity of the battery cell Qchgmin = MIN(Qchg i The smaller capacity of each cell compared to the highest capacity cell is Qchgdiff. i =Qchg i -Qchgmin. According to Qchgdiff i The value indicates the capacity difference between cells, thus reflecting the consistency between cells.

[0013] As a preferred embodiment of the present invention, in S5, according to Qchgdiff i K and PQchgdiff i K calculates and predicts the cell capacity difference PQchgdiff over a future period. i (K+1) Specifically, calculate the prediction gain ZK = Qchgdiff. i K-PQchgdiff i K, then the cell capacity difference PQchgdiff in the future period i (K+1)=f(Qchgdiff i K, ZK, JTime). (via parameter Qchgdiff). i K, ZK, and JTime are used to predict the cell capacity difference over a future period. There are many prediction functions, such as neural networks and Kalman filters, all of which can be used to achieve the prediction of cell capacity difference in this invention.

[0014] As a preferred embodiment of the present invention, S6 specifically involves: determining PQchgdiff. iIf (K+1) > Qset, an alarm is triggered and a warning message is sent to service personnel. If not, S1 is executed. Future capacity differences are predicted based on historical capacity differences, and warnings are issued to service personnel in advance based on the predicted capacity differences to facilitate vehicle maintenance and avoid risks such as reduced range and compromised battery safety due to inconsistencies.

[0015] Therefore, the present invention has the following beneficial effects: it utilizes the large data volume and high-efficiency computing characteristics of cloud platforms to extract historical vehicle data and calculate capacity differences, predicts future capacity differences based on historical capacity differences, and issues early warnings to service personnel for vehicle maintenance based on the predicted capacity differences, avoiding the risk of reduced range and battery safety threats caused by inconsistency; the cloud platform performs capacity calculations and sends the calculation results to the corresponding vehicles, and the vehicles perform equalization processing based on the received data, reducing the computing intensity of the on-board hardware system and reducing dependence on high-performance on-board hardware systems. Attached Figure Description

[0016] Figure 1 This is a flowchart of the method of the present invention;

[0017] Figure 2 This is the OCV-SOC curve of lithium iron phosphate;

[0018] Figure 3 This is a charging voltage-SOC curve diagram of an embodiment of the present invention;

[0019] Figure 4 This is a graph showing the relationship between charging voltage and ΔDQ in an embodiment of the present invention;

[0020] Figure 5 This is a Qchg curve diagram of three battery cells according to an embodiment of the present invention;

[0021] Figure 6 This is a flowchart of a method according to an embodiment of the present invention. Detailed Implementation

[0022] The present invention will now be further described with reference to the accompanying drawings and specific embodiments.

[0023] Currently, the main methods for calculating cell consistency are the OCV-SOC (Open Circuit Voltage, OCV; State of Charge, SOC) lookup table method and the real-time single-cell capacity calculation method. The OCV-SOC lookup table method obtains the SOC of each cell string by looking up the OCV-SOC table using the terminal voltage after resting, and calculates the SOC difference between cells in each string to determine cell consistency. Because lithium iron phosphate batteries have a plateau region, such as... Figure 2As shown, in the plateau region, the SOC cannot be accurately obtained by looking up the OCV-SOC table using the terminal voltage, making this method unsuitable for lithium iron phosphate cells. The real-time single-cell capacity calculation method involves the on-board controller calculating the remaining capacity of each cell string in real time, and then determining the inter-cell differences based on the remaining capacity differences between each string. This method is applicable to different types of batteries, but calculating the remaining capacity of each cell string for hundreds or more battery packs requires significant memory resources, placing memory requirements on the controller selection and increasing the hardware cost of the on-board controller.

[0024] like Figure 1 As shown, this invention proposes a cloud-based method for calculating and predicting battery cell consistency. It utilizes 5G communication and a cloud platform (cloud platform: providing cloud-based services for developers to use when creating applications) to perform battery cell consistency calculations and predictions. By calculating the reciprocal of the charging capacity corresponding to a certain voltage increment during charging, the maximum reciprocal is used to lock the charging capacity of a single battery cell. The difference in capacity between cells, Qchgdiff, is obtained by comparing the locked charging capacities of different cells. Using historical vehicle data and the currently calculated Qchgdiff, the method intelligently predicts Qchgdiff for a future period. If the predicted Qchgdiff value exceeds a threshold, the cloud platform sends an early warning to vehicle service personnel, enabling timely vehicle maintenance.

[0025] Example: Figure 6 The flowchart shown in this embodiment illustrates the method of the present invention, which includes S1: acquiring data uploaded from the entire vehicle to the cloud platform. This data includes the voltage of each battery cell string, the highest temperature, the lowest temperature, the vehicle status, and the vehicle's charging capacity. Based on the acquired cloud platform data, the start charging time is recorded as T0. The system determines that the vehicle charging duration is ≥10 minutes, the lowest temperature is ≥0℃, and the highest temperature is ≤55℃. If these conditions are met, proceed to the next step; otherwise, continue with step S1. S2: processing each battery cell string and recording the current cell voltage as U. i Where i = 1, 2, 3...k, k is the total number of battery cells in series, U i Let Q be the current voltage of the i-th cell in the string, and record the current charging capacity Q. i，y Where y is the number of records, initially set to 1, and Q i，y Let U be the current charging capacity of the i-th cell in the y-th record, when the cell voltage U i >=U i When the charge level increases by 1 and the duration reaches 5 seconds, record the cell voltage and the current charging capacity as Q. i，y+1 S3 includes the following steps: S31: Determine ΔQ when y+1≥n. i,b =Q i，y+1 -Q i，y+1-n , where △Q i,b The charging capacity increment ΔDQ corresponds to the increase of n*x in the voltage of the i-th cell string during charging.i,b =1 / △Q i,b Where b is the number of records, initially set to 1, and △DQ i,b The voltage increase during charging is n*x, which is the reciprocal of the increase in charging capacity. Figure 4 The figure shows the charging voltage and ΔDQ of the first battery cell in this embodiment; otherwise, ΔDQ i,b =0, return to execute S2; S32: determine if △DQ i,b ≥△DQ i,b -1, if so, then Qchg i =Q i，y+1 Qchg i The locked charging capacity value for the i-th cell string is, for example... Figure 5 The diagram shows the charging capacity locked during the charging process of 3-cell batteries; otherwise, Qchg... i =Q i，y S33: Determine whether the current charging time Tcur-T0≥Tdelt1=10min is true, where Tdelt1 is the first time judgment threshold, which is 10min in this embodiment, Tcur is the current charging time, and the current cell voltage is >VOLl=3300mV, where VOLl is the voltage judgment threshold, which is 3300mV in this embodiment. Record △DQ. i If c = 1, perform NUMDQ; otherwise, return to execute S2, where c is the number of times △DQ is determined. i The threshold, △DQ i S34: Record the charging time as T, where n*x is the reciprocal of the charging capacity increment corresponding to the voltage increase of the i-th cell during charging; 1i , judge NUMDQ i The ≥Numv-z criterion is used to determine whether the condition is true, where Numv = ΔVOL / x, Numv is the number of voltage increments x that satisfy the condition, ΔVOL is the difference between two voltage plateaus in the charging state of the battery cell, and z is a value such as 0, 1, 2, 3, etc., to prevent errors in ΔVOL from affecting the NUMDQ criterion. i The impact of NUMDQ i For the i-th cell in the string to satisfy ΔDQ i > the number of times c, Figure 3 For the charging voltage and SOC curve, if they meet the requirements, proceed to the next step; otherwise, return to execute S2. S35: Determine the current charging time Tcur-T. 1i ≥Tdelt2=10min, where Tdelt2 is the second time judgment threshold, which is 10min in this embodiment. If the condition is met, the charging capacity value Qchg is output and locked. i If the condition is not met, return to execute S2. S4: Calculate Qchgmin = MIN(Qchg iQchgmin is the minimum locked charging capacity for all cells in the series, and Qchgdiff is the minimum locked charging capacity for all cells in the series. i =Qchg i -Qchgmin, Qchgdiff i The capacity difference is the smaller capacity of the i-th cell compared to the highest capacity cell, according to Qchgdiff. i The value indicates the capacity difference between battery cells, thus reflecting the consistency between them. This value is transmitted via the mobile network. i The vehicle's infotainment system receives the Qchgdiff data. i Perform equalization processing. S5: Record the current Qchgdiff. i Qchgdiff i K, where K is the exponent, such as 1, 2, 3..., is obtained by calculating Qchgdiff for exponent K-1. i The Kth predicted capacity difference value PQchgdiff is obtained. i K, based on the actual calculation of Qchgdiff i K and predicted value PQchgdiff i K, calculate the prediction gain ZK = Qchgdiff i K-PQchgdiff i K, then calculate Qchgdiff based on the total K times. i ZK, different Qchgdiff i The cell capacity difference PQchgdiff is calculated over a future period of time PTime (7 days) using the interval JTime. i PQchgdiff i (K+1)=f(Qchgdiff i K, ZK, JTime). S6: Determine PQchgdiff i If (K+1) > Qset, an alarm is triggered and a warning message is pushed to the service personnel. If not, return to execute S1. Qset is the capacity difference threshold, which is 10% of the rated cell capacity in this embodiment.

[0026] This invention utilizes the large data volume and high-efficiency computing characteristics of cloud platforms to extract historical vehicle data and calculate capacity differences. By predicting future capacity differences based on historical capacity differences, it issues early warnings to service personnel for vehicle maintenance, avoiding the risks of reduced range and compromised battery safety due to inconsistencies. The cloud platform performs capacity calculations and sends the results to the corresponding vehicles, which then perform equalization processing based on the received data, reducing the computational intensity of the onboard hardware system and decreasing reliance on high-performance onboard hardware systems.

[0027] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions conceived without creative effort should be included within the scope of protection of the present invention.

Claims

1. A cloud-based method for calculating and predicting cell consistency, characterized in that, Including the following steps: S1: Obtain data uploaded by the entire vehicle to the cloud platform and record the start time of charging. The charging status of the vehicle is judged based on the data. If the condition is met, the process proceeds to S2; otherwise, it continues to execute S1. S2: Record the current cell voltage and charging capacity When the cell voltage And when the duration is t, record the cell voltage and the current charging capacity. ; S3: Based on the current charging capacity Perform judgment and calculation; if the condition is met, proceed to S4; otherwise, execute S2. S31: If ,but , Enter S32; Otherwise Execute S2; S32: If ,make ;otherwise ; S33: Determine the current charging time And the current cell voltage If so, record Number of times Proceed to S34; otherwise, execute S2. S4: Calculate the minimum locking capacity of the battery cell Calculate the capacity of each cell that is smaller than the highest capacity cell based on the minimum locked capacity of the cell. ; S5: Record the current status for By calculating the K-1 degree Obtain the Kth predicted capacity difference value ,according to and Calculate and predict the cell capacity difference over a future period ; S6: For the predicted cell capacity difference The system makes a judgment. If the conditions are met, an alarm is triggered and a warning message is sent to the service personnel. If the conditions are not met, S1 is executed.

2. The cell consistency calculation and prediction method based on a cloud platform according to claim 1, characterized in that, The data uploaded by the vehicle to the cloud platform includes the voltage of each battery cell, the highest temperature, the lowest temperature, the vehicle status, and the vehicle's charging capacity.

3. A cell consistency calculation and prediction method based on a cloud platform according to claim 1 or 2, characterized in that, The specific process of judging the vehicle charging status based on data in S1 is as follows: judging whether the vehicle charging duration is sufficient. Is the lowest temperature? Is the highest temperature... If the condition is met, proceed to S2; otherwise, continue executing S1.

4. The cell consistency calculation and prediction method based on a cloud platform according to claim 1, characterized in that, The ,in, To satisfy the number of voltage increments x, This represents the difference between the two voltage plateau regions of the battery cell during the charging state.

5. The cell consistency calculation and prediction method based on a cloud platform according to claim 1, characterized in that, Specifically, S4 involves calculating the minimum locking capacity of the battery cell. The smaller capacity of each cell compared to the highest capacity cell .

6. The cell consistency calculation and prediction method based on a cloud platform according to claim 5, characterized in that, According to the S5 and Calculate and predict the cell capacity difference over a future period Specifically, this involves calculating the prediction gain. The difference in cell capacity in the future period , For different Interval time.

7. The cell consistency calculation and prediction method based on a cloud platform according to claim 1, characterized in that, Specifically, S6 involves: determining... If the conditions are met, an alarm will be triggered and a warning message will be sent to the service personnel; if the conditions are not met, S1 will be executed.

Images